BaanERP Tools

Technical Manual
A publication of:
Baan Development B.V.
P.O.Box 143
3770 AC Barneveld
The Netherlands

Printed in the Netherlands

© Baan Development B.V. 1998.
All rights reserved.
The information in this document
is subject to change without
notice. No part of this document
may be reproduced, stored or
transmitted in any form or by any
means, electronic or mechanical,
for any purpose, without the
express written permission of
Baan Development B.V.

Baan Development B.V.

assumes no liability for any
damages incurred, directly or
indirectly, from any errors,
omissions or discrepancies
between the software and the
information contained in this
document.
Document Information
Code: U7040C US
Group: User Documentation
Edition: C
Date: October, 1998
Table of contents

1 Installing BaanERP Windows (BW) 1-1

Hardware and software requirements 1-1
Installing the BaanERP Windows (BW) software 1-2
Installation directory of BaanERP Windows (BW) 1-9
The BECS utility 1-10
Functionality of BECS 1-10
Registering the BECS utility 1-10
2 BI server 2-1
Introduction 2-1
Architecture 2-1
Requirements for installation 2-4
Installing the BI server 2-5
Moving the BI server to another system 2-7
Configuring the BI server 2-7
Starting the BI server 2-7
BI server as a Windows NT Service 2-9
Adjusting the BI HTML-page 2-10
Client issues 2-11
3 Database tools 3-1
General 3-1
bdbpre 3-1
bdbpost 3-6
bdbreconfig 3-10
refint 3-14
4 Database management 4-1
General 4-1
Accessing tables 4-1
tabledef6.2 4-1
fd6.2.package combination 4-2
Example table access (table and Bshell on the same system) 4-2
Database drivers 4-2
Accessing tables on other systems 4-3
Audit trail 4-3
Database mirroring 4-3
Alternative databases 4-5

Technical Manual
i
Table of contents

Communication protocols 4-5

Socket protocol 4-7
Local sockets 4-7
Remote sockets (only for database servers) 4-7
Pipes protocol 4-8
Message queue protocol 4-8
Example ipc_info file 4-9
5 Native language support 5-1
General 5-1
Composed characters 5-1
Conversion tables 5-2
Input conversion table 5-2
Output conversion table 5-2
Example of total conversion 5-3
NLS editor 5-4
Initial screen options 5-5
Maintenance of NLS conversion tables 5-5
NLS-related files 5-7
Terminal information file 5-7
Printer information file 5-8
Input and output conversion table 5-8
Terminal setup 5-9
Character sets and fonts 5-9
Sort tables 5-9
Shift tables 5-10
ISO 8859-1 character set (0-127) 5-11
DEC vt100/vt200 character set (0-127) 5-13
DEC vt100/vt200 character set (128-255) 5-15
List of compose sequences 5-17
Input/output table vt200 5-20
Output table 5-24
Printer output table MT910 HP Prestige character set 5-25
Overview of Esc/Ctrl codes 5-29
6 Information files 6-1
Terminal information file 6-1
Boolean entries 6-1
Numbers 6-2
Strings 6-3
Keys 6-9
Color support 6-11
Sending colors separately 6-11

Technical Manual
ii
 Table of contents

Colors in combination with code features 6-13

Parameterized strings 6-14
Example of terminal information files 6-16
Printer information files 6-21
Bar codes 6-26
Example of printer information file for mt910 printer 6-29
7 User Interface (UI) Page mode 7-1
Tab processing 7-2
Default Button handling 7-2
Event processing 7-2
How to select the UI Page mode 7-3
How to mark UI objects as synchronizing fields 7-4
8 Customer support tools 8-1
General 8-1
Error logging 8-1
Log file 8-1
Log file layout 8-2
Log file maintenance 8-3
Submitting errors to customer support 8-3
Example error logging 8-4
9 Executable programs 9-1
Database management 9-1
General 9-1
Oracle 9-2
Sybase 9-2
Informix 9-2
DB2 9-3
Logic server (Bshell) 9-3
Bshell6.2 9-3
Debugging the Bshell during run time 9-4
Memory usage 9-9
Miscellaneous 9-10
bshcmd6.2 9-10
badmin6.2 9-13
Installation 9-13
Development 9-14
License management 9-14
licd6.2 9-14
licmon6.2 9-15
Printer management 9-16
Shared memory 9-17

Technical Manual
iii
Table of contents

Network 9-17
TRITON Super Set 9-18
Time zone utilities 9-20
Middleware 9-20
UNIX equivalents 9-20
Miscellaneous 9-21
10 Errors 10-1
General 10-1
UNIX errors 10-1
Database errors 10-4
11 Shared memory management 11-1
General 11-1
Using the shared memory manager 11-1
Installation of shared memory 11-2
Create an entry in the file shm_param file 11-2
Determine shared-memory parameters 11-3
Using shmvalues6.2 11-3
Using shmmanager6.2 11-4
Fill the entry in the shm_param parameter file 11-5
Installation of shared memory 11-6
Starting shmtimer6.2 11-6
Error messages during installation 11-7
Syntax srdd_tab 11-9
Filling shared memory 11-10
Kernel requirements 11-11
General description of adjusting UNIX kernels 11-11
Kernel parameters 11-12
General kernel parameters 11-17
Example parameter file shm_param 11-18
12 Remote databases 12-1
General 12-1
Settings for remote database configurations 12-2
Local host 12-3
Remote 12-4
Local 12-4
tabledef6.2 12-5
User files 12-5
Remote 12-7
Local and remote tables 12-8
Settings for remote menus/forms/objects 12-8

Technical Manual
iv
 Table of contents

13 Multibyte management 13-1

General 13-1
Triton Super Set 13-1
Using multibyte character sets 13-3
Installation 13-3
Printing multibyte characters 13-3
Utilities 13-3
tsscomp6.2 13-3
tsscvt6.2 13-4
tssinfo6.2 13-5
Character set description 13-7
Locale information 13-9
Storage of multibyte characters 13-10
14 Audit management 14-1
General 14-1
Storage of audit data 14-1
Sequence files 14-1
Information file 14-2
Location of audit files 14-2
Other parameters of audit files 14-3
Audit management 14-4
Table data dictionary as seen by audit server 14-5
Commands logged by audit server 14-6
Data stored by audit server 14-6
Format of audit row 14-7
Examples of the audit row for various operations 14-9
Limit on the size of audit data 14-10
Audit server 14-10
Long transactions and overflow file 14-10
Sequence termination 14-11
Sequence file maximum size reached 14-11
Table audit data dictionary changed 14-11
Sequence terminated by user 14-13
Transaction termination 14-13
Reusing sequence files 14-16
File locking 14-16
Limit on open files 14-17
Opening large number of tables in single session 14-18
Multisession support 14-18
File security for various audit management files 14-18
Audit server debugging options 14-19
Audit errors 14-19

Technical Manual
v
Table of contents

Authorizations 14-20
Format of audit files 14-22
auditdef6.2 14-22
Information file 14-23
Sequence file 14-26
audit_spec file 14-31
15 OLE Automation 15-1
General 15-1
BaanERP as OLE Automation server 15-2
Restrictions 15-4
Example: Import BaanERP users 15-4
DLL function example 15-5
Visual Basic example 15-6
Example: Using BaanERP SQL 15-7
DLL information 15-7

Technical Manual
vi
About this document

This document is a reference guide explaining how you can customize the
BaanERP application for use by specific organizations. It is intended for use by
BaanERP application developers and application managers. You can use this
reference guide to set up and maintain a BaanERP application.
In this document you will find information about database tools and
management, audit management, native language support, and terminal
information, as well as some limited installation instructions.
Chapter 1, “Installing BaanERP Windows (BW),” describes how to install
BaanERP Windows (BW). This assumes that you already have BaanERP
installed on the server.
Chapter 2, “BI server,” summarizes how to install the BaanERP Internet Server
application, which is used as a gateway between the BaanERP Internet client
(BI) and the BaanERP application server (Bshell), to expose the BaanERP
functionality to the Internet domain.
Chapter 3, “Database tools,” describes tools you can use for the following tasks:
n Convert a database table to a sequential file (bdbpre)
n Create a database table from a sequential dump or append data to an existing
database table (bdbpost)
n Reconfigure a database table according to a new data dictionary
(bdbreconfig)
n Check or repair the referential integrity within the database (refint)
Chapter 4, “Database management,” discusses the working and usage of several
protocols that can be used for local and remote communication.
Chapter 5, “Native language support,” describes the communication between the
character set of a terminal and that of Bshell6.2.
Chapter 6, “Information files,” describes terminal information files, which store
information such as the type of terminal, terminal operations, how to access keys
on the keyboard, color support, cursor movements, and whether code features are
blinking, bold, or reverse.

Technical Manual
vii
About this document

Chapter 7, “User Interface (UI) Page mode,” describes the User Interface (UI)
Page mode in BaanERP. In normal mode, the 4GL Engine validates the entered
data per field. In the UI Page mode, the 4GL Engine validates data per page
rather than per field. The synchronous interaction between the UI driver and the
Bshell is therefore significantly reduced, which improves response times.
Chapter 8, “Customer support tools,” describes two functions used for solving
errors in a Bshell environment: error logging and the bserel6.2 program. Error
logging is keeping a record of error messages in a log file. The program bserel6.2
provides information on the BSE environment as installed on your system.
Chapter 9, “Executable programs,” describes all of the executable programs used
for database management.
Chapter 10, “Errors,” describes all the errors that can occur when you are
working in the BaanERP environment, and offers a solution.
Chapter 11, “Shared memory management,” discusses shared memory, which is
a part of the internal memory intended for common usage.
Chapter 12, “Remote databases,” explains how the BaanERP client/server
architecture enables the user to work with databases distributed over one or more
systems.
Chapter 13, “Multibyte Management,” explains how to use the 32-bit wide
character space, which has been implemented in BaanERP Tools to support all
possible character sets.
Chapter 14, “Audit Management,” describes the audit management facility
including the audit server, the audit management utility, and where audit data is
stored.
Chapter 15, “OLE Automation,” describes how BaanERP works with OLE
Automation, an industry standard that applications use to expose their OLE
objects to development tools, macro languages, and other applications that
support OLE Automation.

Technical Manual
viii
1 Installing BaanERP Windows (BW)

This chapter describes how to install BaanERP Windows (BW). It assumes that
BaanERP is already installed on the server. If this is not the case, refer to the
database-specific installation guide.
BaanERP Windows (BW) is a graphical user interface (GUI) program in
BaanERP. BW runs on an Intel compatible PC in a Windows 95 or Windows NT
environment. BW displays BaanERP as an application with the same look and
feel as a native Windows application.
This chapter describes:
n The hardware and software requirements
n How to install the BaanERP Windows (BW) software
n The installation directory of BaanERP Windows (BW)
n The BaanERP Windows (BW) environment variables

Hardware and software requirements

The following software and hardware requirements must be met before you can
install BaanERP Windows (BW).
n The operating system Windows 95 or Windows NT Server version 4.0 must
be installed and running.
n The system on which BaanERP Windows is to be installed, must be at least
an Intel-based system with a 50 MHz 486DX processor. Furthermore,
BaanERP Windows requires at least a SVGA video display, 16 MB of main
memory (RAM), and 8 MB of free disk space.

Technical Manual
1-1
Installing BaanERP Windows (BW)

Installing the BaanERP Windows (BW)

software
To install the BaanERP Windows (BW) software, you can follow one of these
procedures:
n If you have a UNIX BaanERP server, the BaanERP Windows (BW) software
can be copied from the directory $BSE/mswindows/Bw5.xx on the server to
a temporary directory on the client by using ftp. Double-click the file
setup.exe. This starts the BW installation program.
n If the client has a local CD-ROM drive, you can install BaanERP Windows
(BW) directly by using the BaanERP distribution CD-ROM.
n If the client has no CD-ROM drive, you can share the CD-ROM of the server.
NOTE The description below assumes that your client has a local CD-ROM drive and is
running the Windows NT operating system.
Take the following steps to install BaanERP Windows.
1 Close all running applications and insert the BaanERP distribution CD-ROM
in the CD-ROM drive.
2 A Welcome message is displayed. Choose Next.
3 A Licence Agreement message screen is displayed. Choose Yes.
4 The Select environment dialog box is displayed.

Technical Manual
1-2
 Installing BaanERP Windows (BW)

Click Install new environment if you perform a fresh installation, and enter
a name for the new installation. If you want to upgrade an existing BW
installation, you must click Select existing environment to choose the
existing environment to be upgraded. Choose Next to continue.
5 The Choose Destination Location dialog box is displayed. The layout of the
dialog box depends on whether you are installing BW on a PC client or a
Windows NT server.
If you are installing on a PC client, the following dialog box is displayed:

Technical Manual
1-3
Installing BaanERP Windows (BW)

If you are installing on a Windows NT server, the following dialog box is

displayed:

Accept the default path or choose Browse to enter another path. Choose Next
to continue.
6 The Setup Type dialog box is displayed:

Technical Manual
1-4
 Installing BaanERP Windows (BW)

You can select the following server types:

− User Interface
This type is a BaanERP Windows (BW) client.
− User Interface and Logic
This type is a stripped version of a BaanERP Windows NT server (with
local application logic).
− Full installation (including application files and database)
This type is the complete BaanERP Windows NT server version.
7 Select User Interface and choose Next. The Select Program Folder dialog
box is displayed. Select the Program Folder to which the BaanERP Windows
icons must be added (default BaanERP) and choose Next.
8 The Setup Information dialog box is displayed. Choose Next to continue
with the displayed setup (or choose Back to change the displayed setup).
9 After the software is installed you must choose OK to exit the installation
script. The BaanERP program folder is displayed. Double-click the BW
Configuration Tool icon. The BW Configuration dialog box is displayed.

10 Supply the following information.

On the Application Server tab you must fill in the following boxes:
− Hostname
Enter the host name of the BaanERP server.

Technical Manual
1-5
Installing BaanERP Windows (BW)

− Connect As
Enter a valid user name as known on the BaanERP server.
− Ask password at startup
If this check box is selected, you can only supply the password during the
BaanERP login procedure.
− Use saved password
If this check box is selected, the BaanERP logon procedure is skipped,
and the saved password is used.
− Current user
Select this check box to make use of the Windows Integrated Security
feature. In this case you will use your current user identity to log on to the
BaanERP Windows NT server.
NOTE For Windows 95 the unified logon feature will only work if you are
actually logged on to Windows 95.
NOTE You can only make use of the Windows Integrated Security feature, if
your BaanERP server is a Windows NT server.
− BSE
Enter the path to the BSE environment on the BaanERP server (equals the
BSE environment variable on the BaanERP server).
− Bshell name
Enter the application logic to use on the remote BaanERP host (changing
the default of bshell will only be useful during debugging).
− Command
Enter the option you want to pass to the bshell. You can, for example,
define environment variables by using the following syntax:
-- -set Environment Variable1=xxx -set Environment Variable2=yyy

On the NLS tab, you must fill in the following box:

− Locale
Select the character set to be used (the default character set is
International [ISO]).
On the Font tab, fill in the following box:
− Font Type
Enter the desired font type. You can change the setting by choosing the
button at the right side of the font boxes.

Technical Manual
1-6
 Installing BaanERP Windows (BW)

On the Automation tab, you can fill in the following box:

− Class Name
This box contains the name of the Baan automation object
Baan.Application.server, where server is the hostname of the BaanERP
server you supplied on the Application Server tab.
NOTE Choose Save As on the Application Server tab to save the new configuration
in a different configuration file. However, only one configuration can be
active per BaanERP session.
11 Choose RUN! to start your BaanERP session.

NOTE The BaanERP .bwc files are registered by the Windows NT or Windows 95
operating system. You can find the .bwc files in the directory Lib\User under
the installation directory of the BaanERP client software. This implies that
you can start a BaanERP logon procedure by double-clicking the icon in front
of the relevant .bwc file in the File Manager. You can also make BaanERP
program icons by dragging the .bwc configuration files to a Program Group.

Technical Manual
1-7
Installing BaanERP Windows (BW)

12 From the BaanERP Tools menu, choose Software InstallationÐ

Installation. Open the Download Help Files (ttadv8245m000) session. This
session has two options:
− Download help files
This command installs the windows Help files on your client. Select the
packages for which you want to see Help and choose Download Help
Files from the Specific menu.
− Edit HTML Help transfer script
When you install HTML help on an HTTP server, you must edit the
HTML Help transfer script. This script is used for transferring the HTML
help files from the application server to the HTTP server. To execute this
script, you need access to the HTTP server.

Technical Manual
1-8
 Installing BaanERP Windows (BW)

Installation directory of BaanERP Windows

(BW)
The installation directory C:\Program Files\Baan\BW5 is created during the
setup process of BaanERP Windows (BW). During setup you can change this
installation directory. This section assumes you have chosen the default name for
the installation directory of BaanERP Windows (BW).
The following subdirectories are created in directory C:\Program
Files\Baan\BW5 during setup:
n C:\Program Files\Baan5\BW\bin
This directory contains the bw.exe program to start the BaanERP application.
It also contains the bwconfig.dll for the configuration program.
n C:\Program Files\Baan5\BW\lib
This directory contains the files bw.ini (initialization file), bw5.reg (object
registration file), bw.tlb (type library information for OLE Automation), and
tss_bw (language support).
n C:\Program Files\Baan\BW5\log
This directory contains logging information about BaanERP Windows.
n C:\Program Files\Baan\BW5\samples
This directory contains a Microsoft Excel spreadsheet, which is an example
of the use of OLE Automation to access BaanERP dynamic link libraries.
n C:\Program Files\Baan\BW5\tmp
This directory contains the temporary files.
n C:\Program Files\Baan\BW5\help
This directory contains the Help files.
The following subdirectories are created in the directory C:\Program
Files\Baan\BW5\lib.
n C:\Program Files\Baan\BW5\lib\nlsinf
This directory contains the special character set file xwindows.in.
n C:\Program Files\Baan\BW5\lib\user
This directory contains the configuration files.

Technical Manual
1-9
Installing BaanERP Windows (BW)

The BECS utility

If you want to use both a BAAN IV and a BaanERP server from a single
workstation, you must install two different BW versions on your workstation.
Before the BECS utility was available, this generated the problem that both BW
versions wanted to be the owner of the BW configuration files (the files with
extension .bwc, in which your configuration settings are stored). Therefore, it
was unpredictable which BW version would be started if you double clicked a
.bwc file.
This problem is solved by the BECS utility: it enables you to run different BW
versions on a single workstation.

Functionality of BECS
The BECS utility offers the following functionality:
n It takes the ownership of files with the extension .bwc. By doing this, it
resolves the potential conflict described above. After BECS is correctly
installed, you will start BECS by double clicking on a .bwc file. Based on the
location of your .bwc file, BECS will locate the associated BW version, and
start this BW version accordingly. For this, your .bwc file must be stored in
the lib\user subdirectory of the directory where you installed the BW version.
BECS will start bin\BW.exe in this same directory.
n It can show you the different BW-versions installed on your system. Once the
BW versions have been registered, you will get an overview of all the .bwc
files for the BW versions if you start BECS. This enables you to go to the
configuration screen for each of the .bwc files, and to launch BW using the
selected configuration screen.

Registering the BECS utility

To make BECS.exe the owner of all .bwc files, you must follow the steps below:
1 Copy BECS to the location you want to have it, for example,
C:\Program Files\Baan\BECS.exe.
2 In the same directory where you found this file you can find the file
BECS.reg. At the end of this file, you will find the following lines:
HKEY_CLASSES_ROOT\BW_Configuration\shell\Open\command =
"$(BSE)\bin\BECS.exe" "%1"
HKEY_CLASSES_ROOT\BW_Configuration\shell\Config\command =
"$(BSE)\bin\BECS.exe" -config "%1"

Technical Manual
1-10
 Installing BaanERP Windows (BW)

3 Please edit this file, and replace the portion $(BSE)\bin\BECS.exe by the
path of your choice, for example, C:\Program Files\Baan\BECS.exe.
4 Execute the file BECS.reg by double clicking it. This will make BECS.exe
the owner of all .bwc files.

Registering BW environments
To let BECS.exe show all available .bwc files, you must create registry keys for
all BW environments. You can do so by following the steps below:
1 Start regedit
2 Go to HKEY_LOCAL_MACHINE\SOFTWARE\Baan
3 Create a new key for each BW environment you want to register, for
example, both Baan4 and Baan5. Within each key you must add a new string
value named BSE, containing the patname to the directory where you
installed the BW version. For example, the settings could look like the ones
below:
[HKEY_LOCAL_MACHINE\SOFTWARE\Baan\Baan4]
"BSE"="C:\\Program Files\\Baan\\Bw4"
[HKEY_LOCAL_MACHINE\SOFTWARE\Baan\Baan5]
"BSE"="C:\\Program Files\\Baan\\Bw5"

Technical Manual
1-11
Installing BaanERP Windows (BW)

Technical Manual
1-12
2 BI server

Introduction
The BaanERP Internet Server ap
plication (BI server) is used as a gateway between the BaanERP Internet client
(BI) and the BaanERP application server (Bshell), to expose the BaanERP
functionality to the Internet domain.

Architecture
The architecture of the BI server is shown in the following figure:

Figure 1, BI server architecture

The following components are included in the architecture:

n A Bshell running on an application-server system. This is a normal BaanERP
application server to which both BaanERP Windows (BW) clients and BI
applets can connect. For each client (BI or BW) there is always one
connection with a Bshell.
n An HTTP-server system, on which an HTTP-server application runs, and one
or more BI server applications. This setup is used as a gateway between the
BaanERP applications and Internet browsers. BI applets always connect to
the Bshell through the BI server. There is never a direct connection between
Bshell and BI applet.

Technical Manual
2-1
BI server

One of the HTML pages, which is managed by the HTTP server, contains a
reference to the BI applet. When a browser performs a GET for this page, the
BI applet is downloaded to the requesting browser. The HTTP-server system
and the application-server system can be one system or they can be two
distinct systems.
n A client system running a Java-enabled HTML browser. The BI applet runs
within this browser. The BI applet establishes a direct TCP/IP connection
with the BI server running on the HTTP-server system or on another system.
The following figure shows a timing diagram for setting up the connection
between a BI applet and a Bshell using the BI server.

Figure 2, Connection using the BI server

Technical Manual
2-2
 BI server

Security is important when you offer application functionality over the Internet.
Security provision is available at three levels:
1 The HTTP server security mechanisms can be used to offer secure access to
the HTML page containing the BI applet.
2 Secure communication between the BI server and the BI applet is supported
using the Secure Sockets Layer (SSL) v3.0 protocol.
3 Application-server system authentication.
The first level can be used to restrict access to the BI applet to a set of registered
users. This feature is completely dependent on the HTTP server and is not
described here.
The second level offers secure communication between the BI server and the BI
applet. At this level the SSL protocol is supported. The SSL protocol provides
connection security that has these basic properties:
n The connection is private. Encryption is used after an initial handshake to
define a secret key. Symmetric cryptography is used for data encryption.
n The peer’s identity can be authenticated by using asymmetric or public key
cryptography.
n The connection is reliable. Message transport includes a message integrity
check using a message authentication code (MAC).
The BI server uses a CipherSuite that does not offer server-side certification or
client-side authentication. Furthermore, the CipherSuite uses exportable versions
of the cryptographic algorithms so that the software can be delivered to all
countries.
Note that Secure Sockets Layer (SSL) client-side certification is not used,
because this would require that each client get and store a certificate from a
certification authority. Instead, client authentication relies on the user name and
password validation used to connect to the application-server system. When you
use a Secure Sockets Layer (SSL) connection, this user name and password are
encrypted when they are transmitted from the BI applet to the BI server. In
addition, you can restrict access to the HTML-page containing the BI applet by
using the HTTP server security mechanisms, which might also require client
authentication.
The BI server can be configured to use:
n No Secure Sockets Layer (SSL) security
n Secure Sockets Layer (SSL) security

Technical Manual
2-3
BI server

The third level uses the authentication mechanism of the application-server

system. Each BI user provides a user name and password to log on to the
application-server system.

Configuration
The BI server allows for configuration of important communication parameters
by a system administrator. These configuration parameters include:
n Listener port
n Enable/disable Secure Sockets Layer (SSL) security
n Application-server system name or IP address
n Enable/disable logging of access history and errors
n When logging is enabled, the log file name

Requirements for installation

Before you start the installation, you must note the following requirements:
n Required related products
n Performance requirements
n Installation requirements
These requirements are explained in the following sections.

Required related products

The BI server depends on the following related products, which must be installed
on the HTTP-server system:
n A Java run time environment that conforms to the JDK 1.1.
n An HTTP server capable of downloading Java applets to a browser

Performance requirements
The number of clients that can be supported by a single BI server application is
not restricted by the BI server. However, the Java virtual machine or the
underlying operating system can limit the number of simultaneous clients
supported by one instantiation of the BI server.

Technical Manual
2-4
 BI server

Installation requirements
These are the requirements for installing the BI server and BI client components
on the HTTP-server system.
Before installing the BI components on the HTTP-server system, you must
install and configure the following third-party products:
n An HTTP server. The operating system of the HTTP-server system
determines which HTTP server must be installed. To run the HTTP server on
a Windows NT server, you must install Microsoft Internet Information Server
(IIS). If your server is running on a UNIX system you can, for example,
install Netscape Enterprise servers or Apache.
n A Java virtual machine compatible with Java 1.1. The operating system of the
HTTP-server system determines which Java virtual machine must be
installed. Contact your operating system supplier to get an implementation of
a Java virtual machine.

Installing the BI server

After you have prepared the HTTP-server system as described in the section
“Requirements for installation,” perform the following steps to install the BI
components on the HTTP-server system.
NOTE Both the HTTP-server system and the BaanERP application server can be UNIX
or Windows NT systems. The following example installation assumes that you
are installing on a UNIX platform.
1 Copy the BI installation file $BSE/internet/bisetup.class from the BaanERP
server to the HTTP-server system.
2 Start the BI installation program by entering the following on the command
line:
java bisetup

The java string is the name of the Java interpreter. This name is supplier-
dependent (for example, jview for the Microsoft Java virtual machine).

Technical Manual
2-5
BI server

3 The installation program asks for a destination directory for the BI

components. This must be a directory in the HTML tree that you can access
through the HTTP server (a directory below the document root directory of
the HTTP server). If this directory does not yet exist, it is created. After
completing the installation program, the directory structure on the target
system is similar to the following structure:
$PUBLIC_HTML/$TARGETDIR/client/ie4/bi.cab
/grid.cab
/ssl.cab
/netscape/bi.jar
/bi.html
/readme.txt
/images/logo.jpg
/server/*.class
In this structure you can access the $PUBLIC_HTML directory in the HTML
tree through the HTTP server, in a directory below the document root
directory of the HTTP server.
$TARGET_DIR is a directory below $PUBLIC_HTML where the BI
package has been installed.
4 At this point you must determine on which TCP port you want the BI server
to listen. Choose a free port.
5 Edit the HTML file that you can find at
$PUBLIC_HTML/$TARGETDIR/client/bi.html. All the parameters
mentioned in the section “Adjusting the BI HTML-page” must be filled in
according to the actual configuration.
6 Start the BI server program. See the section “Starting the BI server” for
details. The arguments must correspond with the values entered in the HTML
file.

Technical Manual
2-6
 BI server

Moving the BI server to another system

Some firewall configurations require that the BI server run on a system other
than the HTTP-server system. In this case you must perform the following steps
after you have performed steps 1-3 described in the section “Installing the BI
server.”
1 Install a virtual machine compatible with Java 1.1 on the system where you
want to run the BI server.
2 Copy the BI-server class files from the HTTP-server system to the same
system. These class files can be found on the HTTP-server system at:
$PUBLIC_HTML/$TARGETDIR/server. You must copy this complete
directory tree, preserving the directory structure.
3 Determine the TCP port on which you want the BI server to listen. Choose a
free port.
4 Edit the HTML file on the HTTP-server system, which you can find at
$PUBLIC_HTML/$TARGETDIR/client/bi.html. All the parameters
mentioned in the section “Adjusting the BI HTML-page” must be filled in
according to the actual configuration.
5 Start the BI server program. See the section “Starting the BI server” for
details. The arguments must correspond with the values entered in the HTML
file.

Configuring the BI server

This section describes the additional steps to configure the BI server after the
installation, and contains the following sections:
n Starting the BI server
n Adjusting the BI HTML-page

Starting the BI server

The following command shown starts the BI server, listening on one TCP/IP port
and forwarding incoming connections to the application-server system
appserver. The java string is the name of the Java interpreter. This name is
supplier dependent, for example, jview for the Microsoft Java virtual machine.
java BI server [–s ] [–l logfile] [–c] [–t level] –p port –a appserver

Technical Manual
2-7
BI server

Options
–s
When this option is specified, the Secure Sockets Layer (SSL) is enabled . When
this option is not specified, SSL is disabled. When security is enabled, the BI
applet must also use SSL for communication with the BI server. The BI applet
gets this information through the applet parameters in the HTML page. See the
parameter SECURITY in the section “Adjusting the BI HTML-page.”
–l logfile
When this command is specified, logging of access history and errors is
redirected to the file defined by the logfile placeholder. When this command is
omitted, logging information is sent to the standard output, the Java console. The
level of logging detail is set with the –t command.
–c
When this option is specified, the logging file indicated with the option –l is first
cleared, that is, truncated. If this option is not specified and the logging file
specified with the –l option already existed, all logging information is appended
to this file.
–t level
When this option is specified the level of detail for logging can be specified. A
higher level gives more detail. Level 0 disables logging. The default for this
option is level 1. The maximum level is 3.
–p port
This mandatory option is used to specify the port number on which this BI server
listens for incoming BI applet connections. Usually a port number higher than
1024 must be chosen, because lower numbers are reserved for system services.
The BI applet must also use this port number when connecting to the BI server.
The BI applets gets this port number via the applet parameters in the HTML
page. See the parameter PORT in the section “Adjusting the BI HTML-page.”
–a appserver
This mandatory parameter specifies the application-server system. The appserver
can be the hostname of the application-server system or its IP address.

Technical Manual
2-8
 BI server

BI server as a Windows NT Service

If you are running the BI server on a Windows NT system, you can install the BI
server as a Windows NT Service. Before you do this, make sure that you can
start the BI server in your environment by manually starting the server. To install
the BI server as an NT Service you must perform the following steps:
1 Install the Windows NT Resource Kit. This section assumes it is installed at
C:\ntres.
2 From the Windows NT Resource Kit, install the executable SRVANY.EXE
as an NT Service. You can do this by entering the following commands at the
DOS command prompt:
− C:\>cd \ntres
− C:\ntres>INSTSRV BiService1 c:\ntres\srvany.exe
3 Start the Control Panel and choose the Services applet. Now perform the
following steps:
− In the Services applet, choose BiService1
− Choose Startup
− In the Startup dialog box, select the startup type Automatic
− Select Log on as System Account. Do not allow this service to interact
with Desktop
− Close the Control Panel

4 Start a registry editor, for example, regedt32, and perform the following
steps:
− Look up HKEY_LOCAL_MACHINE\SYSTEM\
CurrentControlSet\Services\BiService1.
− Add another key under this key named: Parameters
− Add the values listed in the following table to this key:

Registry values
Value name Value type Value string

Application REG_SZ C:\winnt\jview.exe

AppParameter REG_SZ BI server –l <logfile> –c –a <hostname> –p
<portnum>
AppDirectory REG_SZ $PUBLIC_HTML\$TARGETDIR\server

Close the registry editor.

5 When you restart your Windows NT system, the service is started
automatically.

Technical Manual
2-9
BI server

If you want more than one BI server to be started as a Windows NT Service,

repeat steps 1 - 5, and use another name for BiService1.

Adjusting the BI HTML-page

The following example shows the content of an HTML-page loading the BI
applet in a browser.
<applet
archive="netscape/bi.jar"
code=BiLogon
width=400
height=200>
<param name=cabinets value="ie4/bi.cab, ie4/grid.cab, ie4/ssl.cab">
<param name=HOSTNAME value=””>
<param name=PORT value=2000>
<param name=SECURITY value=NONE>
<param name=BSE value=/usr/bse>
<param name=BSHELL value=bshell>
<param name=COMMAND value="">
</applet>
The parameters that must be passed to this applet are as follows:
n HOSTNAME
The value of this parameter is the TCP/IP hostname or IP address of the
system on which a BI server application is listening. When this parameter is
omitted or set to an empty string, the BI applet connects to the server, from
which the HTML-page containing the applet code was running. This is the
default value. However, some firewall configurations require that the BI
server runs on a system other than the HTTP-server system. See the section
“Moving the BI server to another system”.
n PORT
The value of this parameter is the port number on the HTTP server system on
which a BI server application is listening. This must correspond with the port
number passed to the BI server with the –p option. The BI applet always
connects to the server from which the HTML-page containing the applet code
was running. The default value is 2000.
n SECURITY
The value of this parameter is used to determine the security mode of the
connection between the BI applet and the BI server. If this value is “SSL,”
the BI applet starts a Secure Sockets Layer (SSL) connection with the BI
server. (The BI server must have been started with the –s option for the
related port).

Technical Manual
2-10
 BI server

If this value is “NONE” or this parameter is omitted, the BI applet starts a

normal TCP/IP connection with the BI server. (The BI server must have been
started without the –s option for the related port).
The default value is NONE.
n BSE
The value of this parameter is the BSE directory path on the application-
server system. It is used to set the correct Bshell environment. The default
path is /usr/bse.
n BSHELL
The value of this parameter is the Bshell name on the application-server
system. It is used to indicate which Bshell must be started (tag in
$BSE/lib/ipc_info). The default value is Bshell.
n COMMAND
The value of this parameter is the command string to be passed to the Bshell.
This can be the environment setting for the Bshell and the name of the first
session to be started. The default command is empty.

Client issues
At the time of writing this document, no detailed information about the supported
browsers at the client side could be given. For the most up-to-date information
about the supported browsers refer to the file
$PUBLIC_HTML/$TARGET_DIR/client/readme.txt, installed on the HTTP-
server system.

Technical Manual
2-11
BI server

Technical Manual
2-12
3 Database tools

General
The database tools are designed for database management, and are independent
of how the database is organized. At run time the system selects a database driver
depending on the contents of the file tabledef. You can then use these database
tools to:
n Convert a database table to a sequential file (bdbpre)
n Create a database table from a sequential dump or append data to an existing
database table (bdbpost)
n Reconfigure a database table according to a new data dictionary
(bdbreconfig)
n Check or repair the referential integrity within the database (refint)
If there are changes in the data dictionary, for example, if fields are added or
field lengths are changed, you can reconfigure the old table to a new one. The
data dictionary changes are incorporated in the new table, which also contains
the data dictionary table data.
The system you use determines the extension for the database tool name. For
example, bdbpre used on Windows NT is called bdbpre.exe, and on UNIX it is
called bdbpre6.2.
The following is a description of each of the database tools.

bdbpre
NAME
bdbpre
Convert database tables to a sequential file.
SYNOPSIS
bdbpre [–q output file] [–uUvVsrxyzkK][–p pack_comb][–t sep][–o dir]
[–d driver type][–N table [table ..]][–I file][–E file][–O file]
[–C company number list/range]

Technical Manual
3-1
Database tools

DESCRIPTION
The bdbpre tool reads tables for given company numbers, converts them into
sequential data and writes them to standard output. This output can be redirected
to a file that can be used as input for bdbpost.
The bdbpre tool prints information such as names, the number of records, and
any errors. You can use the –s option to suppress the messages produced by
bdbpre at run time. This option is useful when you use the output of bdbpre as
the direct input to bdbpost.
The file fd6.2.package combination is always used for searching dictionary
information. The following examples use $BSE/lib/tabledef6.2 to get
information about the database type and driver. If you want to use a specific
database driver for a particular conversion, you can use the –d option. If you use
this option, tabledef6.2 is not read.
You can use bdbpre in the following ways:
n You can convert a database table for more than one company number by
specifying a range of company numbers. For example:
bdbpre –Ntimcs099 –C000-005 > timcs099_dump
n You can convert a database table for given company numbers. For example:
bdbpre –Ntimcs088 –C000 004 005 > timcs088_dump
n You can specify all the tables for which you want to create a sequential dump
in some file in package module format. The –I option reads this file and
creates a sequential dump for each table and the specified company numbers.
For example:
bdbpre –Iseqfile –C000-009 > BIGdump
Type cat seqfile to view the contents of seqfile:
timcs001
timcs002
timcs003
........
timcs099

Technical Manual
3-2
 Database tools

In the previous examples the database driver type is retrieved from

$BSE/lib/tabledef6.2. If you want to use a specific database driver for a
particular conversion you can use the –d option. If you use this option,
tabledef6.2 is not read.
The database driver can be of the following types:
n Sybase (sybase)
n Oracle (oracle)
n Informix Online (informix)
n DB2 (db2)
n Microsoft SQL Server (msql)
The name between the brackets is the entry that must exist in the
$BSE/lib/ipc_info file.
Enter the full database specification or host name, as in tabledef6.2 (optional
with specifications). For example:
–d oracle
–d "oracle(ORACLE_HOME=/usr/oracle)"
–d bv8c03
Possible options:
–u/U
Print usage information.
–v/V
Print information about the version of bdbpre.
–p <pack_comb>
define package combination.
–s
Suppress error messages, statistics, and so on.
–C
Company numbers for given table, in two formats:
− Specific company numbers (for example, 001 002 003)
− Range of company numbers (for example, 001-999)
–d
Database driver type, as follows:
O (Oracle) I (Informix) S (Sybase) D (DB2) M (Microsoft SQL-Server)
You can use the -d option to copy data from one instance of a database to
another, for example:
-d “oracle(ORACLE_HOME=/usr/oracle, ORACLE_SID=D1)”

Technical Manual
3-3
Database tools

–N
Database table name. Specify specific table names, for example, –N timcs000
timcs016. The –I and –N options are mutually exclusive.
–I
Input file with table names. The –I and –N options are mutually exclusive.
–E file
Redirects errors to file
–O file
Redirects output to file
–q output file
Redirect terminal output to file output file
–k
Drop table after dump is created.
–K
Drop table. Create a backup before dropping. This is only possible if the DBMS
supports this.
–t
Used to insert a separator (Default \0). Useful for separating the fields when
downloading records from the Bshell.
EXAMPLE You want to download your database to UNIFY which uses | as a separator.
Whenever you specify the –t parameter, a sequential file (.S) is created for each
table in the current directory unless the –o option is specified.
EXAMPLE bdbpre –t"|" –Ntimcs016 –C000 creates timcs016000.S in the current
directory. This sequential file is an ASCII dump of table timcs016000, in which
case the fields are separated by |. You can load this file directly into Oracle,
UNIFY, and so on, using SQL.
The options –x and –t are mutually exclusive.
–x
Creates an ASCII dump with a fixed-length record and without any separators.
This option is useful for downloading to databases such as dBASEIII, which
require fixed-length records without any separators.
In some cases a value is printed in the dump file that does not match the number
of digits before decimal points (DIGV). For example, suppose DIGV=3 and the
printed value=1234. The whole value, more than 3 digits, is written in the dump
file, and an error message appears. To prevent this, you can use the following
options in combination with the –x option:

Technical Manual
3-4
 Database tools

–y
Skip out-of-domain records for fixed-length dumps. See –x.
–z
Set value to 0 (zero) if value and DIGV do not match, that is, nullifies out-of-
domain records. Creates individual, fixed-length ASCII File with the extension
.F. See –x.
EXAMPLE bdbpre –x –Ntimcs016 –C000 creates timcs016000.F, which is an ASCII dump
with fixed-length records. This file can be loaded directly into dBaseIII.
–o
Can be specified with the options –x and –t to specify the directory name in
which the sequential file is created (.S or .F).
EXAMPLES Convert Informix database to a sequential dump. The database driver is explicitly
Informix.
bdbpre –dinformix –Ntimcs000 –C000-004 > timcs000_dump
The database driver is retrieved from $BSE/lib/tabledef6.2
bdbpre –Ntimcs000 –C000-005 > timcs000_dump
Convert Oracle to Informix.
bdbpre –s –doracle –Nttadv099 –C000 001 | bdbpost –dinformix
Transport the file to another database:
bdbpre –t"|" –d"oracle(ORACLE_SID=D1)" –Ntimcs016 –C000 –oor_dump_016
bdbpre –x –dinformix –Ntimcs016 –C000 –oci_dump_016
SEE ALSO
bdbpost
NOTES If you are piping your bdbpre output directly to bdbpost without giving the –s
option, messages on the screen are jumbled.
If the –d option is not used, the database driver is taken from
$BSE/lib/tabledef6.2. Also, if a remote driver is specified in tabledef6.2 the
database table from the remote machine is used, but the data dictionary is taken
from the current machine, which can cause great difficulties.

Technical Manual
3-5
Database tools

bdbpost
NAME
bdbpost
Create a database table from a sequential dump or append data to an existing
database table.
SYNOPSIS
bdbpost [–I input file][–O output file][–{qE} output file]
[–uUvVARfilxnmkK][–p pack_comb][–e file][–d driver type][–D seq_dir]
[–t sep] [–c compnr][–C compnr range][pattern]
DESCRIPTION
The bdbpost tool reads from either the argument you supply or from the
standard input and creates a new database table if that table does not exist. If the
append option is turned on, it appends data to an existing table. The bdbpost
tool also compares current data dictionary information with the information in
the dump. If they do not match, the bdbpost tool gives an error. If the current
data dictionary is not present, it creates a data dictionary based on the dump. For
each table, bdbpost prints information such as the table name, indexes, the
number of records, and any errors.
The bdbpost tool can be run using a sequential dump created by bdbpre or by
using sequential dumps from other databases. The different methods are shown
in the following examples:
This example shows how to use a sequential dump created by bdbpre.
On system 1:
bdbpre –doracle –Ntimcs016 –C000-003 > timcs_dump
On system 2:
bdbpost < timcs_dump
This example shows the use of sequential dumps (–x or –t option of bdbpre)
from other databases. In this case the –D option is mandatory to get the directory
name in which .S files are stored.
bdbpost –dinformix –t"|" –D./seqdir
bdbpost –doracle –x –D./seqdir
Options –x, –t, and –D are required when you upload ASCII files from another
database to the Bshell format.

Technical Manual
3-6
 Database tools

–D
Path for sequential files, which have the extension .S or .F.
–t
Separator. If you want to load a sequential dump containing separators
(for example, dump from UNIFY has separator |), make sure the file name is the
table name with the extension .S.
EXAMPLE If you want to load a UNIFY sequential dump into the table ttimcs016000, first
move the sequential dump to the ttimcs directory with name ttimcs016000.S.
bdbpost –Dttimcs –t"|" searches for a .S file in ttimcs and if that file is found,
the corresponding tables are created or appended.
CAUTION With the –D option all .S files in that directory are used to create/append the
tables, so be sure to remove unwanted .S files before running bdbpost.
–x
If you want to load a sequential dump with a fixed-length record without any
separators (.F files, for example dump from dBase-III), make sure the file name
is the table name with the extension .F.
EXAMPLE Suppose you want to load a dBASEIII sequential dump into the timcs016000
table (Oracle organization). bdbpost –doracle –Ddbase –x searches for an .F
file in dbase and if that file is found the corresponding table is created or
appended.
In some of these examples the database driver type is retrieved from
$BSE/lib/tabledef6.2. If you want to use a specific database driver for a
particular conversion, you can use the –d option.
The database driver can be of the following types:
n Sybase (sybase)
n Oracle (oracle)
n Informix Online (informix)
n DB2 (db2)
n Microsoft-SQL Server (msql)
Enter the driver name of the database driver between the parentheses. See also
the section “bdbpre.”
The following parameters can be used for bdbpost:
–u/U
Print usage information.
–v/V
Print information about the version of bdbpost.

Technical Manual
3-7
Database tools

–p
Define package combination
<pattern>
Pattern to specify tables that are filtered out of the dump. Wildcards such as *
and ? are allowed.
–c
Company number for the tables to be created.
–C
Range of customer numbers on which to perform the bdbpost operation. This
must be the last option specified in the command, other than the <pattern>
option.
–d
Database driver type, as follows:
O (Oracle) I (Informix) S (Sybase) D (DB2) M (Microsoft SQL-Server)
You can use the -d option to copy data from one instance of a database to
another, for example:
-d “oracle(ORACLE_HOME=/usr/oracle, ORACLE_SID=D1)”
–x
Load ASCII file (.F) with fixed-length records.
–D
The directory name for ASCII files to be loaded.
–e
File to store the names of unsuccessfully created tables.
–k
The existing tables are deleted.
–K
The existing tables are deleted after a backup is made (if DBMS supports this).
–l
Display contents of input.
–I <file>
Redirects input from input file <file>.
–O <file>
Redirects output to output file <file>.
–E <file>
Redirects errors to error file <file>.

Technical Manual
3-8
 Database tools

–q <output file>
Redirect terminal output to output file.
–m
Disable domain constraints.
–i
Ignore domain range error and skip record.
–n
Ignore referential integrity constraints.
–A
Append to an existing table. If duplicate records exist, do not overwrite them
from the dump. Print duplicate record summary at the end. Create table if it does
not exist.
–R
Append to an existing table or create a new one. If a record already exists, the
record in the dump replaces it. A summary is given at the end. Note that only the
existence of the primary key is checked. If a primary key exists, the record is
replaced. If the primary key does not exists but a secondary key exists, error 100
(duplicate record) occurs.
–t
Specify the used separator. The –t option is used when loading an ASCII file
from another database.
–f
Fast mode. Tables are created by first inserting all rows and then creating the
indexes (if DBMS supports this). When using the –f option be aware of the
following:
n Interrupting bdbpost results in table inconsistency
n An index cannot be created in case of a duplicate conflict
n For large tables, the adding of indexes can take a long time (> 15 minutes)
EXAMPLES bdbpre –doracle –Nttadv000 –C000-010 > dump
bdbpost –doracle < dump
Creates (appends) all tables in dump
bdbpost –l < dump
Gives you names of tables in dump
bdbpost –C000-005 < dump
Creates/appends tables only in the given customer range
bdbpre –doracle –Nticom000 –C000-010 > dump

Technical Manual
3-9
Database tools

bdbpost –C000-005 ticom* < dump

Creates/appends tables only in the given customer range and where the table
name matches the pattern ticom.
bdbpost –R –C000-005 ttadv* < dump
Creates/appends tables only in the given customer range and where table name
matches the pattern ttadv. If duplicate records exist, they are replaced by records
from the dump.
bdbpre –dsybase –Nttadv099 –C000-999 | bdbpost –doracle
Converting from one database organization to another (From Sybase to Oracle
organization).
bdbpost –doracle –Dunify –t"|"
Loading from other databases (UNIFY to Oracle)
bdbpost –Doracle –x –dsybase
(Oracle to Sybase)
CAUTION If a remote driver is specified in tabledef6.2, the database table from the remote
machine is used but the data dictionary is taken from the current machine, which
can cause great difficulties.
When the –m or –n option is used, the data in the database can violate the Baan
integrity constraints. Data can violate the Baan domains or it can violate Baan
referential integrity.
SEE ALSO
bdbpre

bdbreconfig
NAME
bdbreconfig
Reconfigure database table according to new data dictionary.
SYNOPSIS
bdbreconfig [–uUvVpscinmfTZ][–t dir][–o file][–e file][–d driver type]
[–N table [table [..]] [–I file][–O file][–E file][–S number][–C company
numbers]

Technical Manual
3-10
 Database tools

DESCRIPTION
The bdbreconfig tool reads a table name and company numbers as arguments
and converts them into a new table definition that matches a new data dictionary
definition. The bdbreconfig tool requires the following:
n A new data dictionary with a .new extension
n The current data dictionary (for example dtimcs016)
n Tables matching those in the current data dictionary
After successful completion, old tables are deleted and new tables are generated.
The bdbreconfig tool finds the optimum way to reconfigure. If there is no real
change in the old and new data dictionaries, it prints the message “No conversion
required.” It also sees to it that a database remains consistent.
You can perform some reconfigurations by modifying the old data dictionary.
Otherwise the reconfiguration steps are:
1 Compare the current and new data dictionaries.
2 If reconfiguration is required, create a sequential dump
(R.<table name>, for example R.timcs016244) for the specified tables.
3 Delete old tables.
4 From the sequential dump, build new tables that match the new data
dictionary definition.
The following options are available with bdbreconfig:
–U/u
Usage information.
–V/v
Version information.
–p
The name of the used package combination.
–p <pack. comb.>
Define the package combination.
–s
This suppresses error messages and other information.
–C
Company numbers for a given table in these formats:
n Specific company numbers (for example 001 002)
n Range of company numbers (for example 001-010)

Technical Manual
3-11
Database tools

–d
Database driver type, as follows:
O (Oracle) I (Informix) S (Sybase) D (DB2) M (Microsoft SQL-Server)
You can use the -d option to copy data from one instance of a database to
another, for example:
-d “oracle(ORACLE_HOME=/usr/oracle, ORACLE_SID=D1)”
–T
If the domain changes, the table is not reconfigured.
–B
Backup table instead of dropping.
–Z
Reorganize, that is, clean up, the table.
–I
The file with table names.
–i
Ignore errors during reconfiguration, for example, duplicates.
–n
Ignore referential integrity constraints.
–m
Disable domain constraints.
–f
Fast rebuild of table (first rows, then indexes).
–N
List of tables to be reconfigured.
–c
This compares two data dictionaries and checks whether reconfiguration is
required.
–t
The directory name for temporary dump files.
–I
File with table names to reconfigure.
–o
File with successfully reconfigured tables.
–e
File with unsuccessfully reconfigured tables.

Technical Manual
3-12
 Database tools

–O file
Redirects output to output file file.
–E file
Redirects errors to error file file.
–S number
number is the maximum size per sequence. number can be specified as any
number or a number followed by K for kilobytes, M for megabytes, and G for
gigabytes. The default size is 2G per dump sequence.
The bdbreconfig tool reconfigures only one table of given company numbers at
a time.
EXAMPLE bdbreconfig –Ntimcs016 –C001
Reconfigures timcs016 for company no. 001.
bdbreconfig –Ntimcs016 –C000-010
Reconfigures timcs016 for a range of company numbers.
If bdbreconfig is used without the –c option, the exit status is either 0 or 1,
depending on successful or unsuccessful completion. If the exit status is 1
(unsuccessful) the file specified at the –o option contains tables that can be
reconfigured, and the file at the –e option contains the tables that cannot be
reconfigured.
If bdbreconfig is used with the –c option, the file specified at the -o option
contains the table names which are found correct, and the file specified at the
-e option contains the table names that must be reconfigured.
CAUTION Make a copy of the dump before executing the following.
In case of interrupted reconfiguration while the R.table file still exists, the
following command uses the dump R.ttadv100222 to rebuild the table.
bdbreconfig –N ttadv100222+
EXAMPLE bdbreconfig –Ntimcs016 –C000-003
Reconfigures timcs016 for some company numbers according to the new data
dictionary, provided that dtimcs/dtimcs016 and dtimcs/dtimcs016.new are
present.
bdbreconfig –dsybase –Ntimcs016 –C000
Reconfigures Sybase table timcs016000.
When the –m or –n option is used, the data in the database can violate the Baan
integrity constraints. Data can violate the Baan domains or it can violate Baan
referential integrity.

Technical Manual
3-13
Database tools

refint
NAME
refint
Check or repair the referential integrity within the database.
SYNOPSIS
refint [–uU][-vV][-b|-B file][-a |-A file][-l |-L file][-c][-n][-r][-s]
[-p package][-I infile][-O outfile][-E errfile][-N table [table [..]]]
[-C compnr [compnr [..]]]
DESCRIPTION
The refint tool checks the integrity of references in the database. You can also
use it to repair the integrity of a corrupted database. Integrity refers to the
accuracy or validity of data.
The environment variable KEY_BUFFER specifies the number of keys stored in
a buffer before updating the database. Default KEY_BUFFER=10000.
You can use the following options:
–u/U
Print usage information.
–v/V
Print version information.
–p
Define package combination.
–C
Specify company numbers.
–b/B
List references before repair (-b output file is ref_before).
–a/A
List references after repair (-a output file is ref_after).
–c
Check validity of references.
–n
Nullify undefined references, all references to the specified tables are checked.

Technical Manual
3-14
 Database tools

–r
Repair reference counter, from all specified tables the reference counters are
checked and changed if necessary. Undefined references to the specified tables
are displayed on the screen.
–l/L
File for undefined references. (-l file is ref_undefined).
–I
File containing table names to be checked.
–N <table>
Specify table (format: ttadv100 or ttadv100999).
–s
Handle one single-parent table only.
EXAMPLES refint –Iref_tables –c –n –C100
refint –Ntimcs016 –l

Technical Manual
3-15
Database tools

Technical Manual
3-16
4 Database management

General
Database servers and user interface servers perform processes different from
those performed by the client, the Bshell. They communicate with the Bshell
using a communication protocol. This section discusses the working and usage of
several protocols that can be used for local and remote communication.
Local communication means that server and client are started on the same host
machine. Remote communication means that server and client are on different
machines.

Accessing tables
Tables can be distributed as follows:
n Local: the tables and the Bshell are on one system
n Remote: the Bshell is on one system; all the tables are on another system
n Combination: some tables and the Bshell are on one system; some tables are
on another system
To access a table, the Bshell and database drivers must know the database type
and the location of the table. This information has been stored in the following
files:
n tabledef6.2
n fd6.2.package combination

tabledef6.2
Table references that refer to a database type are stored in the file tabledef6.2.
This file is located in the directory $BSE/lib. To fill this file, use the Tables by
Database (ttaad4111m000) and Database Definitions (ttaad4510m000) sessions
on the BaanERP ToolsÈDatabase ManagementÈDatabase Definitions and
Directories. If the tables are on a different system, you specify the host name of
the remote system instead of entering the database type. The Bshell knows which
database server must be started and in which database the required tables can be
found.

Technical Manual
4-1
Database management

fd6.2.package combination
The file fd6.2.package combination is used to refer the following dictionary
types to a specific directory (path) for a specific package combination:
n Forms
n Menus
n Objects
n Program scripts
n Functions
n Report scripts
The file fd6.2.package combination is located in the directory $BSE/lib. To fill
this file, use the Directories of Software Components (ttadv1115m000) session in
the BaanERP ToolsÈ Application CustomizationÈ Packages and Modules
menu. The specific directory path of the table definitions must be entered in the
Package Combinations (ttaad1520m000) session and is also placed in the
fd6.2.package combination file. You can access this session through the
BaanERP ToolsÈ Application DevelopmentÈ Packages and Modules menu.

Example table access (table and Bshell on the same

system)
This example roughly explains how the Bshell accesses a table if they are on the
same system.
Suppose you work with company number 001, and read in an application table
ttmod003. The contents of tabledef6.2 are:
*:001:oracle7(ORACLE_HOME=/a1/oracle/product/7.3.3,ORACLE_SID=baan5):N
First, Bshell reads tabledef6.2, searches for the reference and in this case starts
the Oracle server. A connection between Bshell and server is made. Then the
server accesses the table and the database action is carried out.

Database drivers
When an action is carried out on a table, a database driver (server) is started in
the background for the database type in question.

Technical Manual
4-2
 Database management

Accessing tables on other systems

When the database is on a different machine, the communication takes place as
described in the following example.
Suppose that in a Bshell program running on the system host_local, you want to
read a record from the table ttmod000111 (table ttmod000 in company 111). This
table is stored on the remote system, whose host name is host1. First, the Bshell
searches tabledef6.2 on the system host_local for the table in question. The file
contains the following reference:
ttmod:111:host1
This means that all tables of ttmod with company number 111 are stored on the
remote system host1. Next, the Bshell accesses the system host1, where it reads
tabledef6.2. This file contains the reference:
ttmod:111:oracle
This implies that the desired tables are stored in the Oracle database. The oracle
server is then started. The desired table is searched and accessed using
tabledef6.2 on the system host1. The record is read and its data is sent to the
Bshell program on the system host_local.

Audit trail
Audit trail is a method by which actions on a record are logged in several audit
files using an audit server. To work with audit trails, you must place the server
name in the file $BSE/lib/ipc_info. In addition, you must specify in tabledef6.2
that you are using audit trail. To do this, set the Audit Trail field to Y or make
the audit server the mirror server. See the section, “Database mirroring.” When
the client performs an action on a table, the action is logged in an audit file
through the audit server. See also the chapter “Audit management.”

Database mirroring
Tables can be stored in one or more database systems. The file tabledef6.2
contains references to the database where the tables are saved and/or read. For
each reference you can specify one or more database systems where the tables
are positioned. An example is:
tccom:*:oracle&host2.
This reference indicates that all tables of the tccom are in the Oracle database and
on the system host2.

Technical Manual
4-3
Database management

The tables from the previous example are copies. A change applied to a table
from tccom is carried through in the Oracle database on the local system and on
the remote system host2. Also, a change to the tccom tables on the remote system
must be carried through in the database on the local system. This means that
tabledef6.2 on the remote system must contain a reference to the local system.
If the system refers to a remote system, tabledef6.2 on that system is read. This
file contains a reference to both the tables from tccom and to the systems where
copies of the tccom tables are stored.
In case of a read action, if there is a local database for the tables, it is used
without a lock to enhance performance. If no local database has been defined, a
remote database is used.
The principle of database mirroring is outlined in the following figure.

Figure 3, Database mirroring

In this example, a change applied to a table in tccom is carried through in both

the Sybase database on the local system and in the Oracle database on the remote
system. A table is read, without a lock, from the Sybase database.

Technical Manual
4-4
 Database management

Alternative databases
The file tabledef6.2 contains references to the database where the tables are
saved and/or read. For each reference you can specify an alternative database
driver or host where the table is located.
Assume tabledef6.2 contains the following line:
*:*:host1|oracle
The previous example indicates that a table is searched or stored on host1. If that
operation fails, the table is searched or stored in a local Oracle database. If this
also fails, the table cannot be accessed.
If you are using mirroring databases, you can specify an alternative database
driver or host for each database system.
*:*:oracle&host1|host2
The previous example means that a table must be accessible in a local Oracle
database. Besides that, it is searched or stored on host1. If host1 is not available,
the table is searched or stored on host2.

Communication protocols
Communication between the client and the server includes both local and remote
communication. Local communication means that both the client and the server
are on the same system, which does not apply to remote communication.
Depending on the hardware configuration, you can choose the following
protocols for the communication between the client and the server:
n Sockets (local or remote)
n Pipes
n Message queues
Only the remote socket protocol can be used for remote communication.
For the communication between client and servers, the following files and/or
directories are needed:
n ipc_info (in the $BSE/lib directory)
n A remote user file (ruser name). This file is located in the $BSE/lib/user
directory, and is described in detail in the chapter “Remote databases.”

Technical Manual
4-5
Database management

The ipc_info file contains the following data:

n Server name
n Operation mode
n Semaphore key
n Message key
n Protocol
n Path of the server program
The server name is the name of the server, which can be used in other files, for
example in tabledef6.2 for database servers. Only the single mode can be used as
operation mode, because the multimode is not yet implemented. In single mode,
one server can communicate with one client. The client can communicate with
more than one server, however, as shown in the following figure.

Figure 4, Single operation mode

The semaphore key and message key have not been implemented either.
You can choose from the following protocols:
n Socket (s)
n Message queue (m)
n Pipes (p)
These protocols are described in more detail in the following sections.
The server’s path can include environment variables, for example $BSE, which
must be used as ${BSE}. An example of the ipc_info file is shown in Example
ipc_info File.

Technical Manual
4-6
 Database management

Socket protocol
A socket is a UNIX function that enables processes to communicate in both
directions. The socket protocol is subdivided into local and remote sockets. Local
sockets are used for local communication; remote sockets for remote
communication. The client uses a local socket address if it is free at that moment.
Both sockets are described in the following paragraphs.

Local sockets
When a client (Bshell) opens a table, the client reads tabledef6.2. If this file
refers to a certain server, the client reads data about that server from the file
ipc_info. The file ipc_info contains data such as the protocol to be used and the
position of the executable server. Next, the client uses a socket to start the server
and establish communication with the server.

Remote sockets (only for database servers)

Suppose that the client on host A wants to open a table on host B as in the
following figure.

Figure 5, Remote socket protocol

Suppose that the client on host A wants to open a table on host B.

This is done through sockets. The client reads tabledef6.2 on host A, which
indicates that the table is on host B. The client logs on to host B using the
password found in the file r<user’s name> in the $BSE/lib/user directory.

Technical Manual
4-7
Database management

The client uses a remote execute call (rexec()) to start ipc_boot6.2 on host B,
which starts fs6.2. fs6.2 is used to open, read, and write data from a remote $BSE
directory.
The ipc_boot process reads ipc_info to find the directory in which fs6.2 is
located. Then an exec() call is done, so that the ipc_boot6.2 process changes to
fs6.2.
The client searches tabledef6.2 on host B, to determine which driver must be
used to open the desired table, for example, Oracle.
Another remote execute call is done; this time to start the Oracle driver on host
B. Again ipc_boot6.2 is started and reads ipc_info for the path of the Oracle
driver. Now icp_boot6.2 does an exec() call to the Oracle driver.
Communication then proceeds normally between client and Oracle driver.

Pipes protocol
The pipes protocol is analogous to the local socket protocol. In pipes protocol,
the client and server communicate through unnamed pipes streams. A pipe is a
UNIX function that enables processes to communicate in one direction. The
communication between the client and the server takes place through two pipes.

Message queue protocol

The message queue protocol is only used for local communication. With this
protocol the client starts the server. The client and server communicate through a
message queue using a semaphore key. This key indicates whether a message
queue contains information, so that the client or server knows that information
has been sent. See the following figure.

Figure 6, Message queue protocol

Technical Manual
4-8
 Database management

For example, suppose the client wants to send information to the server. This
information or message is put in a message queue, and the semaphore key is
given the value of 1. The semaphore key shows the server that information has
been sent, so that the server can read it. After the server reads it, the semaphore
flag is reset and released, so that the client knows its information has been read.
If the client process is canceled, the client will not put data in the message queue.
However, the semaphore key is decremented using an undo value. The server
reads the message queue and finds no message. Thus, the server knows that the
client process has been terminated and stops its own process.

Example ipc_info file

Server name Protocol (s/m/p) Path of executable server

oracle7 s00s ${BSE}/bin/ora7_srv6.2
informix s00s /usr1/bse/bin/inf_srv6.2
Bshell s00s ${BSE}/bin/bshell6.2
fs6.2 s00p ${BSE}/bin/fs6.2

Technical Manual
4-9
Database management

Technical Manual
4-10
5 Native language support

General
Native Language Support controls the communication between the character set
of a terminal and that of Bshell6.2.
Bshell6.2 has the 8-bit character set of ISO 8859-1. Most terminals and printers
do not support this character set, so Native Language Support techniques are
implemented.
The character set of the terminal is uniform. However, various country-
dependent keyboards can be used with the same terminal. Not all the characters
of the terminal character set can be found on the keyboard. The remaining
characters must be composed. How to compose characters is described in the
section “Composed characters.”
An input conversion table is used to convert the character set of a terminal to that
of the Bshell. In this table a character from the terminal character set is converted
to the corresponding value from the ISO 8859-1. This means that characters from
the Bshell must be converted again to correctly control the terminal or printer.
Both the printer and the terminal need an output conversion table. For a
description of the conversion tables, see the section “Conversion tables.”
To maintain the conversion tables you can use the NLS editor. This editor is
described in the section “NLS editor.”

Composed characters
Characters that are not found on a keyboard must be composed by a unique
combination of two or more existing characters. An example of a list of
composed characters is contained in the section “List of compose sequences.”
Prior to entering the combination, you must press a compose key. The compose
key is defined in the terminal information file of the terminal. For more
information see the section “NLS-related files.”
The compose key can be linked to a special key on the keyboard, for example, a
function key. You can also compose characters by using the Compose character
key, if it is present on the keyboard. This key has the same function as the
compose key.

Technical Manual
5-1
Native language support

The manufacturer of the terminal has defined the composed characters. The user
documentation of the terminal contains a list of these composed characters.

Conversion tables
You need two kinds of conversion to make the character sets of Bshell and the
terminal or printer compatible: input conversion and output conversion.

Input conversion table

The input conversion table converts the entered characters to their corresponding
values in the ISO 8859-1 character set. Composed characters and characters that
are not compatible with the ISO character set must be included in this table.
To record and/or modify an input conversion table, use the NLS editor, described
in the section “NLS editor.” An example of an input conversion table is given in
the section “Input/output table vt200.”

Output conversion table

The output conversion table converts the characters of the ISO character set to
their corresponding values in the terminal or printer character set. You, therefore,
need two output conversion tables:
n Terminal output table
n Printer output table
In this way a character is displayed or printed as entered.
To record and/or modify an output conversion table you can use the NLS editor,
described in the section “NLS editor.”
An example of an output conversion table is given in section “Input/output table
vt200” and the section “Printer output table MT910 HP Prestige character set.”

Technical Manual
5-2
 Native language support

Example of total conversion

Terminal type: vt200 with North American keyboard
Printer type: Mannesmann mt910
Required character: ÿ = ISO8859-1 code 255 = DEC vt200 code 253
The character ÿ is not found on a North American keyboard so it must be
composed. The compose sequence can be y".
To convert the composed character (DEC code 253) to its ISO 8859-1
representation (code 255), the input conversion table of the terminal must contain
the following line:
\255 y"
The result of the input conversion is that the keyboard input Compose[y]["] is
stored as ISO character y on disk or Bshell memory.
To display the entered (required) character on your screen, ISO code255 must be
converted to DEC code 253. Therefore the terminal output conversion table must
contain the following line:
\255 \253
To print the entered (required) character on a Mannesmann mt910 printer the
printer output table must contain the following line:
\255 y\b\168
This means that character ÿ is printed as:
[y] <Backspace>["]
y \b \168

Technical Manual
5-3
Native language support

This example is shown in the following figure.

Figure 7, Example NLS conversion

NLS editor
Use the NLS editor to maintain input and output conversion tables. To start the
editor, type nlsedit6.2. To maintain printer output tables, use the –p option
followed by the printer name. To maintain a terminal conversion table, the shell
variable TERM must have the value corresponding to your terminal type.
For an X-terminal to display the characters of ISO 8859-1, you must have the
right font. The section “NLS-related files” contains an example of suitable fonts.
After you have started the editor, the upper part of the ISO character set (160-
255) is displayed. If a character cannot be displayed on your screen, it might be
because it is not found in the character set of the terminal.
You can display feasible options by typing ?. Exit the editor by typing Q.

Technical Manual
5-4
 Native language support

Initial screen options

Feasible initial screen options are:
-d/h/o set decimal/hex/octal display mode
-t edit nls_in and nls_out tables
-r toggle display of characters
-s toggle character set
With the options -D, -H, and -O you can change the display mode of the
character values into decimal, hexadecimal, or octal.
The -T option is used to edit the conversion tables. A second screen appears.
The initial screen shows the characters displayed by the terminal as defaults.
With the -R option you can set this mode on/off.
The default characters displayed on the initial screen are characters of the upper
part (160-255) of the ISO character set. With the -S option you can switch
between the lower part (32-127) and the upper part of this character set.
You can display the initial screen options by typing ?. The window that appears
shows also your terminal setup, NLS table names, and shell variable BSE.

Maintenance of NLS conversion tables

By entering the -T option on the initial screen you can maintain the input and
output conversion tables. Five columns are displayed as illustrated in the
following example.

input int s # output

A‘ 192 À 0 \192
A' 193 Á 1 \193
A\^ 194 Â 0 \194

The left column, input, contains the compose sequence of the input characters.
This column also contains single input characters not transparent with the ISO
character set.
The second column, int, displays the values of the characters in the ISO character
set.
The third column, s, shows the characters as they are represented after output
conversion.

Technical Manual
5-5
Native language support

The fourth column, #, contains the character set number. Character sets are
defined at the terminal or printer. The possible values are 0-9, and the default is
0.
The last column, output, contains the output sequence of the characters. You can
type the output sequence in decimal, hexadecimal or octal code.
The bottom of the screen shows the display mode in which the values of the
characters are displayed. The last line indicates the marked character. The
numeric representations and description are also included.
Feasible options to maintain conversion tables are:
–D/H/O set decimal/hexadecimal/octal display mode
-N toggle numeric/alpha mode compose sequence
–ARROW DOWN next line
–ARROW UP previous line
–CTRL+N next page
–CTRL+P previous page
–K single key
–C enter compose sequence
–S enter output sequence
–W write nls_in and nls_out file
–+/- increment/decrement character set number
With the options -D, -H, and -O you can change the display mode of the
character values into decimal, hexadecimal, or octal.
The -N option converts alphanumeric characters from the input and output table
to their numeric values, and vice versa.
To move the bar to the following or previous line, use the key ARROW DOWN or
ARROW UP.
Pressing CTRL+N or CTRL+P moves the bar 15 lines forward or backward.
To enter the input sequence of a single key, press -K. The editor asks you to
press the required key. That character is included directly in the table. You do not
need to press ENTER after you have entered the character.
After pressing -C you can enter the compose sequence. You can enter up to 35
characters.

Technical Manual
5-6
 Native language support

To type the output sequence, enter -S. You can enter up to 35 characters.
After any modification, you can store the input and output tables by entering -W.

NLS-related files
To work with NLS you must have the following files in your BSE environment:
n Terminal information file$BSE/lib/terminf/...
n Printer information file$BSE/lib/printinf/...
n Input conversion table$BSE/lib/nlsinf/...
n Output conversion table$BSE/lib/nlsinf/...
You must also have the right terminal setup, shell variable TERM, and font. For
some additional features you can use sort tables and shift tables.

Terminal information file

The terminal information files have been placed in subdirectories of the directory
$BSE/lib/terminf. For example, terminal information files starting with the letter
v are placed in subdirectory v.
Two lines must be added to each terminal information file, to indicate the
conversion tables used by the terminal:
nls_in=<terminal type>.in
nls_out=<terminal type>.out,
For example, for a vt200 terminal, the terminal information file vt200 can
contain the following lines:
nls_in=vt200.in
nls_out=vt200.out,
The compose key code is also defined in the terminal information file:
kcompose=<compose key code>,
The code can be anything that does not conflict with other special keys. For
example, suppose you want to define the compose key as ESC+C. Add the
following line to the terminal information file:
kcompose=\ec,

Technical Manual
5-7
Native language support

Printer information file

The printer information files have been placed in subdirectories of the directory
$BSE/lib/printinf. For example, printer information files starting with the letter
m are placed in subdirectory m.
The following line must be added to each printer information file, to indicate the
output conversion table used by the printer:
nls_out=<printer type>.out,
For example, for an mt910 printer, the printer information file mt910 must
contain the following line:
nls_out=mt910.out,

Input and output conversion table

The input and output conversion tables are stored in directory $BSE/lib/nlsinf.
Possible table names in this directory are:
<terminal type>.in (input conversion terminal)
<terminal type>.out (output conversion terminal)
<printer type>.out (output conversion printer)
For each terminal there is one input and one output conversion table. For each
printer there is only an output conversion table.
The format of every line in a conversion table is:
<string><tabs or spaces><string>
A string can consist of one or more:
n ASCII codes (A-Z, a-z, and 0-9...)
n Octal codes (\01, \012, \012...)
n Decimal codes (\1, \12, \12...)
n Hexadecimal codes (\0x1, 0x12…)
n Control codes (^A, ^B, ^C…)
n Escape codes (\E, \e, \s, \n, \r, \f, \v, \b)
n Special codes (\\, \^ )
A list of control and escape codes is given in the Overview of Esc/Ctrl codes.”
The first column contains the numeric representation of the ISO 8859-1 character
set.
In an input table, the second column contains the compose sequence of the input
characters. In an output table, the second column contains the numeric
representation of the output characters. The columns are separated by spaces or
tabs.

Technical Manual
5-8
 Native language support

Examples of conversion tables are given in the section “Input/output table

vt200.”

Terminal setup
The communication mode of the terminal must be 8-bit no parity. The shell
variable TERM must have the name of the terminal information file of the
terminal.

Character sets and fonts

Line drawing characters do not need to be included in the character set. These
characters are drawn by the Bshell itself. To check the suitability of character
sets, type the following command:
xfontsel –pattern –*-*-iso8859-1
To choose the font and set the shell variable FONT, type the following
command:
csh: setenv FONT "`xfontsel –pattern *-c-*-iso8859-1-print`"
sh: FONT="`xfontsel –pattern *-c-*-iso8859-1-print`" export FONT
After reducing the number of options, a character set can be chosen in this
program.
Suitable character sets are:
-bitstream-terminal-medium-r-normal--18-140-100-100-c-110-iso8859-1
-misc-fixed-medium-r-normal--14-130-75-75-c-70-iso8859-1
-misc-fixed-medium-r-normal--15-120-100-100-c-90-iso8859-1

Sort tables
You can use a sort that differs from the default sort method. To do this you can
use a sort table to specify the weight of characters.
The file with the sort table has the default name:
$BSE/lib/nlsinf/sort.tab
You can make your own sort table by filling the environment variable
SORT_TABLE with the name of your own sorting file. A sort table has two
columns. The first one contains the character, the second contains the weight of
that character.

Technical Manual
5-9
Native language support

For example, to sort the B before the A, create a sort table with the following
contents:
\65 \66 or A B
\66 \65 B A
You can use escape and control codes, like \012, \e, ^A. See the section “Input
and output conversion table” for more information.

Shift tables
Shift tables can be used to specify lowercase and uppercase characters that
belong together.
The file with the shift table has the default name:
$BSE/lib/nlsinf/shift.tab
You can make your own shift table by filling the environment variable
SH_TABLE with the name of the file with your own shift table. A shift table has
two columns. The first one contains the uppercase character, the second contains
the lowercase character.
For example, to specify the lowercase and uppercase character c with cedilla,
create a shift table with the following contents:
\199 \231

Technical Manual
5-10
 Native language support

ISO 8859-1 character set (0-127)

0 16 32 48 64 80 96 112
SP 0 @ P ‘ p
1 17 33 49 65 81 97 113
! 1 A Q a q
2 18 34 50 66 82 98 114
“ 2 B R b r
3 19 35 51 67 83 99 115
# 3 C S c s
4 20 36 52 68 84 100 116
$ 4 D T d t
5 21 37 53 69 85 101 117
% 5 E U e u
6 22 38 54 70 86 102 118
& 6 F V f v
7 23 39 55 71 87 103 119
‘ 7 G W g w
8 24 40 56 72 88 104 120
( 8 H X h x
9 25 41 57 73 89 105 121
) 9 I Y I y
10 26 42 58 74 90 106 122
* : J Z j z
11 27 43 59 75 91 107 123
+ ; K [ k {
12 28 44 60 76 92 108 124
‘ < L \ l |
13 29 45 61 77 93 109 125
- = M ] m }
14 30 46 62 78 94 110 126
. > N ^ n -
15 31 47 63 79 95 111 127
/ ? O _ o

Technical Manual
5-11
Native language support

128 144 160 176 192 208 224 240

NBSP ° À Ð à ð
129 145 161 177 193 209 225 241
¡ ± Á Ñ á ñ
130 146 162 178 194 210 226 242
¢ ² Â Ò â ò
131 147 163 179 195 211 227 243
£ ³ Ã Ó ã ó
132 148 164 180 196 212 228 244
¤ ' Ä Ô ä ô
133 149 165 181 197 213 229 245
¥ µ Å Õ å õ
134 150 166 182 198 214 230 246
¦ ¶ Æ Ö æ ö
135 151 167 183 199 215 231 247
§ · Ç × ç ÷
136 152 168 184 200 216 232 248
¨ , È Ø è ø
137 153 169 185 201 217 233 249
© ¹ É Ù é ù
138 154 170 186 202 218 234 250
ª ° Ê Ú ê ú
139 155 171 187 203 219 235 251
« » Ë Û ë û
140 156 172 188 204 220 236 252
¬ ¼ Ì Ü ì ü
141 157 173 189 205 221 237 253
SHY ½ Í Ý í ý
142 158 174 190 206 222 238 254
® ¾ Î Þ î þ
143 159 175 191 207 223 239 255
| ¿ Ï ß ï ÿ

Technical Manual
5-12
 Native language support

DEC vt100/vt200 character set (0-127)

0 16 32 48 64 80 96 112
NUL DLE SP 0 @ P ‘ p
0 10 20 30 40 50 60 70
1 17 33 49 65 81 97 113
SOH DC1 ! 1 A Q a q
1 11 21 31 41 51 61 71
2 18 34 50 66 82 98 114
STX DC2 “ 2 B R b r
2 12 22 32 42 42 62 72
3 19 35 51 67 83 99 115
ETX DC3 # 3 C S c s
3 13 23 33 43 53 63 73
4 20 36 52 68 84 100 116
EOT DC4 $ 4 D T d t
4 14 24 34 44 54 64 74
5 21 37 53 69 85 101 117
ENQ NAK % 5 E U e u
5 15 25 35 45 55 65 75
6 22 38 54 70 86 102 118
ACK SYN & 6 F V f v
6 16 26 36 46 56 66 76
7 23 39 55 71 87 103 119
BEL ETB ‘ 7 G W g w
7 17 27 37 47 57 67 77
8 24 40 56 72 88 104 120
BS CAN ( 8 H X h x
8 18 28 38 48 58 68 78
9 25 41 57 73 89 105 121
HT EM ) 9 I Y I y
9 19 29 39 49 59 69 79
10 26 42 58 74 90 106 122
LF SUB * : J Z j z
A 1A 2A 3A 4A 5A 6A 7A
11 27 43 59 75 91 107 123
VT ESC + ; K [ k {
B 1B 2B 3B 4B 5B 6B 7B

Technical Manual
5-13
Native language support

12 28 44 60 76 92 108 124
FF FS ‘ < L \ l |
C 1C 2C 3C 4C 5C 6C 7C
13 29 45 61 77 93 109 125
CR GS – = M ] m }
D 1D 2D 3D 4D 5D 6D 7D
14 30 46 62 78 94 110 126
SO RS . > N ^ n –
E 1E 2E 3E 4E 5E 6E 7E
15 31 47 63 79 95 111 127
SI US / ? O _ o DEL
F 1F 2F 3F 4F 5F 6F 7F

Technical Manual
5-14
 Native language support

DEC vt100/vt200 character set (128-255)

128 144 160 176 192 208 224 240
DCS ° À à
80 90 A0 B0 C0 D0 E0 F0
129 145 161 177 193 209 225 241
PU1 ¡ ± Á Ñ á ñ
81 91 A1 B1 C1 D1 E1 F1
130 146 162 178 194 210 226 242
PU2 ¢ ² Â Ò â ò
82 92 A2 B2 C2 D2 E2 F2
131 147 163 179 195 211 227 243
STS £ ³ Ã Ó ã ó
83 93 A3 B3 C3 D3 E3 F3
132 148 164 180 196 212 228 244
CCH Ä Ô ä ô
84 94 A4 B4 C4 D4 E4 F4
133 149 165 181 197 213 229 245
MW ¥ µ Å Õ å õ
85 95 A5 B5 C5 D5 E5 F5
134 150 166 182 198 214 230 246
SPA ¶ Æ Ö ö
86 96 A6 B6 C6 D6 E6 F6
135 151 167 183 199 215 231 247
EPA § · Ç ç
97 A7 B7 C7 D7 E7 F7
136 152 168 184 200 216 232 248
¨ È Ø è ø
98 A8 B8 C8 D8 E8 F8
137 153 169 185 201 217 233 249
© ¹ É Ù é ù
99 A9 B9 C9 D9 E9 F9
138 154 170 186 202 218 234 250
ª ° Ê Ú ê ú
9A AA BA CA DA EA FA
139 155 171 187 203 219 235 251
CSI « » Ë Û ë û
9B AB BB CB DB EB FB

Technical Manual
5-15
Native language support

140 156 172 188 204 220 236 252

¼ Ì Ü ì ü
9C AC BC CC DC EC FC
141 157 173 189 205 221 237 253
OSC ½ Í Ý í ý
9D AD BD CD DD ED FD
142 158 174 190 206 222 238 254
PM Î î
9E AE BE CE DE EE FE
143 159 175 191 207 223 239 255
APC ¿ Ï ß ï
9F AF BF CF DF EF FF

Technical Manual
5-16
 Native language support

List of compose sequences

Character Description Composition
NBSP Nonbreaking space \s\s
| Broken bar ||
" Diaeresis ""
# Quotation mark ++
@ Commercial at sign Aa
[ Opening bracket ((
\ Backslash //
] Closing bracket ))
{ Opening brace (-
| Vertical bar (or pipe) /^
} Closing brace )-
¡ Inverted exclamation mark !!
¢ Cent sign C/
£ Pound sign L-
¥ Yen Sign Y-
§ Section sign So
¤ Currency sign Xo
© Copyright sign CO
® Registered trademark sign RO
ª Feminine ordinal indicator A_
« Left angle quotation mark <<
¬ Not sign -,
SHY Soft hyphen -\
- Macron =-
° Degree symbol, Ring above 0^
± Plus or minus +-
² Superscript two 2^
³ Superscript three 3^
' Acute accent ''
× Multiplication sign Xx
÷ Division sign :-
µ Micron sign /u

Technical Manual
5-17
Native language support

Character Description Composition

¶ Paragraph mark P!
· Middle dot .^
º Masculine ordinal indicator o_
, Cedilla ,,
¹ Superscript one 1^
» Closing chevrons >>
¼ Common fraction one-quarter 14
½ Common fraction one-half 12
³ Common fraction three-quarter 34
¿ Inverted question mark ??
À Uppercase A with grave accent A`
Á Uppercase A with acute accent A'
 Uppercase A with circumflex accent A^
à Uppercase A with tilde A~
Ä Uppercase A with diaeresis A"
Å Uppercase A with ring above A*
Æ Uppercase ligature A with E AE
Ç Uppercase C with cedilla C,
È Uppercase E with grave accent E`
É Uppercase E with acute accent E'
Ê Uppercase E with circumflex accent E^
Ë Uppercase E with diaeresis E"
Ì Uppercase I with grave accent I`
Í Uppercase I with acute accent I'
Î Uppercase I with circumflex accent I^
Ï Uppercase I with diaeresis I"
Ð Uppercase Icelandic ETH D-
Ñ Uppercase N with tilde N~
Ò Uppercase O with grave accent O`
Ó Uppercase O with acute accent O'
Ô Uppercase O with circumflex accent O^
Õ Uppercase O with tilde O~
Ö Uppercase O with diaeresis O"
Ø Uppercase O with oblique stroke O/

Technical Manual
5-18
 Native language support

Character Description Composition

Þ Uppercase Icelandic thorn TH
Ù Uppercase U with grave accent U`
Ú Uppercase U with acute accent U'
Û Uppercase U with circumflex accent U^
Ü Uppercase U with diaeresis U"
Ý Uppercase Y with diaeresis Y"
ß Lowercase German sharp s ss
à Lowercase a with grave accent a`
á Lowercase a with acute accent a'
â Lowercase a with circumflex accent a^
ã Lowercase a with tilde a~
ä Lowercase a with diaeresis a"
å Lowercase a with ring above a*
æ Lowercase ligature a with e ae
ç Lowercase c with cedilla c,
è Lowercase e with grave accent e`
é Lowercase e with acute accent e'
ê Lowercase e with circumflex accent e^
ë Lowercase e with diaeresis e"
ì Lowercase i with grave accent i`
í Lowercase i with acute accent i'
î Lowercase i with circumflex accent i^
ï Lowercase i with diaeresis i"
ð Lowercase Icelandic ETH d-
ñ Lowercase n with tilde n~
ò Lowercase o with grave accent o`
ó Lowercase o with acute accent o'
ô Lowercase o with circumflex accent o^
õ Lowercase o with tilde o~
ö Lowercase o with diaeresis o"
ø Lowercase o with oblique stroke o/
þ Lowercase Icelandic thorn th
ù Lowercase u with grave accent u`
ú Lowercase u with acute accent u'

Technical Manual
5-19
Native language support

Character Description Composition

û Lowercase u with circumflex accent u^
ü Lowercase u with diaeresis u"
ý Lowercase y with acute accent y'
ÿ Lowercase y with diaeresis y"

Input/output table vt200

\160 \s\s
\161 !!
\162 c/
\163 L-
\164 xo
\165 Y-
\166 ||
\167 So
\168 ""
\169 cO
\170 a_
\171 <<
\172 -,
\173 -\\
\174 rO
\175 =-
\176 0\^
\177 +-
\178 2^
\179 3^
\180 ’’

Technical Manual
5-20
 Native language support

\181 /u
\182 p!
\183 .\^
\184 ,,
\185 1^

\186 o_
\187 >>
\188 14
\189 12
\190 34
\191 ??
\192 A‘
\193 A’
\194 A\^
\195 A~
\196 A"
\197 A*
\198 AE
\199 C,
\200 E‘
\201 E’
\202 E\^
\203 E"
\204 I‘
\205 I’
\206 I\^
\207 I"

Technical Manual
5-21
Native language support

\208 D-
\209 N~
\210 O‘
\211 O’
\212 O\^
\213 O~
\214 O"
\215 xx
\216 O/
\217 U‘
\218 U’
\219 U\^
\220 U"
\221 Y’
\222 TH
\223 ss
\224 a‘
\225 a’
\226 a\^
\227 a~
\228 a"
\229 a*
\230 ae
\231 c,
\232 e‘
\233 e’
\234 e\^

Technical Manual
5-22
 Native language support

\235 e"
\236 i‘
\237 i’
\238 i\^
\239 i"
\240 d-
\241 n~
\242 o‘
\243 o’
\244 o\^
\245 o~
\246 o"
\247 :-
\248 o/
\249 u‘
\250 u’
\251 u\^
\252 u"
\253 y’
\254 th
\255 y"

Technical Manual
5-23
Native language support

Output table
\160 \s
\164 \168
\166 |
\168 "
\172 \183
\173 -
\174 \183
\175 \183
\180 \183
\184 \183
\190 \183
\208 D
\215 x
\222 P
\240 \183
\247 /
\253 y
\254 p
\255 \253

Technical Manual
5-24
 Native language support

Printer output table MT910 HP Prestige

character set
\160 \s
\161 \184
\162 c\b|
\163 \175
\164 \186
\165 \188
\166 |
\167 \189
\168 \171
\169 \s
\170 a\b_
\171 <<
\172 \176
\173 -
\174 \s
\175 \176
\176 \179
\177 \254
\178 \s
\179 \s
\180 \168
\181 u
\182 \s
\183 .
\184 ,

Technical Manual
5-25
Native language support

\185 \s
\186 \250
\187 >>
\188 \247
\189 \248
\190 \s
\191 \185
\192 \161
\193 \224
\194 \162
\195 \225
\196 \216
\197 \208
\198 \211
\199 \180
\200 \163
\201 \220
\202 \164
\203 \165
\204 \230
\205 \229
\206 \166
\207 \167
\208 \227
\209 \182
\210 \232
\211 \231

Technical Manual
5-26
 Native language support

\212 \223
\213 \233
\214 \218
\215 x
\216 \210
\217 \173
\218 \237
\219 \174
\220 \219
\221 \89
\222 \240
\223 \222
\224 \200
\225 \196
\226 \192
\227 \226
\228 \204
\229 \212
\230 \215
\231 \181
\232 \201
\233 \197
\234 \193
\235 \205
\236 \217
\237 \213
\238 \209

Technical Manual
5-27
Native language support

\239 \221
\240 \228
\241 \183
\242 \202
\243 \198
\244 \194
\245 \234
\246 \206
\247 /
\248 \214
\249 \203
\250 \199
\251 \195
\252 \207
\253 y\b\168
\254 \241
\255 \239

Technical Manual
5-28
 Native language support

Overview of Esc/Ctrl codes

ASCII Dec code Hec code Oct code Esc code Ctrl code
mnemonic
SOH 0001 0x01 0001 ^A
STX 0002 0x02 0002 ^B
ETX 0003 0x03 0003 ^C
EOT 0004 0x04 0004 ^D
ENQ 0005 0x05 0005 ^E
ACK 0006 0x06 0006 ^F
BEL 0007 0x07 0007 ^G
BS 0008 0x08 0010 \b ^H
HT 0009 0x09 0011 ^I
LF 0010 0x0A 0012 \n ^J
VT 0011 0x0B 0013 \v ^K
FF 0012 0x0C 0014 \f ^L
CR 0013 0x0D 0015 \r ^M
SO 0014 0x0E 0016 ^N
SI 0015 0x0F 0017 ^O
DLE 0016 0x10 0020 ^P
DC1 (Xon) 0017 0x11 0021 ^Q
DC2 0018 0x12 0022 ^R
DC3 (Xoff) 0019 0x13 0023 ^S
DC4 0020 0x14 0024 ^T
NAK 0021 0x15 0025 ^U
SYN 0022 0x16 0026 ^V
ETB 0023 0x17 0027 ^W
CAN 0024 0x18 0030 ^X
EM 0025 0x19 0031 ^Y
SUB 0026 0x1A 0032 ^Z
ESC 0027 0x1B 0033 \e ^[
FS 0028 0x1C 0034 ^\
GS 0029 0x1D 0035 ^]
RS 0030 0x1E 0036 ^^
SPACE 0032 0x20 0040 \s

Technical Manual
5-29
Native language support

Technical Manual
5-30
6 Information files

Terminal information file

A terminal information file stores information including the type of terminal,
terminal operations, how to access keys on the keyboard, color support, cursor
movements, whether code features are blinking, bold, reverse, and so on.
Terminal information files are stored in the directory $BSE/lib/terminf.
To create a new terminal information file, create a subdirectory in the terminf
directory whose name is the first letter of the future terminal information file.
The terminal information file is stored in that subdirectory. For example, the
terminal information file for a vt100 terminal is stored in the directory:
$BSE/lib/terminf/v/
A terminal information file can contain three types of entries:
n Boolean entries
n Numbers
n Strings
Each entry is closed by a comma (,). Comment text can be added, preceded by a
number sign (#). You can include another terminal information file at any
position in the current file with the line, for example:
include=vt100,

Boolean entries
A Boolean entry reflects a feature of the terminal. A Boolean entry is set to true
by including it in the terminal information file. It is set to false either by not
including it or by prefixing it by an exclamation point (!), for example, !vt100.
Here are the available Booleans and what they mean when set to true:
aux_dual
Indicates that output to the aux port also appears on the terminal screen. The
terminal screen is disabled during aux printing.
color_cf
Provides color control in combination with code features.

Technical Manual
6-1
Information files

color_sep
Colors and code features are controlled separately.
mcf
You can move the cursor when the code feature (cf) is active.
mgr
You can move the cursor when the program is in graphics mode.
nclrwc
The terminal does not clear the screen when the width is switched from 80 to 132
columns or vice versa (No CLeaR on Window (width) Change).
vt100
After a character is entered on the last position (80 or 132), the cursor remains on
that position. The following character causes the cursor to jump to the next line.
xhp
Code features and graphic characters are not removed when they are overwritten
by characters in normal mode, but have to be removed explicitly (HP terminals).

Numbers
Numbers are variables that indicate the size of the terminal or particular features.
The following numbers are available:
cf_pos=
Number of positions required to switch a code feature on and off. The default
value is 0. The maximum value depends on the terminal type.
cols=
Number of available columns on the screen. The maximum is 132. If this value is
set to 80 you cannot switch the display to 132 positions.
disp_vt100=
Number of columns the cursor is to move if, in the last column (80th or 132nd),
the terminal receives a backspace (vt100 mode). The default value is 1. The
maximum value depends on the terminal type.
gr_pos=
Number of positions required to switch the graphics mode on and off. The
default value is 0. The maximum value depends on the terminal type.
lines=
Number of lines on screen (variable). The environment variable LINES can
overwrite this value.

Technical Manual
6-2
 Information files

tmo=
Timeout sequence after pressing ESC. The given number divided by 10 specifies
the time (in seconds) in which a sequence can be typed in after ESC. If no
character is pressed between this time, the pressing of ESC is handled normally.
The default timeout sequence is 5 (0.5 second).

Strings
String variables are filled with a sequence of characters that can be used to
perform particular terminal operations.
You can use the following special characters to assign string variables:

Escape codes: Control codes: ASCII-code

\b backspace ^A t/m ^Z (1-26)
\e or \E escape ^[ (27)
\f form feed ^\(28)
\n new line ^](29)
\r carriage return ^^ (30)
\t tab ^_ (31)
\s space
\xx decimal value xx= 1-255
\0xx octal value xx= 1-377
\xxx hexadecimal value xx= 1-FF (for example \x1F)

You can also use an octal value instead of an alphabetical representation. For
example, “escape” becomes \033. The control codes are converted to the ASCII
codes 1 to 31. For example, ^B equals the ASCII character with decimal value 2
(STX).
The following string variables are possible:
aux_off=
Closes the auxiliary port. All data is displayed on the terminal.
aux_on=
Opens the auxiliary port. All data sent to the terminal is passed to the aux port.
bell=
Audible signal (usual value: ^G).
cadr=
Absolute column positioning string. See the section “Parameterized strings”
about the use of a parameterized string.

Technical Manual
6-3
Information files

cbl=
Character used in graphics mode to display the lower left corner (+). If it is not
available, a viable alternative is +. Do not leave the variable empty, because this
can result in screen layout problems.
cbr=
Character used in graphics mode to display the lower right corner (+). If it is not
available, a viable alternative is +. Do not leave the variable empty, because this
can result in screen layout problems.
cf_b=
Code feature, blinking.
cf_br=
Code feature, blinking, and reverse.
cf_bru=
Code feature, blinking, reverse, and underlined.
cf_bu=
Code feature, blinking, and underlined.
cf_B=
Code feature, bold.
cf_Bb=
Code feature, bold, and blinking.
cf_Bbr=
Code feature, bold, blinking, and reverse.
cf_Bbru=
Code feature, bold, blinking, reverse, and underlined.
cf_Bbu=
Code feature, bold, blinking, and underlined.
cf_Br=
Code feature, bold, and reverse.
cf_Bru=
Code feature, bold, reverse, and underlined.
cf_Bu=
Code feature, bold, and underlined.
cf_n=
Code feature, normal.
cf_r=
Code feature, reverse.

Technical Manual
6-4
 Information files

cf_ru=
Code feature, reverse, and underlined.
cf_u=
Code feature, underlined.
clear=
Clears the entire display and moves the cursor to the top left corner of the screen.
csr=
Character sequence for changing the scrolling region. The first parameter
indicates the start line. The second parameter indicates the end line of the
scrolling region. See also the section “Parameterized strings.”
ctl=
Character used in graphics mode to display the upper left corner (+). If it is not
available, a viable alternative is +. Do not leave the variable empty, because this
can result in screen layout problems.
ctr=
Character used in graphics mode to display the upper right corner (+). If it is not
available, a viable alternative is +. Do not leave the variable empty, because this
can result in screen layout problems.
cu_off=
Cursor off (invisible).
cu_on=
Cursor on (visible).
cub=
The cursor moves back the specified number of columns. See the section
“Parameterized strings.”
cub1=
The cursor moves back one position.
cud=
The cursor moves down the specified number of rows. See the section
“Parameterized strings.”
cud1=
The cursor moves down one row.
cuf=
Move the cursor forward by the specified number of columns. See the section
“Parameterized strings.”
cuf1=
The cursor moves forward one position.

Technical Manual
6-5
Information files

cup=
Absolute cursor positioning. See the section “Parameterized strings.”
cuu=
The cursor moves up by the specified number of rows. See the section
“Parameterized strings.”
cuu1=
The cursor moves up one row.
delch=
Deletes the character below the cursor.
delli=
Deletes one line.
dt=
Character used in graphics mode to display the lower T-bar. If it is not available,
a viable alternative is +. Do not leave the variable empty, because this can result
in screen layout problems.
el=
Deletes to the end of the row.
es=
Deletes to the end of the screen.
ewin=
Erases the current window.
flash=
Flashes the screen instead of beeping when an error occurs.
gr_off=
Turns graphics mode off.
gr_offB=
Turns bold graphics mode off.
gr_on=
Turns graphics mode on.
gr_onB=
Turns bold graphics mode on.
hb=
Character used in graphics mode to display the horizontal bar. If it is not
available, a viable alternative is –. Do not leave the variable empty, because this
can result in screen layout problems.

Technical Manual
6-6
 Information files

home=
Positions the cursor in the upper left corner of the screen.
if=
Name of file containing codes to initialize the terminal. This initialization is done
after the initialization string in init and before the string in init2. Within the file
you can, for example, specify a large number of codes that do not fit in a string
of soft font codes.
inch=
Insert one character, usually a space at the current cursor position.
init=
Initialization string sent to the terminal to set it up for proper use by the Bshell.
This string is sent before any other output.
init2=
Initialization string sent to the terminal to set up. This string is sent during
initialization of the terminal, after the string init and the codes in the file
specified at entry if.
inli=
Insert one line.
kr=
Character used in graphics mode to display the center crossbar (+). If it is not
available, a viable alternative is +. Do not leave the variable empty, because this
can result in screen layout problems.
ladr=
Absolute line positioning string. See the UNIX manual TERMINFO(4) and the
section “Parameterized strings” about defining such a string with one parameter.
lt=
Character used in graphics mode to display the left T-bar. If it is not available, a
viable alternative is +. Do not leave the variable empty, because this can result in
screen layout problems.
nls_in=
Name of NLS input conversion table. See the NLS4.2 user manual.
nls_out=
Name of NLS output conversion table. See the NLS4.2 user manual.
rc=
Restores cursor position and attributes saved with the character sequence
specified with entry sc.

Technical Manual
6-7
Information files

reset=
Reset initialization string. The terminal is set up in such a way that it can be used
in UNIX or other programs. The string is sent after the Bshell terminates
execution.
rt=
Character used in graphics mode to display the right T-bar. If it is not available, a
viable alternative is +. Do not leave the variable empty, because this can result in
screen layout problems.
sc=
Save cursor position and attributes.
set0=, to set9=
Set a character set of the terminal. The default set is set0.
set80=
Set the display width to 80 column.
set132=
Set the display width to 132 columns.
ut=
Character used in graphics mode to display the upper T-bar. If it is not available,
a viable alternative is +. Do not leave the variable empty, because this can result
in screen layout problems.
vb=
Character used in graphics mode to display the vertical bar. If it is not available,
a viable alternative is | (pipeline). Do not leave the variable empty, because this
can result in screen layout problems.
The following dim code features are used in graphics mode if the normal code
features are defective. (Poppy and Wyse terminals):
cblB=. See cbl, but bold.
cbrB=. See cbr, but bold.
ctlB=. See ctl, but bold.
ctrB=. See ctr, but bold.
dtB=. See dt, but bold.
hbB=. See hb, but bold.
krB=. See kr, but bold.
ltB=. See lt, but bold.
rtB=. See rt, but bold.

Technical Manual
6-8
 Information files

utB=. See ut, but bold.

vbB=. See vb, but bold.

Keys
This section describes how you can access keys on the keyboard. Use the string
assigned to each entry to specify either the sequence that simulates the key or the
sequence that is generated when the user presses the key.
If you want to use a special key in the Bshell, you have to specify this key in the
terminal information file. If the key is present on the keyboard, you must assign
the corresponding sequence to it. If the key is not present, you can assign another
sequence.
For several key entries, a default value is given along with the ASCII code and
decimal code. For example:
CTRL+J (LF, 10d)
The following keyboard entries are possible:
k1=, to k20=
Specifies the function keys F1 to F20.
k1c=, to k20c=
Specifies the function keys when the CTRL key is held down.
k1s=, to k20s=
Specifies the function keys when the SHIFT key is held down.
k1sc=, to k20sc=
Specifies the function keys when the SHIFT and CTRL keys are held down.
kbs=
Backspace; moves cursor one position back. Default: CTRL+H (BS, 8d).
kbtab=
Specifies the back tab key.
kcompose=
Key compose; used to define a code for NLS character composition.
kcs=
Key column switch. If it cannot display all 132 columns at once, switches
function between the left and right half of the display (usually \eC or \ec).
kd=
Key down; moves cursor one position down. Default: CTRL+J (LF, 10d).

Technical Manual
6-9
Information files

kdc=
Key delete character; removes one character. Default: <Del> (127d).
kdo=
Key do.
kesc=
Key escape.
kf=
Key forward; moves cursor one position to the right. Default: CTRL+L (FF, 12d).
kfind=
Key find.
khlp=
Key help.
khome=
Key home; moves the cursor to the upper left corner of the screen. Default:
CTRL+^ (RS, 31d).
kid=
Key interrupt debugging.
kinshere=
Key insert.
kkill=
Key kill.
kl=
Key left; moves the cursor one position to the left. Default: CTRL+H (BS, 8d)
knscreen=
Key page down.
kpr=
Key print.
kpscreen=
Key page up.
kre=
Key refresh; rewrites screen (usually \eR or \er).
kremove=
Key remove.
kselect=
Key select.

Technical Manual
6-10
 Information files

ktab=
Key tab.
ku=
Key up; moves the cursor up one line. Default: CTRL+K (VT, 11d).
pf1=, to pf4=
DEC vt100 “gold” keys. NOTE (only for Baan-C programmers):
Using event functions, your program can catch the keys defined in your terminal
information file. When the program catches a key-press event, it returns a code
for that key. The codes used for the keys are shown in the header file
$BSE/include6.2/bic_key.
For example, a line in the header file is:
#define KEY_F(1) 1024
This indicates that the F1 function key (k1 in a terminal information file) is
identified by code 1024. You can use the macro KEY_F(1) in your program to
check if the defined sequence for the F1 function key is given.
See the section ”Example of terminal information files” for an extensive example
of terminal information files.

Color support
Colors can be sent to the screen separately or in combination with code features,
depending on the capabilities of the terminal. This is set in the terminal
information file with the Boolean values color_cf and color_sep.
color_cf
The code feature can be sent together with the color codes in one single escape
sequence.
color_sep
The color codes cannot be combined with the code features.

Sending colors separately

You can use the following string entries in the terminal information files to set
the foreground and background color:
bg_black=
bg_red=
bg_green=

Technical Manual
6-11
Information files

bg_yellow=
bg_blue=
bg_magenta=
bg_cyan=
bg_white=,
fg_black=
fg_red=
fg_green=
fg_yellow=
fg_blue=
fg_magenta=
fg_cyan=
fg_white=
You can also assign escape sequences to these entries. For example:
color_sep
bg_black=\e[40m
bg_red=\e[41m
bg_green=\e[42m
bg_yellow=\e[43m
bg_blue=\e[44m
bg_magenta=\e[45m
bg_cyan=\e[46m
bg_white=\e[47m,
fg_black=\e[30m
fg_red=\e[31m
fg_green=\e[32m
fg_yellow=\e[33m
fg_blue=\e[34m
fg_magenta=\e[35m
fg_cyan=\e[36m

Technical Manual
6-12
 Information files

fg_white=\e[37m,

Colors in combination with code features

The following numbers are used as %p1 and %p2 parameters in combined code
feature and color strings (color_cf).
black=0
red=1
green=2
yellow=3
blue=4
magenta=5
cyan=6
white=7
The colors can be sent to the terminal along with code features with the
following string entries. The foreground and background color must be specified
as parameters.
cf_co_n=, # colors as specified and code feature normal
cf_co_B=, # code feature bold
cf_co_b=, # code feature blinking
cf_co_Bb=, # code feature bold and blinking
cf_co_r=, # code feature reverse
cf_co_Br=, # code feature bold and reverse
cf_co_br=, # code feature blinking and reverse
cf_co_Bbr=, # code feature bold, blinking and reverse
cf_co_u=, # code feature underlined
cf_co_Bu=, # code feature bold underlined
cf_co_bu=, # code feature blinking underlined
cf_co_Bbu=, # code feature bold, blinking underlined
cf_co_ru=, # code feature reverse underlined
cf_co_Bru=, # code feature bold, reverse underlined
cf_co_bru=, # code feature blinking, reverse underlined

Technical Manual
6-13
Information files

cf_co_Bbru=, # code feature bold, blinking, reverse underlined

For example:
cf_co_n=\e[0;22;%p1%{30}%+%d;%p2%{40}%+%dm
cf_co_B=\e[0;1;%p1%{30}%+%d;%p2%{40}%+%dm
cf_co_b=\e[0;5;22;%p1%{30}%+%d;%p2%{40}%+%dm
cf_co_Bb=\e[0;1;5;%p1%{30}%+%d;%p2%{40}%+%dm
cf_co_Br=\e[0;1;7;%p1%{30}%+%d;%p2%{40}%+%dm,
In the previous example, %p1 is foreground color; p1=0-7 and %p2 is
background color; p2=0-7. See the section, “Colors in combination with code
features.” For use of parameters within a string entry, see the section
“Parameterized strings.”

Parameterized strings
Some of the string entries described in the section “Terminal information file”
require a number of parameters. For example, to address the cursor, the cup entry
requires two parameters: the row and column to address to. The use of
parameters within a string entry is described in this section.
The parameter mechanism uses a stack and special codes, starting with % to
manipulate the stack.

Technical Manual
6-14
 Information files

The codes have the following meanings:

Code Meaning
%d Take a number from the stack (pop())
%2d Take a number with a minimum of two digits
%3d Take a number with a minimum of three digits
%02d Take a number with exactly two digits
%03d Take a number with exactly three digits
%c Take a character from the stack
%s Take a string from the stack
%P[a-z] Take a variable [a-z] from the stack
%p[1-9] Push a parameter [1-9] onto the stack
%g[a-z] Get a variable [a-z] and push it onto the stack
%{nn} Push a decimal constant nn onto the stack
%+ Take two numbers from the stack and push their sum
%- Take two numbers from the stack and push their subtraction; push(pop() – pop())
%* Take two numbers from the stack and push their multiplication
%/ Take two numbers from the stack and push their division; push(pop() / pop())
%m Take two numbers from the stack and push the modulus of their division;
push(pop() mod pop())
%& Take two numbers from the stack and push the result of their logical AND.
%| Take two numbers from the stack and push the result of their logical OR.
%\^ Take two numbers from the stack and push the result of their logical EXOR.
%= Take two numbers from the stack and compare them. Dependent on the equality,
TRUE or FALSE is pushed onto the stack.
%> Take two numbers from the stack and compare them. If the first number is greater
than the second, TRUE is pushed onto the stack, otherwise FALSE is pushed.
%< Take two numbers from the stack and compare them. If the first number is less
than the second, TRUE is pushed onto the stack, otherwise FALSE is pushed.
%! Take a number from the stack and push its negative value onto the stack.
%\~ Take a number from the stack and push the result of its logical NOT.
%I Add 1 to the first two parameters (for ANSI terminals). Binary operations are in
postfix form with the operands in the usual order. That is, to get x – 5, you must
use %gx%{5}%-

EXAMPLES %i\e[%p1%d;p2%dr
This example is used for changing the scrolling region on C.Itoh terminals. The
first parameter +1 specifies the start row, the second parameter +1 specifies the
end row of the region. Both are used as numerics in the escape sequence.

Technical Manual
6-15
Information files

\e[0;22;%p1%{30}%+%d;%p2%{40}%+%dm
The sequence in this example is used for setting a color and code feature. The
code feature normal is set, the foreground color is set to the first parameter +30,
and the background color is set to the second parameter +40.

Example of terminal information files

This section contains examples of terminal information files for some terminal
types. Examples are given of include files, color usage, function keys, setting
fonts, and so on.
Comments and explanations are added where the number sign (#) precedes the
text.
#*******************************************************
# @(#)
# @(#) File : vt100
# @(#)
# @(#) Terminal : VTxxx terminals (CITOH / DEC)
# @(#) Setup : VT100 – 7 bit;
# @(#) Multinational Char. Set
# @(#) Emulation Mode : ANSI 8-bit
# @(#) Character Set : ISO Latin-1
# @(#)
# @(#) Copyright : Baan Service b.v. (c)
# @(#) Customer : Baan International b.v.
# @(#)
# @(#) Comment : If the terminal has a DEC Supplemental char.set
# @(#) then vtansi_d must be included instead of vtansi
# @(#)
#*******************************************************
# Include files
#*******************************************************
include=vtansi,
#*******************************************************
# @(#)
# @(#) File : vtansi
# @(#)
# @(#) Comment : Generic terminf-file for VTxxx terminals
# @(#) supporting ANSI mode and ISO character set
# @(#)
#*******************************************************
#*****************************
# Include files
#*****************************
include=vtansi.fkeys,

Technical Manual
6-16
 Information files

#*****************************
# Booleans
#*****************************
vt100
#nclrwc
mcf
mgr,
#*****************************
# Numbers
#*****************************
cols=132
lines=24
cf_pos=0
gr_pos=0
tmo=5
disp_vt100=0,
#*****************************
# Strings
#*****************************
# Setup
#
# init \e[?3l set 80 cols
# \e[0;1m reset attributes and set bold
# \e> set keypad mode to numeric
# \e[1l set cursor key mode to normal
# \e[?25l set cursor mode to visible
# init2 \e(B ASCII –> G0
# (fonts) \e-A Latin-1 –> G1
# ^O (SI) invoke G0 into GL
# gr_on \e(0 Spec.Graph –> G0
# reset \e[0m reset attributes
# \e[H cursor position 0,0
# \e[2J reset entire screen
#
init=\e[?3l\e[0;1m\e>\e[1l\e[?25l
init2=\e(B^O\e-A
reset=\e[0m\e[0;0H\e[2J\e(B^O\e-A,
set80=\e[?3l
set132=\e[?3h,
bell=^G
tmo=20,
csr=%i\e[%p1%d;%p2%dr,
# Line Draw Chars (Char.set: DEC Special Graphic)
#
gr_on=\e(0^O
gr_ond=\e(0^O\e[0m

Technical Manual
6-17
Information files

gr_off=\e(B^O
gr_offd=\e(B^O\e[0m,
cbl=m
cbld=m
cbr=j
cbrd=j
ctl=l
ctld=l
ctr=k
ctrd=k
dt=v
dtd=v
ut=w
utd=w
lt=t
ltd=t
rt=u
rtd=u
kr=n
krd=n
hb=q
hbd=q
vb=x
vbd=x,
# Code Features
#
cf_n=\e[0m
cf_d=\e[0;1m
cf_b=\e[0;5m
cf_db=\e[0;1;5m
cf_r=\e[0;7m
cf_dr=\e[0;1;7m
cf_br=\e[0;5;7m
cf_dbr=\e[0;1;5;7m
cf_u=\e[0;4m
cf_du=\e[0;1;4m
cf_bu=\e[0;4;5m
cf_dbu=\e[0;1;4;5m
cf_ru=\e[0;4;7m
cf_dru=\e[0;1;4;7m
cf_bru=\e[0;4;5;7m
cf_dbru=\e[0;1;4;5;7m,
# Cursor Control & Positioning
#
cu_on=\e[?25h
cu_off=\e[?25l

Technical Manual
6-18
 Information files

cup=%i\e[%p1%d;%p2%dH,
cuu=\e[%p1%dA
cuu1=\e[1A
cud=\e[%p1%dB
cud1=\e[1B
cuf=\e[%p1%dC
cuf1=\e[1C
cub=\e[%p1%dD
cub1=\e[1D,
home=\e[0;0H
clear=\e[0;0H\e[2J\e[?25l,
# Editing
#
inch=\e[@
delch=\e[P
inli=\e[L
delli=\e[M
el=\e[0K
es=\e[0J
#ewin=,
# Printer Control
#
aux_on=\e[5i
aux_off=\e[4i,
# Nls
#
nls_in=generic.in,
#********************************************************
# @(#)
# @(#) File : vtansi.fkeys
# @(#)
# @(#) Comment : Include-file for VTxxx terminals function-keys
# @(#) (ANSI-compatible)
# @(#)
#********************************************************

# Available Keys
#*****************************
# Function Keys
#
#k6=\e[17~
#k7=\e[18~
#k8=\e[19~
#k9=\e[20~
#k10=\e[21~
k11=\e[23~

Technical Manual
6-19
Information files

k12=\e[24~
k13=\e[25~
k14=\e[26~
#k15=\e[28~
#k16=\e[29~
#k17=\e[31~
#k18=\e[32~
#k19=\e[33~
#k20=\e[34~,
# VT "Gold Keys"
#
pf1=\eOP
pf2=\eOQ
pf3=\eOR
pf4=\eOS,
# VT Cursor Pad
#
khlp=\e[28~
kdo=\e[29~
kfind=\e[1~
kinshere=\e[2~
kremove=\e[3~
kselect=\e[4~
kpscreen=\e[5~
knscreen=\e[6~
ku=\e[A
kd=\e[B
kf=\e[C
kl=\e[D,
#*****************************
#BA6.2 Special Keys
#*****************************
#Misc. Keys
#
# F6 = kmenu F17 = kprocess
# F7 = kcompose F18 = kkill
# F8 = kre F19 = knextdisp
# F9 = kcs F20 = kprevdisp
# F10 = kswitch
#
kmenu=\e[17~,
kcompose=\e[18~,
kre=\e[19~,
kcs=\e[20~,
kswitch=\e[21~,
kprocess=\e[31~,

Technical Manual
6-20
 Information files

kkill=\e[32~,
knextdisp=\e[33~,
kprevdisp=\e[34~,
#*****************************

Printer information files

The directory where printer information files are stored is $BSE/lib/printinf.
Create a subdirectory in the directory printinf. Its name must be the first letter of
the future printer information file. The printer information file is stored in this
subdirectory. For example, the printer information file for a Mannesmann mt910
is stored in the directory $BSE/lib/printinf/m/
The name of the file itself is an abbreviation of the printer name, for example,
mt910.
You can include another printer information file at any position in the current file
with the include command. For example, include=mt910
The variables are filled in accordance with the specifications (escape sequences)
of the printer as listed in the printer manual. Some of them can be omitted,
because they are not required for the majority of the applications. These variables
are marked with *.
A variable definition consists of a name, a =, an escape sequence (containing
special characters), and a comma. Lines are closed by a new line character.
If a printer does not offer an escape sequence for boldface and underlined text,
and the specific variables are left empty, the filter program simulates these
functions by backspace and carriage return. The positions are then covered twice.
You switch this simulation off by entering the value \000 for the definition of
bold on/off or underline on/off, or by entering no value (for example pbold=,).

Technical Manual
6-21
Information files

The escape sequences can contain special characters. They can be divided into
escape codes and control codes.

Escape codes Control codes

\b backspace ^A to ^Z
\E or \e escape ^[
\f form feed ^\
\n newline ^]
\r carriage return ^^
\t tab ^_
\s space
\xx decimal value xx=1-255
\0xx octal value xx=1-377
\xxx hexadecimal value xx=1-FF (for example \x1F)

You can use an octal value instead of an alphabetical representation. For

example, “escape” then becomes \033.
The control codes are converted to the ASCII codes 1 to 31. For example, ^B
equals the ASCII character with decimal value 2 (STX).
A printer information file can have the following Boolean entries:
n rsf_pbold, reset font after pbold
n rsf_prev, reset font after prev
n rsf_punder, reset font after punder
n rsf_pobold, reset font after pobold
n rsf_porev, reset font after porev
n rsf_pounder, reset font after pounder
A printer information file can have the following variables:
barcode_dir=
Bar-code directory (relative to $BSE/lib/barcode)
bin1=
Select first paper bin.
bin2=
Select second paper bin.
bin3=
Select third paper bin.

Technical Manual
6-22
 Information files

bin4=
Select fourth paper bin.
hpos=
Set horizontal cursor position of the printer.
initpage=
String sent before each page.
initpr=
Initialization string preceding the output to the printer. This variable must be
filled with the codes for large typeface.
initprog=X
The output of program X (full path name) is sent to the printer.
initpr2=
The second initialization string sent after initprog.
landscape=
Select landscape printing.
large=
Large typeface (10 cpi) control. Some printers require a code to changethe
typeface size.
nls_out=
NLS output table name.
middle=
Medium typeface (12cpi) control. Some printers require a code to change the
typeface size.
You can use the following codes to set the print color:
p_black=
p_red=
p_green=
p_yellow=
p_blue=
p_magenta=
p_cyan=
p_white=,
pbold=
Print boldface characters. A boldface character can print wider than its original.
This does not imply any proportional spacing. Printing graphic characters (with
ASCII code over 127 decimal) yields additional line feeds on most Mannesmann
printers. If this variable is not used, the Bshell simulates boldface printing by
reprinting each line, making use of the carriage return character.

Technical Manual
6-23
Information files

pcbl=
Print +. If this variable cannot be filled because the printer cannot print it, for
example, a + is used.
pcbr=
Print +. If this variable cannot be filled because the printer cannot print it, for
example, a + is used.
pctl=
Print +. If this variable cannot be filled because the printer cannot print it, for
example, a + is used.
pctr=
Print +. If this variable cannot be filled because the printer cannot print it, for
example, a + is used.
pdbl_wide=
Double wide mode on.
pdt=
Print –. If this variable cannot be filled because the printer cannot print it, for
example, a + is used.
pfont1=, to pfont16=
Select font 1 to font 16. These fonts are user definable.
phb=
Print –. If this variable cannot be filled because the printer cannot print it, for
example, a – is used.
pitalic=
Italic mode on.
pkr=
Print +. If this variable cannot be filled because the printer cannot print it, for
example, a + is used.
plt=
Print |. If this variable cannot be filled because the printer cannot print it, for
example, a + is used.
pnlq=
NLQ mode on.
pobold=
Switch off boldface printing without affecting character size.
podbl_wide=
Double wide mode off.

Technical Manual
6-24
 Information files

pofont1=, to pofont16=
Deselect font 1 to font 16. These fonts are user definable.
poitalic=
Italic mode off.
ponlq=
NLQ mode off.
porev=
Reverse mode off.
portrait=
Select portrait printing.
posubscript=
Subscript mode off.
posuperscript=
Superscript mode off.
pounder=
Underlined printing off.
prev=
Reverse mode (white on black).
prt=
Print |. If this variable cannot be filled, for example, because the printer cannot
print it, a + is used.
psubscript=
Subscript mode on.
psuperscript=
Superscript mode on.
punder=
Print underlined characters. If this variable cannot be filled, the Bshell simulates
underlining by printing underscore characters on the next line.
put=
Print –. If this variable cannot be filled because the printer cannot print it, for
example, a + is used.
pvb=
Print vertical bar. If this variable cannot be filled because the printer cannot print
it, for example, a | (pipeline) is used.

Technical Manual
6-25
Information files

q1=, to q13=
These 13 variables can be filled with strings that specify whether a character
must be printed in double width, italicized, and so on The variables are user-
definable. These codes have been replaced and they will be removed in a future
release.
resetpr=
Resets the printer. The string is sent to the printer after all output. Fill this
variable with a form feed (\014) followed by the Large typeface codes.
resetprog=X
The output of program X (full path name) is sent after the reset string.
resetpr2=
The second reset string, sent after resetprog.
set0=, to set9=
Set a character set of the printer. The default set is set0.
small=
Small typeface (16.66-17 cpi) control. Some printers require a code to change
typeface size.
usr1=, to usr16=
Define user entry1 to user entry16.

Bar codes
The report writer interfaces with the printer driver using bar-code scripts. A bar-
code script saves the cursor position, prints a bar code with given height and
returns to the saved cursor position. These scripts must be available in the
directory barcode_dir. This directory is relative to $BSE/lib/barcode, so if
barcode_dir = hp_barcode, the bar-code directory is
$BSE/lib/barcode/hp_barcode. Bar-code scripts are prefixed by the word “type”
and post fixed by a two-digit number. For example:
type01.

Technical Manual
6-26
 Information files

Following is an example of such a bar-code script. It saves the cursor position

and restores it after the bar code has been printed. Bar codes can only be
implemented when such a scheme is possible.
#!/bin/sh
#
# Sample driver for HP Laserjet 4
# with "Bar Codes & More Font Cartridge"
#
# Prints EAN/UPC with the code under it.
# $1 means the bar code
# $2 means the height of the bar code (number of lines)
#
code=$1
height=$2
Push()
{
# push cursor position (max 20x)
echo "\033&f0S\c"
}
Pop()
{
# pop cursor position
echo "\033&f1S\c"
}
NextRow()
{
# move to next row, relative
echo "\033&a+1R\c"
}
# – save cursor position – filter assumes same pos. after
bar code print
Push
str1=`echo $code | awk '{ print substr($1, 1, 5) }'`
str2=`echo $code | awk '{ print substr($1, 6, 5) }'`
# – select EAN/UPC 13 mil font
echo "\033(8Y\033(s1p12v0s3b0T\c"
while [ $height –gt 1 ]
do
Push
echo "($str1-$str2(\c"
Pop
NextRow

Technical Manual
6-27
Information files

height=`expr $height – 1`
done
# – last (empty) line of bar code
Push
echo "(- (\c"
Pop
# – select Courier 12cpi.
echo "\033(10U\033(s0p12.00h10.0v0s0b3T\c"
# – move x+25 dots
echo "\033*p+25X\c"
# – print code (text)
echo "$str1 $str2\c"
# – restore cursor position for filter
Pop
exit 0

Technical Manual
6-28
 Information files

Example of printer information file for mt910

printer
This shows what a printer information file can consist of. You can add comments
in the same manner as in a terminal information file.
#mt910
small=\E{s16.66H\EL08
large=\E{s10H
middle=\E{s12H
initpr=\E{s12H\EF66
resetpr=\E{s12H\EF66\014
pctl=+
pctr=+
pcbl=+
pcbr=+
put=+
prt=+
plt=+
pdt=+
phb=-
pvb=|
pkr=+
pbold=\E"g1B
pobold=\E"g0B
prev=\E"g1K
porev=\E"g0K
punder=\EI
pounder=\EJ
pitalic=\E{s1S
poitalic=\E{s0S
landscape=\E"g1R
portrait=\E"g0R
nls_out=mt910.out,

Technical Manual
6-29
Information files

Technical Manual
6-30
7 User Interface (UI) Page mode

This chapter describes the User Interface (UI) Page mode in BaanERP. The UI
Page mode improves the response times on networks where information
processing can be delayed, for example, a wide area network (WAN).
In normal mode, the 4GL Engine validates the entered data per field. In the UI
Page mode, the 4GL Engine validates data per page, and not per field. The
synchronous interaction between the UI driver and the Bshell is therefore
significantly reduced.
The UI Page mode is designed for experienced BaanERP users, because the
interaction between BaanERP and the user is not as extensive as in the normal
mode.
In the normal mode, an error message is displayed immediately when data
entered in the active field is not correct. In the UI Page mode, an error-logging
window appears, which shows all errors that occurred during the validation of all
the data entered on the page.
In the UI Page mode, the UI driver groups together a number of UI objects that
form a logical unit. A UI object is for example, a field, a button, or a check box.
The logical unit is known as a page. A page can be an overview window or a tab
in a details session with more than one tab.
If the UI Page mode is selected, each UI object automatically belongs to a page.
If the UI object is positioned on a tab, the object is considered part of the page
associated with the related tab. In any other case, the UI object is part of the page
associated with the overview window.
The following functions are implemented in the UI Page mode to improve
network performance:
n The UI handles the movement from an active element to another element
without any communication with the Bshell. An active element is a portion of
the screen that is currently operational or subject to command operations.
Usually the cursor or a highlighted section shows the active element on the
display screen. Active elements are, for example, fields, command buttons,
check boxes, or list boxes.
n The UI client handles the behavior of the default button for the command
buttons on a page.

Technical Manual
7-1
User Interface (UI) Page mode

n Events that result from status changes of UI objects are delayed until a
complete page is filled, or another synchronization event occurs. The delayed
events are sent into a minimum number of network frames.

Tab processing
When the UI page mode is selected, the UI driver processes Tab key movements
in the details window. The Tab key sequence is identical to the creation order of
the UI objects on a page. Other navigation keys, such as Page Up, Page Down,
Home and End are reported to the Bshell with the appropriate keystroke.

Default Button handling

When the UI Page mode is selected, the UI driver handles the default command
button. This is the command button that is highlighted when the dialog box is
initially displayed. It can also be the command button with the bold border,
indicating that it is automatically selected if you press Enter.

Event processing
If the UI Page mode is selected, events are treated differently by the UI driver
than when normal mode is selected. There are four different event categories:
n Suppressed events
n Delayed events
n Synchronization events
n Bypass events
Suppressed events
If the UI Page mode is selected, a UI object does not send suppressed events to
the Bshell . Instead, suppressed events are handled locally by the UI driver on the
client. An example of a suppressed event is moving from an active element to
another active element.
Delayed events
Status changes of UI objects can be delayed. As soon as a synchronization event
occurs, for example, by using the Validate button, the changed UI objects are
requested to synchronize their status with the 4GL engine. The delayed events
are sent in the same order as the Tab key sequence of the UI objects on the page.
This is not necessarily the order in which the user enters data. If possible, the
delayed events are sent into one network frame, to reduce response times.

Technical Manual
7-2
 User Interface (UI) Page mode

Synchronization events
The synchronization events make sure that the statuses of the changed UI objects
are sent to the Bshell for validation.
Synchronization events are:
n Selecting a menu item
n Clicking a button on a toolbar, or a pushbutton
n Moving to another tab
n Moving from a synchronizing field to another field
n Starting a browse session
n Pressing a key that is not handled by the UI client
n Resizing a window
n Using the scroll bar
Bypass events
The following bypass events are sent immediately to the Bshell without first
being synchronized with the delayed events:
n Using the online Help
n Closing a Windows application that was started through the application Start
feature of the UI client
n Using OLE objects

How to select the UI Page mode

The UI Page mode is selected per user to make sure that inexperienced BaanERP
users cannot use it. You must define the UI Page mode in the user profile of the
user. The user profile is defined in the User Management module in BaanERP
Tools. Follow these steps to select the UI Page mode for a user:
1 To select a user data template, or to create a new user data template, start the
User Data Template (ttams1110m000) session in the Authorization
Management System (AMS) module. Be sure to select the Use Page mode
check box under Options on the System Data tab.
2 Start the User Data (ttaad2500m000) session in the User Management
module. Double-click the user for which you want to select the UI Page mode
to start the details session.
3 Under Templates, enter the new User Data template, which is page-mode
enabled. Click Save to return to the overview session.
4 Click on the Specific menu Convert Changes to Runtime DD to rebuild the
user profile in the runtime database.

Technical Manual
7-3
User Interface (UI) Page mode

5 Restart BaanERP to activate the changed user profile.

Check that the following are true to verify that the UI Page mode has been
successfully selected:
n All the fields and options in the overview windows and detail windows are
available
n All commands on the menus of the menu bar in an overview window are
available
n On the Window menu the Validate command is available in an overview
window
n The Validate button shows on a details window

How to mark UI objects as synchronizing

fields
You can mark UI objects on an overview window or details window as
synchronizing fields. This means that when you move from one active field or
button to another field or button, all delayed events are sent to the Bshell for
validation. You can mark the following fields and buttons as synchronizing
fields:
n Edit fields
n Check boxes
n Option buttons
n List boxes
n Combo boxes
You must use the Form Editor in BaanERP Tools to mark a form field as a
synchronizing field. Complete the following steps to start the Form Editor:
1 On the BAANERP Tools menu, click Application Development, and then
click Forms to start the Forms (ttadv3500m000) session.
2 Select a form. On the Specific menu click Check Out to release the form for
modification.
3 On the Specific menu, click Edit/View Form... to start the Form Editor.
Depending on the form type, the appropriate Form Editor starts. The following
form editors are available in BaanERP:
n The Static Form Editor, which shows the form in an ASCII format
n The Dynamic Form Editor, which is a graphical editor in a Windows format

Technical Manual
7-4
 User Interface (UI) Page mode

You can carry out one of the following procedures to mark a field as a
synchronizing field:
Static Forms
1 In the ASCII editor, click the field to start the Form Fields (ttadv3501s000)
session.
2 On the General tab, select the Synchronized check box.
3 Click Close to return to the Form Editor.
4 On the Specific menu, click Check In to finish the procedure.
Dynamic Forms
1 In the graphical Dynamic Form Editor, click the field to start the Field
Properties session.
2 On the Miscellaneous tab, select the Synchronized check box.
3 Click OK to return to the graphical Dynamic Form Editor.
4 On the Specific menu, click Check In to finish the procedure.

Technical Manual
7-5
User Interface (UI) Page mode

Technical Manual
7-6
8 Customer support tools

General
This chapter describes two aspects essential to solving errors in a Bshell
environment: error logging and the bserel6.2 program. Error logging is keeping a
record of error messages in a log file. The log file contains information about
when the error occurred and which processes were then active. See the section
“Error logging.”
The program bserel6.2 provides information on the BSE environment as installed
on your system. Among other things, bserel6.2 checks permissions, the
availability of programs, and the correctness of program objects. This data can be
collected in a report and viewed on screen, printed, or sent to a file.
If you want to submit an error to the Customer Support Center of Baan Info
Systems, add both the error logging report and the bselrel6.2 report to your error
report.

Error logging
Certain actions in a Bshell program can result in an error. Usually an error is
displayed and from it you can see what went wrong.
To localize Bshell errors, a status report of the program is written to a log file.
Use this file to find out where and when the error occurred.

Log file
The log files the errors are written to are stored in directory $BSE/log. Its name
format is:
log.<program name>
This means that when an error occurs in Bshell6.2, it is logged in the file
log.Bshell6.2. If it occurs in bx6.2, it is logged in log.bx6.2, and so on.

Technical Manual
8-1
Customer support tools

Log file layout

The header of the log file is: "START of log message", followed by:
n Date, time, logon name, and TTY number
n Source program name with line number
n Keyword
n UNIX process, user, and group identifiers for program
n User type, language code, logon code and TTY number of the user
n error number and bdb error number
n error message as displayed on screen
If, for instance, you end a Bshell program with an Interrupt (CTRL+\), this action
is written as an error in the log file. The message keyword is “core dump.” The
log file also contains the following blocks:
n Process List
n Screen Dump
n Process Information
The Process List block is a list of the active Bshell processes at the moment the
error occurred. This list contains the following information for each Bshell
process:
n Process identifier
n Parent
n Status
n cpu.flags
n Process name
n Line number (debug mode only)

Technical Manual
8-2
 Customer support tools

The Screen dump block contains the screen active at the moment the error
occurred. The last block, Process Information, contains the individual process
information about each process from the list. It can consist of:
n The line from the process list for the process involved
n Session code, process ID, and parent number
n Object
n Main table
n Zoom field
n Return zoom field
n Default printer
n Forms
n Reports
n Tables
n Variables, including values in force when the error occurred (empty strings,
strings with value 0, and strings consisting of tabs are not logged). The
section “Example error logging” contains an example of error logging

Log file maintenance

If the log file is not regularly purged it grows enormously, which affects your
storage capacity. Also, a large log file is difficult to read. You are therefore
advised to rename the log file to log.save on a weekly basis, using the UNIX mv
command. After a backup has been made of the system including log.save, you
can remove it from the system. The backup tape should be kept for two weeks.

Submitting errors to customer support

If you are confronted with a Bshell error message for which you have no
solution, you can contact the Customer Support Center (formerly known as
Response Center). To your report you add the part of the log file relating to it.
You can print this part using the UNIX grep command (for example by date,
time, logon code, or TTY number).
The following section contains an example of error logging. The log file contains
two error reports. Suppose you want to print the first one. You can do this using
grep in four ways: by date, time, user name, or TTY number. The following
example shows the command lines for each of these:
cat log.bshell6.2 | grep 91-07-29 |pr |lp –d{printer name} date
cat log.bshell6.2 | grep 16:53:14 | pr |lp –d{printer name} time
cat log.bshell6.2 | grep bsp | pr |lp –d{printer name} name
cat log.bshell6.2 |grep 22 | pr | lp –d{printer name} TTY

Technical Manual
8-3
Customer support tools

Example error logging

NOTE Each line of the following output starts with date, time, logon name, and TTY
number, such as the following:
98-03-04[15:06:54]etimban:
98-03-04[15:06:54]etimban: ******* S T A R T of Log message *******
98-03-04[15:06:55]etimban: Log message called from /vobs/tt/lib/al_1/al_fpath.c: #241
keyword: sopen
98-03-04[15:06:55]etimban: Pid 19313 Uid 10575 Euid 10575 Gid 1700 Egid 1700
98-03-04[15:06:55]etimban: user_type S language 2 user_name etimban TTY ote locale
ISO88591/NULL
98-03-04[15:06:55]etimban: Errno 2 (No such file or directory) bdb_errno 0
98-03-04[15:06:55]etimban: Log_mesg:
98-03-04[15:06:55]etimban: No definition in definition file for sopen(F_BRDD:dtffas402,
dtffas402)
98-03-04[15:06:55]etimban: ********** E N D of Log message **********

Technical Manual
8-4
9 Executable programs

Database management
This chapter describes the executable programs used for database management,
which are located in the directory $BSE/bin. Whether you have all the executable
programs described in this chapter depends on your particular installation set.
The programs are sorted by type of database server. References are included for
programs described in other documents.

General
audit_srv6.2
This is the name of the server that handles logging of actions on the database.
See the chapter “Audit management.”
bdbpre6.2
Program to convert database tables to a sequential dump. See the chapter
“Database tools.”
bdbpost6.2
Program to create database tables from a sequential dump. See the chapter
“Database tools.”
bdbreconfig6.2
Program to reconfigure database tables. See the chapter “Database tools.”
dpt6.2
Test program of database actions. Program for internal use, not for the end user.
refint6.2
Function to check the referential integrity of database tables. See the chapter
“Database tools.”
gcommand6.2
Database test program used internally to solve problems. Not intended for the
end user.
qptool6.2
Database test program used internally to solve problems. Not intended for the
end user.

Technical Manual
9-1
Executable programs

Oracle
ora7_inst6.2
The program used to install the Oracle driver.
ora7_deinst6.2
The program used to deinstall the Oracle driver.
ora7_srv6.2
This is the driver for Oracle Version 7.
ora7_maint6.2
The Oracle version of the program used to maintain the database.
ora8_srv6.2
This is the driver for Oracle Version 8.
ora8_maint6.2
The program used to maintain Oracle version 8.

Sybase
syb_install6.2
The program used to install the Sybase driver.
syb_srv6.2
This is the Sybase driver itself.
syb_admin6.2
The shell script for doing administration.
syb_maint6.2
The Sybase version of the program used to maintain the database.

Informix
inf_install6.2
The program used to install the Informix driver.
inf_srv6.2
This is the Informix driver itself.
inf_admin6.2
The shell script for doing administration.
inf_maint6.2
The Informix version of the program used to maintain the database.

Technical Manual
9-2
 Executable programs

DB2
db2_install6.2
The program used to install the DB2 driver.
db2v5_srv6.2
This is the driver executable program that functions with DB2 V5.
db2_admin6.2
The shell script for doing administration with DB2.
db2v5_maint6.2
This is the maintenance executable program, to be used with the DBA module,
which works with DB2 V5.

Logic server (Bshell)

Bshell6.2
This is a virtual processor between the user interface, database drivers,
applications, and operating system. This program makes applications, including
BaanERP Tools, independent of the operating system. The Bshell functions as a
shell between the application and the operating system.
When you start the user interface, it connects with the Bshell on a machine on the
network. This is the usual way to start the Bshell. Note that you can run Bshell
and user interface on different machines. The following settings are required:
n Add a line to your local or remote $BSE/lib/ipc_info file similar to:
bshell s 0 0 s ${BSE}/bin/bshell6.2
n When running on a remote host fill the r user> file as described in the
chapter “Remote databases”.
n You can specify where the application logic server is located in the User Data
Template (ttams1110m000) session on the Authorization Management
SystemÈ Definition of TemplatesÈ Miscellaneous menu. You must then
connect the template to the user using the User Data (ttaad2500m000) session
on the BaanERP ToolsÈ Database Management menu.
The environment variable DS_AS overrules these settings of the user file
(example: DS_AS=host!bshell).

Technical Manual
9-3
Executable programs

SYNOPSIS
Bshell6.2 [options] [program [program arguments]]
Options for Bshell6.2 are described in the following sections:
n Debugging the Bshell during run time
− Baan CPU
− Scheduler
− File I/O
− Miscellaneous
− Message and log extract
n Memory usage
n Miscellaneous
The al_1 function parse_arguments recognizes the following options by default:
-- : end of parameter list
-set var=val : set environment variable var to val
To pass options to the Bshell, enter them in the command line of the BW
Configurator, for example:
-- –set CORE=1 –dbgcpu –dbgfun ttaad2100m000
− pass "-set CORE=1 –dbgcpu –dbgfun" options to Bshell
− and run Appication Configuration (ttaad2100m000) session
Bshell logs all its output to $BSE_TMP/bshell.PID. This log file is removed
when the Bshell exits normally. You can instruct the Bshell to keep the logfile
with the –keeplog option.

Debugging the Bshell during run time

To debug the Bshell while the BaanERP application is running, display the
Option Dialog dialog box, which appears as an icon while you running
BaanERP, and click Debug Bshell. The Runtime debugging of Bshell dialog
box displays. You can also choose these options from the command line.

Baan CPU
Debug functions
In the Option Dialog dialog box, choose Debug Bshell. On the Bshell Debug
Levels tab, select Debug Functions. Or, from the command line, run –dbgfun.
Use debug version of the CPU
In the Option Dialog dialog box, choose Debug Bshell. On the Bshell Debug
Levels tab, select Debug CPU. Or, from the command line, run –dbgcpu.

Technical Manual
9-4
 Executable programs

Show Bshell CPU instructions

In the Option Dialog dialog box, choose Debug Bshell. On the Bshell Debug
Levels tab, select Debug Instructions. Or, from the command line, run –
dbginstr.
Dump 3GL stack traces on function entry
In the Option Dialog dialog box, choose Debug Bshell. On the Bshell Debug
Levels tab, select Dump Stack Traces. Or, from the command line, run –
dbgstack.
Debug get.var and put.var functions
In the Option Dialog dialog box, choose Debug Bshell. On the Bshell Debug
Levels tab, select Getvat/Putvar debug. Or, from the command line, run –
dbggpvar.
Show program flow
In the Option Dialog dialog box, choose Debug Bshell. On the Bshell Debug
Levels tab, select Program Flow. Or, from the command line, run –dbgflow.
Scheduler
Debug scheduler
In the Option Dialog dialog box, choose Debug Bshell. On the Bshell Debug
Levels tab, select Debug Scheduler. Or, from the command line, run –
dbgsched.
Show process actions (for example, activate, sleep, kill)
In the Option Dialog dialog box, choose Debug Bshell. On the Bshell Debug
Levels tab, select Show process actions. Or, from the command line, run –
dbgmulact.
File I/O
Show all files that are currently opened by the Bshell
In the Option Dialog dialog box, choose Debug Bshell. On the Bshell Debug
Levels tab, select Show opened files. Or, from the command line, run –dbgfile.
Debug file access
In the Option Dialog dialog box, choose Debug Bshell. On the Bshell Debug
Levels tab, select Debug file access. Or, from the command line, run –dbgfdev.

Technical Manual
9-5
Executable programs

Miscellaneous
Show object information
In the Option Dialog dialog box, choose Debug Bshell. On the Bshell Debug
Levels tab, select Show object information. Or, from the command line, run –
dbgobj.
Show whether domains, data definitions and objects are loaded from disk or
shared memory
In the Option Dialog dialog box, choose Debug Bshell. On the Bshell Debug
Levels tab, select Show SRDD use. Or, from the command line, run –
dbgsrdduse.
Show various TSS debugging information
In the Option Dialog dialog box, choose Debug Bshell. On the Bshell Debug
Levels tab, select Debug TSS. Or, from the command line, run –dbgtss.
Show data input options (not for fields)
In the Option Dialog dialog box, choose Debug Bshell. On the Bshell Debug
Levels tab, select Debug data.input(). Or, from the command line, run –
dbgdata.
Stop debugger, if possible, when a message is sent to the message window
In the Option Dialog dialog box, choose Debug Bshell. On the Bshell Debug
Levels tab, select Debug Messages. Or, from the command line, run –dbgmesg.
Show loaded resources
In the Option Dialog dialog box, choose Debug Bshell. On the Bshell Debug
Levels tab, select Debug Resources. Or, from the command line, run –dbgres.
Do not remove the logfile after ending the Bshell
In the Option Dialog dialog box, choose Debug Bshell. On the Bshell Debug
Levels tab, select Keep Log File. Or, from the command line, run –keeplog.
Add time stamps to Bshell log output
In the Option Dialog dialog box, choose Debug Bshell. On the Bshell Debug
Levels tab, select Add Time Stamps. Or, from the command line, run –logtime.
Database related
Show the Baan database activities initiated from the Bshell
In the Option Dialog dialog box, choose Debug Bshell. On the Bshell Debug
Levels tab, select Show BDB Actions. Or, from the command line, run –
dbgbdbact.
Show actions related to enums
In the Option Dialog dialog box, choose Debug Bshell. On the Bshell Debug
Levels tab, select Print Enums. Or, from the command line, run –dbgenums.

Technical Manual
9-6
 Executable programs

Show locking errors

In the Option Dialog dialog box, choose Debug Bshell. On the Bshell Debug
Levels tab, select Show locking errors. Or, from the command line, run –
dbglck.
Show database server type
In the Option Dialog dialog box, choose Debug Bshell. On the Bshell Debug
Levels tab, select Show BDB server type. Or, from the command line, run –
dbgsrv.

BDB/SQL Tracing
To trace BaanERP database and SQL actions, display the Option Dialog box and
click Debug Bshell. The Runtime debugging of Bshell dialog box appears. You
can also choose these options from the command line.
BDB Debug Flags
Shows the drivers and parameters currently in use
In the Option Dialog dialog box, choose Debug Bshell. On the BDB/SQL
Tracing tab, select Driver Type. Or, from the command line, run
BDB_DEBUG=01.
Show database actions such as Insert, Update, Delete, Commit, and Abort
In the Option Dialog dialog box, choose Debug Bshell. On the BDB/SQL
Tracing tab, select Database Actions. Or, from the command line, run
BDB_DEBUG=02.
Show information on currently set locks
In the Option Dialog dialog box, choose Debug Bshell. On the BDB/SQL
Tracing tab, select Delayed Locks. Or, from the command line, run
BDB_DEBUG=04.
Show references between tables
In the Option Dialog dialog box, choose Debug Bshell. On the BDB/SQL
Tracing tab, select References. Or, from the command line, run
BDB_DEBUG=010.
Show all tables using native storage format
In the Option Dialog dialog box, choose Debug Bshell. On the BDB/SQL
Tracing tab, select Multibyte Storage. Or, from the command line, run
BDB_DEBUG=040.
Show the permissions and roles allowed for each user
In the Option Dialog dialog box, choose Debug Bshell. On the BDB/SQL
Tracing tab, select Permissions/Roles. Or, from the command line, run
BDB_DEBUG=0100.

Technical Manual
9-7
Executable programs

BDB_DEBUG Value
Shows the current setting that can be used in a –set BDB_DEBUG=<value>
from the command line.
TT SQL TRACE Flags
Show the full text of a query with an ID number
In the Option Dialog dialog box, choose Debug Bshell. On the BDB/SQL
Tracing tab, select Show query with ID. Or, from the command line, run
TT_SQL_TRACE=040.
Show how long a query has been running
In the Option Dialog dialog box, choose Debug Bshell. On the BDB/SQL
Tracing tab, select Query execution times. Or, from the command line, run
TT_SQL_TRACE=0200.
Show main SQL functions such as SQLExec, SQLParse, SQLFetch, and
SQLBind
In the Option Dialog dialog box, choose Debug Bshell. On the BDB/SQL
Tracing tab, select Internal SQL functions. Or, from the command line, run
TT_SQL_TRACE=02000.
Show the best possible design for indexing, joins, and so on
In the Option Dialog dialog box, choose Debug Bshell. On the BDB/SQL
Tracing tab, select Query evaluation plan. Or, from the command line, run
TT_SQL_TRACE=04000.
Show all full table scans
In the Option Dialog dialog box, choose Debug Bshell. On the BDB/SQL
Tracing tab, select Show full table scans. Or, from the command line, run
TT_SQL_TRACE=020000.
Show low-level communication between client and driver
In the Option Dialog dialog box, choose Debug Bshell. On the BDB/SQL
Tracing tab, select Show BDB communication. Or, from the command line, run
TT_SQL_TRACE=040000.
Show current time to level of milliseconds: YYYYMMDDhhmmss.mmm
In the Option Dialog dialog box, choose Debug Bshell. On the BDB/SQL
Tracing tab, select Add time stamps (SQL). Or, from the command line, run
TT_SQL_TRACE=0400000.
TT_SQL_TRACE value
Shows the current setting that can be used in a –set TT_SQL_TRACE=<value>
from the command line.

Technical Manual
9-8
 Executable programs

Memory usage
Show currently running processes in Bshell
In the Option Dialog dialog box, choose Debug Bshell. On the Bshell Debug
Levels tab, select Dump Process List. This option is not available from the
command line.
Show total memory usage
In the Option Dialog dialog box, choose Debug Bshell. On the Bshell Debug
Levels tab, select Total Memory. Or, from the command line, run -dbgmemtot.
Show free memory list
In the Option Dialog dialog box, choose Debug Bshell. On the Bshell Debug
Levels tab, select Free Memory. Or, from the command line, run –
dbgmemfree.
Show used memory list
In the Option Dialog dialog box, choose Debug Bshell. On the Bshell Debug
Levels tab, select Memory Used. Or, from the command line, run –
dbgmemused.
Show memory usage per block
In the Option Dialog dialog box, choose Debug Bshell. On the Bshell Debug
Levels tab, select Memory Block List. Or, from the command line, run –
dbgmemblk.
Show all memory statistics
In the Option Dialog dialog box, choose Debug Bshell. On the Bshell Debug
Levels tab, select All Memory Info. Or, from the command line, run –dbgmem.
Add remark to log file (enter the remark in the field, then click the button)
In the Option Dialog dialog box, choose Debug Bshell. On the Bshell Debug
Levels tab, select Write Remark to Log. This option is not available from the
command line.
Display shared-memory information
In the Option Dialog dialog box, choose Debug Bshell. On the Bshell Debug
Levels tab, select Display Shared Mem. This option is not available from the
command line.

Technical Manual
9-9
Executable programs

Miscellaneous
–vV : version information.
–r : show resources
–deftext : show Bshell tex.ts.
–mdebug : display Bshell messages sent to display server.
–dbgref : show reference paths.
–dbgrefer : show references.
–dbgbdbact : show database actions.
–dbgenums : show loading of enums.
–dbgpty : debug pseudo terminals (pty).
–dbgorb : debug ORB integration (where available) .
–set var=val : set environment variable var to val.
–logfile <file> : log stdout/stderr output in file.
–appendlog : append to logfile (only useful with –logfile option).
–nolog : stdout and stderr go to the controlling terminal.
–delay sec : delay for sec seconds before continuing.

bshcmd6.2
This command can be used to change the log facilities of the Logic Server
(Bshell) or kill one or more Bshell processes. The actions done with this
command are performed while the Bshell is runing.
SYNOPSIS
bshcmd6.2 [options] <bshell_pid>
Possible options:
–v : print version information.
–p : show process list.
–m : show memory usage.
–d <dbglvl> : set DEBUG_LEVEL to (octal) <dbglvl>.

Technical Manual
9-10
 Executable programs

The following <dbglvl> values are available:

0000000001 : show data input actions.

0000000002 : show object information.

0000000004 : show reference paths.

0000000020 : debug functions.

0000000040 : database server information.

0000000100 : show delayed locks.

0000000200 : show process actions (sleep, kill, and so on).

0000000400 : database reference information.

0000001000 : show database actions.

0000002000 : debug file access.

0000004000 : show loaded resources.

0000010000 : show loading of enums.

0000020000 : show Bshell CPU instructions.

0000040000 : use debug version of the CPU.

0000100000 : show whether domains, data definitions and objects are loaded
from disk or shared memory.
0000200000 : debug get.var & put.var functions.

0000400000 : debug scheduler.

0001000000 : debug pseudo terminals.

0002000000 : show opened sequential files.

0004000000 : debug TSS functions.

00020000000: debut ORB integration.

00040000000: show stack traces.

00100000000: debug messages.

00200000000: show program flow.
–k pid : Kill Bshell process ID pid.
–e : kill all Bshell processes.
–M "message" : Send message to Bshell.

Technical Manual
9-11
Executable programs

–W sec : Wait until the previously issued command is executed. After this the
previous command is overwritten.
–w sec : Wait sec seconds for Bshell to execute command.
–u sec : Send SIGUSR1 to Bshell (wakeup). Only to be used in combination with
–w option. Waits sec seconds to see if the command is executed.
–s : Show entire contents of log file (if accessible).
Only to be used in combination with –p and –w option.
-l : Print log file name of Bshell.
Only to be used in combination with –p and –w option.
-T cmdstr : Modify BDB_DEBUG or TT_SQL_TRACE (Bshell) and
DBSLOG, TT_SGL_TRACE, {DBMS}STAT (drivers) tracing variables. cmdstr
can contain multiple commands of the form:
trace variable=value: set variable to value
trace variable+value: add bits to variable
trace variable-value: remove bits from variable
REMARKS All output is stored in the Bshell's log file (default: $BSE_TMP/bshell.pid).
Only one command can be active at a time. New commands overwrite previous
ones.
Check the return value to see if a command has been processed.
The bshell_pid process number is a member of the output of the UNIX ps
command.
EXAMPLES bshcmd6.2 –s –p –w 10 <bshell_pid>
Show the process list, and wait 10 seconds for response. When no actions are
done within 10 seconds, no output is given.
bshcmd6.2 –d 02000 <bshell_pid>
Set DEBUG_LEVEL to 02000.
bshcmd6.2 –M "Hello" –u 10 –w 10 <bshell_pid>
Send a message to a specific Bshell.
bshcmd6.2 –k <pid> <bshell_pid>
Kill the Bshell process <pid>. The <pid> process number can be accessed from
the shell program (ttstpshell) with the ps command.
bshcmd6.2 –T “TT_SQL_TRACE+02000” <bshell_pid>
Set TT_SQL_TRACE to log interface calls.

Technical Manual
9-12
 Executable programs

badmin6.2
Used to perform administration of the Bshell. The usage is as follows: badmin6.2
[-UuVv] [-qo outfile] [-qe errfile] –chkuser <user> | –chkgroup –ostype
Available options are:
–U or –u : print usage.
–V or –v : print release number.
–qo outfile : redirect standard output to file outfile.
–qe errfile : redirect error output to file errfile.
–chkuser user : returns a 0 if user exists, otherwise returns a 1.
–chkgroup group : returns a 0 if group exists, otherwise returns a 1.
–ostype ostype : returns 0 if operating system is ostype, otherwise returns a 1.
ostype can be either NT or UNIX.

Installation
cmt6.2
Script for component merge tool. Used to migrate the sources from user-
customized applications to new BaanERP versions.
install6.2
Script for installing the BaanERP software.
install.help
ASCII file containing Help for the installation procedure.
sh_server6.2
Shell server used to execute system-dependent commands.
bsp.setperm6.2
Script used to assign the correct permissions to all the BaanERP software.
binperm6.2.
Script used to assign the correct permissions to binaries in the $BSE/bin
directory after a new porting system is set up.

Technical Manual
9-13
Executable programs

Development
bic6.2
Program compiler for Baan C compiler.
repgen6.2
Report generator.
std_gen6.2
4GL-program preprocessor.
bic_cstub6.2
Converts a BaanERP library to a C library.
bic_info6.2
Shows information for specified object.
bic_jstub6.2
Generates a Java class from a BaanERP DLL object.

License management
brand6.2
Program to authorize an environment. You must be logged in as the root user to
run this program.
hostid6.2
Program to print the host identification number.
Available options:
–vV : version information.
–o : print this number in an octal format.
–h : print this number in a hexadecimal format.

licd6.2
Daemon process to watch over the license. Because sockets are used, the
Ethernet software must be installed before you can use the license daemon.
The file $BSE/lib/licence6.2 must contain the name of the host on which the
license daemon is running, as follows:
n host name (must be present in /etc/hosts)
n internet address (x.x.x.x)
Note that brand6.2 expects to find the host name in licence6.2.

Technical Manual
9-14
 Executable programs

Available options:
–vV : version information.
–d : debug.
–f : keep licd6.2 running in foreground. This is useful with the –d option.

licmon6.2
Monitor to retrieve information from the license daemon. The options that can be
used when starting the monitor are:
–vV : version information.
–b : show brand information.
–B : show brand information from brand file and shared memory.
–C : clear brand information from shared memory. Use this option when shared
memory is incorrect.
–w : show users.
–W : show other connections.
–k : kill license daemon.
–u : show user count.
–s : show statistics.
–d : print debug information.
–h host : retrieve information from the given host.
–p n : ping the license server n times (test connection).
The commands that can be entered within the license monitor are:
brand information: show the brand information of this host.
help : show the available commands.
ping [n] : ping the license server [n times] (test connection).
quit : quit license monitor.
stats : show server statistics.
users : show user count who [all]: list users [all: list other logons].

Technical Manual
9-15
Executable programs

Resources:
licence_timeout : specifies the number of seconds the license daemon waits for a
reply from another license daemon, or the number of seconds a Bshell waits for a
reply from the license daemon. The default is 30 seconds. Do not lower this
value. With wide area networks (WAN) it is advisable to set this value to 60.
You can change or set the licence_timeout setting in the resource file.

Printer management
filter6.2
Program to translate BaanERP native codes to printer codes used by the
pdaemon6.2 program.
lp6.2
System spooler interface for the pdaemon6.2 program.
pdaemon6.2
Daemon process to print requests from the environment. This program must be
started by the user root.
The options that can be used at calling the printer daemon are:
–v : print version and porting information of the daemon.
–k : kill the running printer daemon.
–f : Start daemon as foreground process (for debugging).
–d n: Print debugging information to stderr. The higher n is, the more
information is output.
–r : Remove the lock file and start daemon (use this in rc.start).
–h : print the above options.
Resources:
maxproc: maximum number of background (filter) jobs. Default: 20.
maxtries: maximum number of minutes to retry opening the database files.
Default: 30 minutes.
sleeptime: sleep time between two polls. Default: 10 seconds.
shell_cmd: shell which runs the lp6.2 system interface. Default: /bin/sh.
strict: The interval in seconds the printer daemon command (up/down) is
checked for all devices. If this resource is not set, the printer daemon command
of the device in question is checked as soon as a print job is found.

Technical Manual
9-16
 Executable programs

Change default values in the file: $BSE/lib/defaults/pdaemon6.2

IMPORTANT It is highly recommended that you group physical printers into logical printers.
The daemon recognizes if a printer is busy and chooses another physical printer
device. It also calculates the size of the several queues for physical printers, so it
chooses the queue with the lowest number of bytes.

Shared memory
shmmanager6.2
Manager program for shared memory. See the chapter “Shared memory
management.”
shmtimer6.2
Daemon process that stores and updates the current time in shared memory. See
the chapter “Shared memory management.”
shmvalues6.2
Program to generate entries for the "shm_param" file. See the chapter “Shared
memory management.”
srdd_init6.2
Program to load the data dictionary for the Bshell from shared memory.

Network
client6.2
Test program of the client part.
fs6.2
Program to handle remote file I/O. With this program you can approach files on
another system. On the other system you must put the program fs6.2 in the
ipc_info file. See the chapter “Database management.”
ipc_boot6.2
Program to start remote processes. See the chapter “Database management.”
server6.2
Test program of the server part.

Technical Manual
9-17
Executable programs

TRITON Super Set

tsscomp6.2
Compiles and checks TSS information file tss6.2. See the chapter “Multibyte
management.”
tsscvt6.2
TSS conversion filter. See the chapter “Multibyte management.”
tssinfo6.2
Gives information about current TSS settings. See the chapter “Multibyte
management.”
uniinfo6.2
The uniinfo6.2 option requires an argument: –u. The usage is as follows:
uniinfo6.2 [-vahsdow] [-l locale] [-u Unicode hex value] [-n Native hex value]
uniinfo6.2 shows all Unicode values from 0x0000 to 0xffff and the
corresponding native character values, UTF-8 values, and TSS character values
in hexadecimal format. By default, uniinfo6.2 shows only values from 0x000 to
0x00ff.
uniinfo6.2 converts multibyte character strings into TSS strings or vice versa.
You can determine the input/output character set by setting the locale with the –l
flag, otherwise it is read from the user file. The default mode is to convert to the
character set of the Bshell. This can be reversed with the –o option. Some
additional options are:
–v : show version information.
–d : show additional debug information.
–w : do not accept any conversion warnings, but exit(1) instead.
–a : show all Unicode from 0x0000 to 0xffff.
–h : show Unicode from 0xE800 to 0xEA00.
–s : do not show when native character is zero.
–o : show the indexed contents of mapping tables.
–l : choose another locale.
–u : shows Native, UTF-8, and TSS value for the Unicode value.
–n : shows Unicode, UTF-8, and TSS value for the native value.
–t : shows Unicode, Native, and UTF-8 value for the TSS value.
–8 : shows Unicode, Native, and TSS value for the UTF-8 value.

Technical Manual
9-18
 Executable programs

–m : use shared memory instead of local application memory.

–p : performance check (you must use –dp option as parameter).
–dp : shows performance measurements of basic Unicode, UTF-8, and TSS
conversion routines.
For more information, see the chapter “Multibyte management.”
unimap6.2
Do not use the –u option with this executable program. The usage is as follows:
unimap6.2 [-vdeow] [mapping table file] [-l locale] [dbcs ranges]
This program generates Unicode mapping tables based on the mapping
information file, which is provided in .txt format in Microsoft Excel. The
mapping file is called locale and is located in $BSE/lib/unicode/. The file
locale.N2U is the Native to Unicode mapping file, and the file locale.U2N is the
Unicode to Native mapping file. The leading byte ranges of the double-byte
character set can be specified as optional parameters.
Additional flags are:
–v : show program version.
–d : show additional debug information.
–e : add 0x8080 to convert JIS0208 values to EUC. This conversion is applied to
Native to Unicode mapping file only.
–f : place no zero value where no mapping value is specified from 0x00 to 0x20
rather than 0x00 to 0xff.
–w : do not accept any conversion warnings, but exit(1) instead.
–o : print mapping tables in ASCII hex format.
–z : place zero value where no mapping value is specified.
See the chapter “Multibyte management.”

Technical Manual
9-19
Executable programs

Time zone utilities

zic6.2
Time zone information compiler.
zdump6.2
Dumps time zone information to standard output.
date6.2
Shows the dates based on BaanERP time zones.

Middleware
orb_srv6.2
Implementation program that supports the server portion of object request broker
(ORB) communication.
bshellorb6.2
Variation of the Bshell used to integrate BaanERP with CORBA programs.

UNIX equivalents
compress6.2
Program to compress data. Equivalent to the UNIX compress command.
diff6.2
Program used to compare two files. Equivalent to the UNIX diff command.
kermit6.2
Serial communications program. Equivalent of the UNIX kermit command.
sort6.2
Sort program. Equivalent of the UNIX sort command. You can use the
environment variable BSE_SORT to specify a path where temporary files are
stored during the sort process.

Technical Manual
9-20
 Executable programs

Miscellaneous
binput6.2
Program to read input. This can be used in shell scripts.
bput6.2
Program to set or print terminal settings.
encrypt6.2
Program to encrypt passwords, which is used to fill the r<user> file. See the
chapter “Database management.”
mirror6.2
Program for demonstrations and courses.
nlsedit6.2
Editor to maintain input and output conversion tables for Native Language
Support. See the chapter “Native language support.”
popup6.2
Program to handle popup screens. This can be used in shell scripts.
sum6.2
Sum program, used to calculate and patch executable programs.

Technical Manual
9-21
Executable programs

Technical Manual
9-22
10 Errors

General
There are three categories of errors:
n Error numbers 1-99 are generated by UNIX or XENIX
n Error numbers 100-899 are database errors
n Error numbers 900-999 are network errors
Error numbers between 34 and 100 are system dependent.

UNIX errors
1 EPERM Not owner
This error indicates an attempt to modify a file that cannot be modified, except
by its owner or super user. This error also appears when ordinary users attempt to
do things allowed only to the super user.
2 ENOENT No such file or directory
This error occurs when a specified file name should exist but does not, or when
one of the directories in a path name does not exist.
3 ESRCH No such process
This error means that no process can be found that corresponds to the one
specified.
4 EINTR Interrupted system call
This error means that an asynchronous signal (such as interrupt or quit), which
the user has chosen to catch, occurred during a system call. If the system resumes
execution after processing the signal, it will appear as if the interrupted system
call returned this error code.
5 EIO I/O error
This error means that there has been some physical I/O error. In some cases, this
error can point to the call following the one to which it actually applies.
6 ENXIO No such device or address
This error means that I/O on a special file refers to a subdevice that either does
not exist, or is beyond the limits of the device. It may also occur when, for
example, a tape drive is not online or no disk pack is loaded on a drive.

Technical Manual
10-1
Errors

7 E2BIG Arg list too long

This error occurs when an argument list longer than 5120 bytes is presented to a
member of the UNIX exec family system calls.
8 ENOEXEC Exec format error
This error means that a request has been made to execute a file which, although it
has the appropriate permissions, does not start with a valid magic number. A
magic number is the first two bytes in a file, used to determine what type of file it
is. See a.out(5).
9 EBADF Bad file number
This error means that either a file descriptor does not refer to an open file, or that
a write request has been made to a file that is opened only for reading that a read
request has been made to a file that is opened only for writing.
10 ECHILD No child processes
This error means that a process that has no child processes waiting, has executed
a wait. A child process is a process spawned by another process (the parent
process).
11 EAGAIN No more processes
This error means that a fork failed, either because the process table of the system
is full or because the user is not allowed to create any more processes.
12 ENOMEM Not enough space
This error means that during an exec or sbrk, a program has asked for more
space than the system is able to supply. This is not a temporary condition. The
maximum space size is a system parameter. The error can also occur when the
arrangement of text, data and stack segments requires too many segmentation
registers, or if there is not enough swap space during a fork.
13 EACCES Permission denied
This error means that an attempt was made to access a file in a manner forbidden
by the protection system.
14 EFAULT Bad address
This error means that the system encountered a hardware fault when it attempted
to use an argument of a system call.
15 ENOTBLK Block device required
This error means that a nonblock file was specified where a block device was
required, for example, in mount.

Technical Manual
10-2
 Errors

16 EBUSY Device busy

This error means that an attempt was made to mount a device that was already
mounted or to dismount a device on which there is an active file (open file,
current directory, mounted-on file, active text segment). The error also occurs
when an attempt is made to enable accounting that is already enabled.
17 EEXIST File exists
This error means that an existing file was mentioned in an inappropriate context,
for example, a link.
18 EXDEV Cross-device link
This error means that an attempt was made to link to a file on another device.
19 ENODEV No such device
This error means that an attempt was made to apply an inappropriate system call
to a device, for example, a read on a write-only device.
20 ENOTDIR Not a directory
This error means that a nondirectory was specified where a directory is required,
for example in a path prefix or as an argument to chdir(S).
21 EISDIR Is a directory
This error means that an attempt was made to write to a directory.
22 EINVAL Invalid argument
This error means that an invalid argument was used, for example, dismounting a
nonmounted device; mentioning an undefined signal in signal or kill; reading or
writing a file for which lseek has generated a negative pointer. This error is also
used by the math functions described in the (S) entries of the UNIX manual.
23 ENFILE File table overflow
This error means that the systems table of open files is full, and temporarily no
more open commands can be accepted.
24 EMFILE Too many open files
This error means that too many file descriptors are open. No process can have
more than 20 file descriptors open at a time.
25 ENOTTY Not a typewriter
This error means that the selected device does not have the properties of a
terminal.
26 ETXTBSY Text file busy
This error means that an attempt was made to execute a pure-procedure program
that is currently open for writing or reading. It can also mean that an attempt was
made to open for writing a pure-procedure program that is being executed.

Technical Manual
10-3
Errors

27 EFBIG File too large

This error means that the size of a file exceeded the maximum (1,082,201,088
bytes) or ULIMIT; see ulimit(S).
28 ENOSPC No space left on device
This error means that when an ordinary file is written, there is no free space left
on the device.
29 ESPIPE Illegal seek
This error means that an lseek was issued to a pipe.
30 EROFS Read-only file system
This error means that an attempt was made to modify a file or directory on a
read-only device.
31 EMLINK Too many links
This error means that an attempt was made to link more than the maximum
number of links to a file. The maximum number of links to one file is 1000.
32 EPIPE Broken pipe
This error means that a write was made on a pipe for which there is no process to
read the data. This condition normally generates a signal. The error is returned if
the signal is ignored.
33 EDOM Math arg out of domain of func
This error means that the argument of a function in the math package is out of the
domain of the function.
34 ERANGE Math result not representable
This error means that the value of a function in the math package cannot be
represented within machine precision.

Database errors
100 EDUPL
This error means that a duplicate value exists.
101 ENOTOPEN
This error means that the table is not open.
102 EBADARG
This error means that an illegal argument has been specified.
103 EBADKEY
This error means that an illegal key description has been specified. Use the
bdbpre and bdbpost tools.

Technical Manual
10-4
 Errors

106 ENOTEXCL
This error means that the table is not exclusively locked action. You can either
wait until the lock on table is released or you can remove the lock yourself.
107 ELOCKED
This error means that the record you are trying to retrieve is locked. You can
either wait until the lock is released or remove the lock yourself.
108 EKEXISTS
This error means that the key already exists.
109 EPRIMKEY
This error means that you are trying to perform an illegal action on a primary
key. Refer to the log file for more information.
110 EENDFILE
This error means that the end of the file has been reached.
111 ENOREC
This error means that no record was found that matches the query criteria.
112 ENOCURR
This error means that there is no current record.
113 EFLOCKED
This error means that the table is locked. You can either wait until the lock is
released or you can remove the lock.
114 EFNAME File name too long
This error means that you are using a file name that is too long. Refer to your
system requirements to check the maximum length for a file name.
116 EBADMEM
This error means that the system cannot allocate memory because it is out of
memory. Try restarting the Bshell as a possible solution.
117 EBADCOLL
This error means that there is a problem with the collating, or sorting order.
123 ENOSHMEM
This error means that no shared memory is initialized. You can use for example
the rc.start script to initialize shared memory.
129 EUSER
This error means that too many sessions have been started.
136 ENOSPACE
This error means that there is no space in shared memory. You can either
initialize the shared memory, or change the shared memory parameters.

Technical Manual
10-5
Errors

140 ETRANSON
This error means that this operation is invalid when the transaction is on.
141 ETRANSOFF
This error means that this operation is invalid when the transaction is off.
142 EADMON
This error means that some administration process is running.
146 ENOSNAPSHOT
This error means that the system cannot take a snapshot of the database, probably
because it is locked by another user. You can either wait until the lock is released
or you can remove the lock.
148 EOFRANGE
This error means that there has been an error in the data type range check.
201 EROWCHANGED
This error means that the record was changed after a delayed lock.
202 EDBLOCKED
This error means that the database is locked. You can either wait until the lock is
released or you can remove the lock yourself.
203 ETRANSACTIONON
This error means that this action is not allowed within a transaction.
204 EISREADONLY
This error means that this transaction is read only.
205 ENOTINRANGE
This error means that the field value is out of range and does not agree with the
domain definition.
206 ENOTLOCKED
This error means that the record is not locked.
207 EAUDIT
This error means that there is an error of the audit trailer.
208 EPERMISSION
This error means that the action you have just attempted is not allowed at this
time.
209 EMIRROR
This error means that there is an error in the mirroring of the database. The tables
are inconsistent. You can use bdbpre and bdbpost to copy the tables correctly.

Technical Manual
10-6
 Errors

210 EMLOCKED
This error means either that the record is locked in the mirrored database, or that
the tables are inconsistent, or that the mirroring definition in tabledef6.2 is not
compatible.
213 ETRANSACTIONOPEN
This error means that the transaction is started, but not updated. This is an
internal Bshell error.
214 EUNALLOWEDCOMPNR
This error means that an operation for mapping company numbers is not allowed.
If the logical company is not equal to the physical company, you are not allowed
to do a drop/clear table operation.
215 EDBDILLEGAL
This error indicates an illegal state that should never occur.
251 EAUDSETUP
This error means that the audit server setup is not correct. See the log.audit file
for more information.
252 EAUDCORRUPT
This error means that an audit file is corrupt. See the log.audit file for more
information.
253 EAUDLOCKED
This error means that the audit file is locked by another user. See the log.audit
file for more information.
254 EAUDABORT
This error means that a commit transaction has failed in the audit server action
See the log.audit file for more information.
301 ESQLQUERY
This is a general SQL error code. This error occurs when there is a problem with
the SQL syntax.
302 ESQLSYNTAX
This error means that the SQL syntax is not correct.
303 ESQLREFER
This error means that a reference in the query cannot be found.
304 ESQLUNDEFINED
This error occurs when something went wrong but no error code can be set.

Technical Manual
10-7
Errors

305 ESQLWRONGROW
This error means that a wrong record was returned. This probably means either
that the table index is corrupt or that the RDBMS has a different sorting order
than the BaanERP software.
501 EMEMORY
This is an internal memory error.
502 EBDBON
This error means that the user is already logged on.
503 EBADADRS
This error means that an illegal address has been used.
504 EBADFLD
This error means that a column is undefined.
505 ENOSERVER
This error means either that no server is specified in tabledef6.1, or that the
server cannot be started. See the log file for more information.
506 ENOTABLE
This error means that the table does not exist.
507 ETABLEEXIST
This error means that the table you are trying to create already exists.
508 EBDBNOTON
This error means that you are not logged on to a database.
509 EBADCURSOR
This error means that you have a bad memory cursor or that a bad table pointer
has been specified.
510 EDBNOTON
This error means that the database is not on. Start the database to correct the
problem.
511 EWRONGVERSION
This error means that the version of client is not the same as the version of the
server.
512 EDDCORRUPT
This error means that the data dictionary is corrupt. Use bdbpre and bdbpost
tools to repair it.
513 ENODD
This error means that the data dictionary was not found.

Technical Manual
10-8
 Errors

514 ESECURITY (Oracle)

This is a security error. It probably means you do not have the correct user or
group permission.
515 ELICENSEERROR
This is a license error. It probably indicates an unpatched binary.
516 EUPDSEGM
This error occurs during the making or filling of rollback segments. It probably
means that the disk is full.
517 EDELAYED
This is a general error indicating delayed locking.
518 ENOSESSION
This error means that an invalid session code has been specified.
519 ENOCOMPNR
This error means that either no company number an illegal company number has
been specified. A valid company number is a number between 0 and 999.
520 EBUFUPD
This error occurs when flushing of buffered updates fails. The flushing can fail
due to a lock or referential integrity constraint.
521 ENOSHM
This error means that shared memory has not been loaded. See the chapter
“Shared memory” for more information about starting shared memory.
522 EBDBDBCONNECTIONLOST
This error means that the connection between the driver and database has been
lost.
600 EREFERENCE
This is a general reference error. For more information, see the log file.
601 EREFLOCKED
This error means that the reference table is locked. You can either wait until the
lock is released or remove the lock yourself.
602 EUNDEFREF
This error means that the reference is not defined. It is probably a problem in the
run time data dictionary. For more information, see the log file.
604 EREFUPDATE
This error means that a reference could not be updated.

Technical Manual
10-9
Errors

605 EREFEXISTS
This error means that the record cannot be deleted while the reference exists. See
the log file for more information.
606 EREFNOTEXISTS
This error means that the reference does not exist.
607 ENOREFTBL
This error means that the reference table was not found. The data dictionary
might not be correct. See the log file for more information.
608 ENOREFCNT
This error means that no reference counter fields are present.
609 EUPDREFCNT
This error can occur when the reference counter is updated.
700 ESETLOCALE
This error can occur when you are setting the locale. See the log file for more
information.
If an error code greater than 1000 is displayed, the error can be retrieved as
follows, depending on the database driver used:
Informix/Oracle: error – 1000 gives DB error
Example UNIX
Error no: 11400
Error: 11400 – 1000 = 10400
UNIX error = 0 (last two digits)
Example Oracle:
Error no: 1979
Error: 1979 – 1000 = 979 Not a GROUP BY

NOTE When a fatal error occurs, more information is stored in the log files in the
directory $BSE/log. For example, if bdbpost causes an error, it is reported in the
file log.bdbpost.

Technical Manual
10-10
11 Shared memory management

General
Shared memory is a part of the internal memory intended for common usage. All
users can read from and write to it. In BaanERP Tools the shared memory
contains part of the data dictionary.
A prerequisite for shared memory is that it must be supported by the hardware
and there must be sufficient internal memory available.
This chapter describes the way shared memory can be used for BaanERP Tools.

Using the shared memory manager

The shared memory manager is located in the $BSE/bin directory and can be
started as follows:
shmmanager6.2 [-iksavr]
The options are:
–i: installation of shared memory.
–k: deletion of created shared-memory blocks. After shared memory is deleted
you must reinstall it. You can only use this option if all processes that use shared
memory have finished. If not, the system displays a message.
–s: display of technical information on the contents of shared memory. This
includes:
− common Bshell pointers
− description of segment table
–a: display of allocation data, such as addresses, segments, size, and so on.
–v: display of machine and porting data.

Technical Manual
11-1
Shared memory management

–r: resetting of shared memory. Removes all data from shared memory except
the first segment. You do not have to reinstall shared memory afterwards. The
result of this option is the same as when using –k option followed by –i option.
You can only use this option if all processes that use shared memory have
finished. If not, the system displays a message:
# of attaches = <number>
Cannot remove ..# of attaches not equal to zero
The –a, –i and –v options are discussed in the section “Installation of shared
memory.”

Installation of shared memory

The Bshell requires at least 4 MB of internal memory to function properly. After
the kernel selects a virtual address, the kernel does not always leave enough
memory for the Bshell. You must therefore specify a virtual address to determine
where shared memory is allocated in internal memory.
You can determine an adequate address using the shared memory manager,
shmmanager6.2. This section includes the instructions for the installation of
shared memory. Following the installation is a description of error messages that
can come up during installation. The section “Kernel parameters” contains the
minimum values of the kernel parameters.

Create an entry in the file shm_param file

The shmmanager6.2 program uses the shm_param parameter file, which
contains parameters to install shared memory for a specific machine. This file is
stored in directory ${BSE}/lib.
Create an empty entry in this file for the machine in question. It must read as
follows:
Machine_id/OS release:;
{
}
NOTE This entry must be preceded by a default entry. See also the section “Example
parameter file shm_param.”
The OS release is only used if the file contains more than one entry for the same
machine ID. You can display the machine ID and the OS release using the
command shmmanager6.2 –v.

Technical Manual
11-2
 Shared memory management

Determine shared-memory parameters

You can determine these parameters by using one of the following commands:
n shmvalues6.2
n shmmanager6.2 –a

Using shmvalues6.2
The program shmvalues6.2 finds the boundaries of shared memory by itself and
generates a default entry, which you can include in the shm_param parameter
file. For example:
default:;
{
}
Machine_id:;
{
}
Machine_id/OS release:;
{
SHM_START = a40000
SHM_STEP = 80000
SHM_BUFSIZE = 512
SHM_MAXMEM = 10
}
SHM_START is the start address of the first shared-memory segment.
SHM_STEP is the difference between the start addresses of two consecutive
segments.
SHM_BUFSIZE is the size of a segment in KB.
SHM_MAXMEM is the maximum number of segments.
The syntax for shmvalues6.2 is:
shmvalues6.2 [-dvV] [-m Seg_size_in_MB] [-s Seg_size_in_KB]
Possible options are as follows:
–d : the –d option provides you with additional information through error
messages when the boundaries of the operating system are exceeded.
–m : the program shmvalues6.2 first tries to allocate 4 MB of memory to make
sure enough memory is left after shared memory has been installed. You can use
the –m option to increase or decrease the amount of allocated memory. You
specify it in megabytes, that is, –m 3 = 3 MB.

Technical Manual
11-3
Shared memory management

–vV : the –v or –V option displays machine and porting data.

–s : Use this option to increase or decrease the SHM_BUFSIZE value (default
512 KB). This value must be less than or equal to the kernel parameter
SHMMAX.
NOTE You can create a parameter file by entering:
shmvalues6.2>shm_param
The entry is written to standard output (stdout) and all error messages are written
to stderr.
CAUTIONS SHM_BUFSIZE is below recommendations; try tuning the kernel.
SHM_MAXMEM is below recommendations; try tuning the kernel.

FATAL ERRORS Can’t allocate <size> MB of memory: <system error message>; Please try tuning the
kernel or lower the value
Couldn’t create any ID of any size: <system error message>
Error while removing shmid <shmid>: <system error message>
Failed to attach any segment: <system error message>
Error while detaching shmid <shmid>: <system error message>
The shmvalues6.2 program returns a positive value in the event of a fatal error.

Using shmmanager6.2
If you cannot determine the shared-memory parameters by using shmvalues6.2,
you can use the shared memory manager. This takes longer, but it yields more
information on the addresses of the segments and on the instances where errors
can occur.
Execute the shmmanager6.2 with the –a option.
You are prompted for the number of kilobytes to be allocated and the number of
segments to be created. For the Bshell, specify 4096 KB and 10 segments,
respectively.

Technical Manual
11-4
 Shared memory management

For example:
# shmmanager6.2 –a
No. of kbytes to be malloc: 4096
No. of shm segments to be created: 10
addr[0]: 0xa40000 id[0]: 40 < SHM_START = a40000
addr[1]: 0xac0000 id[1]: 41
addr[2]: 0xb40000 id[2]: 42
addr[3]: 0xbc0000 id[3]: 43
addr[4]: 0xc40000 id[4]: 44
addr[5]: 0xcc0000 id[5]: 45
addr[6]: 0xd40000 id[6]: 46
addr[7]: 0xdc0000 id[7]: 47
addr[8]: 0xe40000 id[8]: 48
addr[9]: 0xec0000 id[9]: 49 < SHM_MAXMEM = 10
step 0x80000 < SHM_STEP = 80000
step 0x80000
step 0x80000
step 0x80000
step 0x80000
step 0x80000
step 0x80000
step 0x80000
step 0x80000
step 0x80000
ret 0

Fill the entry in the shm_param parameter file

Enter the variables as determined in the shm_param parameter file. The
following variables are mandatory:
n The start address of the first shared-memory segment: SHM_START
n The number of shared-memory segments: SHM_MAXMEM
n The difference between the start addresses of the segments: SHM_STEP
You can retrieve the values of these variables from the information in the
example on the preceding page. For this example, fill the entry in the parameter
file as follows:
Machine_id/OS release:;
{
SHM_START = a40000
SHM_MAXMEM = 10
SHM_STEP = 80000
}

Technical Manual
11-5
Shared memory management

Installation of shared memory

Use the –i option to install shared memory. For example:
# shmmanager6.2 –i
BUFSZ 524288, MAXATTCH 10, START 0xa40000, STEP 0x80000
Start /usr/bse/bin/shmtimer6.2:
Shmtimer started: pid = 22743, time = 790343035
(Tue Jan 17 12:43:55 1995)
Starting successful

NOTE During installation the BSE shell variable must be the same as the one for the
Bshell users. You can avoid installing shared memory manually each time by
including the installation in your system’s start-up procedure. If the BSE variable
is not /usr/bse, it must be set during the start-up procedure and shmmanager6.2
must be started with the –I option.

Starting shmtimer6.2
The UNIX time() function is a system call often called in database servers. On
multiprocessor systems, use of the time() function can be quite heavy. Even
though a process can run successively on different processors, this must be
invisible to the process. The process requires that each processor returns the
same time, using time(). Therefore, the processors have to synchronize their
internal clocks every time the time() function is called.
The shmtimer6.2 program can prevent this overhead. shmtimer6.2 is a daemon
process that writes the current time in shared memory. The time is updated every
second. This reduces the number of calls of the time() function to a maximum of
one per second.
When shared memory (shmmanager6.2 –i) is initialized, the shmtimer6.2
program starts. When shared memory is removed (shmmanager6.2 –k), the
shmtimer6.2 stops first. The following options can be used for shmtimer6.2:
–i : Initialize shmtimer6.2
A shmtimer6.2 program is started, and runs in the background.
–k : Kill shmtimer6.2
The running shmtimer6.2 program is killed. The allocated shared memory is
cleared, but not removed.
–s : Show status as stored in shared memory:
− The current time
− The process ID of the running timer
–u : Show information about shmtimer6.2

Technical Manual
11-6
 Shared memory management

–v : show version and porting data of shmtimer6.2

When shmtimer6.2 stops, the shared memory is cleared, as described for the –k
option. From then on, the system calls the UNIX time() function instead of
reading the time from shared memory.
If shmtimer6.2 is killed, shmtimer6.2 cannot clear its shared memory. Because
shmtimer6.2 is killed, the time in shared memory is never updated, but the time
is still read from shared memory. In that case, shmtimer6.2 –s gives a warning.
A new shmtimer6.2 can be started with shmtimer6.2 –i.

Error messages during installation

The error messages are described in the order in which they can occur.

No more space
# shmmanager6.2 –a
No. of kbytes to be malloc: 4096
no more space
abort – core dumped
This error happens because some computer systems put a limit on the amount of
virtual memory (MAXUMEM) used by a process. Hence the shared memory
manager can not be capable of allocating 4096 K.
Solution: Adjust the kernel parameters. See the section “Kernel requirements.”

Cannot create shared-memory segments of 512K

# shmmanager6.2 –a
No. of kbytes to be malloc: 4096
No. of shm segments to be created: 10
shmget errno 22
etc......
Errno 22 (EINVAL): The kernel cannot create any shared-memory segments of 512 K
(SHMMAX).
Solution: Adjust the kernel parameters. See the section “Kernel requirements.”

Technical Manual
11-7
Shared memory management

Not enough virtual memory

# shmmanager6.2 –a
No. of kilobytes to be malloc: 4096
No. of shm segments to be created: 10
shmat errno XX id 46
addr[0]: 0xa40000 id[0]: 40
addr[1]: 0xac0000 id[1]: 41
addr[2]: 0xb40000 id[2]: 42
addr[3]: 0xbc0000 id[3]: 43
addr[4]: 0xc40000 id[4]: 44
addr[5]: 0xcc0000 id[5]: 45
addr[6]: 0xffffffff id[6]: 46 (first invalid segment)
addr[7]: 0 id[7]: 6747 (junk)
addr[8]: 0 id[8]: 9644 (junk)
addr[9]: 0 id[9]: 20492 (junk)
step 0x80000
step 0x80000
step 0x80000
step 0x80000
step 0x80000
step 0x80000
step 0xff53ffff
step 0x1
step 0
ret 0
Errno XX can be:
12 (ENOMEM): The user process does not have sufficient virtual memory (MAXUMEM)
24 (EMFILE): (In this example) the kernel cannot create more than 6 shared-memory
segments per
process (SHMSEG).
Solution: Adjust the kernel parameters. See the section “Kernel requirements.”
If the kernel cannot allocate 10 segments, as in the previous example, you need
to fill the entry in the shm_param file as follows:
Machine_id/OS release:;
{
SHM_START = a40000
SHM_MAXMEM = 6
SHM_STEP = 80000
}

Technical Manual
11-8
 Shared memory management

Syntax srdd_tab
The srdd_tab6.2 file can be divided into several sections. Each section is
indicated by the package combination for which components have to be loaded
into shared memory. For example:
package=31Sa
This parameter means that the components below this line belong to package
combination 31Sa.
After this line first the domains are specified. For example:
dti.pd
dtt.pd
The following data is the table definitions. For example:
dttaad000
dttaad001
After that the objects have to be placed. For example:
ottstpconv
ottstpdisplay
So the syntax must be:
package={pc}
for domains: <d{p}.pd>
for data definitions: <d{p}{t}>
for objects and reports: <o{p}{m}{o}>
The abbreviations are:
{pc} package combination
{p} package
{m} module
{t} table
{o} object (program/report)
You can also read parts of a data dictionary on a remote system into shared
memory. To do this, write the host name of the remote system at the beginning of
a line. For example:
${BSE}/standard6.2/ddbaan/dti/dti.pd
ibm1!/usr3/bse/standard6.2/ddbaan/dtd/dtd.pd
${BSE}/standard6.2/ddbaan/dttadv/dttadv100
${BSE}/standard6.2/ddbaan/dtppdm/dtppdm100
/usr2/bse/standard6.2/ttB40_a_CP/ottadv/oadv1100
If the file srdd_tab6.2 contains a redirection to a remote system, a remote user
file must be available for the user who executes srdd_init6.2.

Technical Manual
11-9
Shared memory management

Filling shared memory

You can load parts of the data dictionary into shared memory: data definitions,
objects, reports, and domains. In the file $BSE/lib/srdd_tab6.2, specify which
parts to load into shared memory. You can fill the file srdd_tab6.2 using
BaanERP Tools:
Shared Memory Data (ttaad4150m000) session
To load program and report objects.
%SEttaad1120m000 session
To load table definitions and domains.
The parts of the data dictionary loaded into shared memory have an entry
consisting of the inode number, date, time, and name. If the entry of a part in
shared memory does not match the entry for that part on hard disk, the parts on
hard disk are loaded locally, but not into shared memory. The old part remains in
shared memory until you execute shmmanager –k to delete shared memory
blocks. Then you reinstall shared memory using shmmanager –i.
Use the srdd_init6.2 program to load the parts of the data dictionary into shared
memory. The available options are:
–l : print what is done by the program, including error messages, if any. Use only
in combination with the –i option.
–i : initialize the parts that must be loaded in shared memory.
–p : print a list of loaded components.
–v : version information of the program.
Use the shmmanager6.2 –s command or the UNIX command ipcs –m to see
which segments are used and how much space is left to load other parts of the
data dictionary.
Error messages and other messages are written to stderr. You can write them to a
file by entering the command:
srdd_init6.2 –i 2> file name
Directory ${BSE}/etc contains the file rc.start. This program is executed at
system startup. Shared memory is installed, filled, and initialized for a BSE
environment specified in this file. The programs shmvalues6.2,
shmmanager6.2, and srdd_init6.2 are used.

Technical Manual
11-10
 Shared memory management

Kernel requirements
General description of adjusting UNIX kernels

Static kernels
The kernel of a UNIX system is a program that runs from booting to shutting
down. The kernel performs all actions in which user programs access resources.
Internally, the kernel keeps lists of the actions that are executed simultaneously.
Because those lists have limited capacity, a very large number of simultaneous
actions can cause an overflow.
When there is an overflow, the kernel cannot process any new actions.
Consequently, when another action is added to the list the system displays an
error message. For example, if the number of files opened at the same time is
greater than permitted by the kernel, the system responds with error message 23
(ENFILE File table overflow).
To avoid this error message, the user can create a new kernel that allows a larger
number of files to be opened simultaneously. The kernel is stored on the system
in different forms.
First, it includes sources, libraries, an include file and, most importantly, a file
containing kernel parameters.
Second, the kernel is present as an executable and can be generated from the
previously-mentioned files.
Third, a kernel runs in the internal memory.
Adjusting the kernel requires a three-step procedure:
1 Adjust the file with kernel parameters.
2 Generate the kernel as an executable from the parameter file.
3 Reboot the computer, allowing the new kernel to be loaded into internal
memory and started.

Dynamic kernels
With a static kernel, boundaries are set through a kernel parameter. With a
dynamic kernel, the user can adjust boundaries during run time. The kernel of an
IBM system is fully dynamic (except for a few parameters), so the user does not
have to reboot the computer to activate changes. You can use a command to
modify the kernel of an IBM while the system is still running. Furthermore, the
kernel is automatically adjusted if the required number of resources is greater
than permitted by the kernel.

Technical Manual
11-11
Shared memory management

The kernels of Digital and Sun systems are semidynamic. When the system is
booted, some kernel parameters are read from a file and the kernel is adjusted
accordingly. The user can adjust other parameters while the kernel is running.

Kernel parameters
The recommended values for the parameters presuppose the following
conditions:
n A standard BaanERP installation
n One active session per Bshell
You can use the sar command while working in BaanERP to keep track of
parameter usage and adjust the parameters where necessary. This is of particular
interest for BaanERP customization. When using other databases, take into
account the kernel tuning references from the corresponding documentation.
NOTE When using raw devices, you must set the parameters relating to the cache such
as NBUF or BUFPAGES to low values, according to the database-specific
documentation, not to the relatively high values suggested in this document.
The following descriptions for parameters refer to the number of users. If there is
more than one active session per Bshell, instead of number of users, read number
of active sessions.

Processes
NPROC
The NPROC parameter specifies the maximum number of processes that can be
run simultaneously in the system.
NPROC = 64 * users & NPROC = 1.1 * MAXUP
Adjust: error 11
MAXUP
The MAXUP parameter specifies the maximum number of processes that can be
run per logon code. This restriction does not apply to the superuser. If several
users log on using the same logon code, those users must share this number of
processes.
MAXUP=64
Adjust: error 11, message ’fork failed: too many processes’

Technical Manual
11-12
 Shared memory management

REGION, NREGION
The REGION parameter specifies the maximum number of program regions that
can be active at the same time. Each UNIX process has at least three regions:
text, data, and stack. The text part is shared if a program runs more than once.
Shared memory also uses the region resources.
REGION=3* * NPROC

Files
NFILE
The NFILE parameter specifies the maximum number of files that can be
opened simultaneously on the system.
NFILE = 512 * users
Adjust: Message “ file table overflow”
The command sar –v 1 1 shows the number of open files and the maximum
number of files to be opened, under the header “file-sz.”
NOFILES, SFNOLIM(SVR4), MAXFILES (HP)
The NOFILE parameter specifies the maximum number of files that can be
opened per process.
NOFILES >+ 256
Adjust: error 24, Too many connected sessions (INGRES).
On an SVR4 system the user can use not only the kernel parameter sfnolim, but
also a command to change the maximum number of open files per process. That
command is “limit – 256”, including the quotation marks. This new value only
applies in this shell.

Technical Manual
11-13
Shared memory management

Buffers
NBUF
Under SVR3 and SVR4, the NBUF parameter has a different meaning.
SVR3
The NBUF parameter specifies the number of system buffers available for
block I/O.
NBUF = 1/3 of total internal memory (SVR3)
NBUF = 0 (HP)
SVR4
In UNIX SVR4, buffers contain only indexes, superblocks, and file header
information. They do not contain file data. The actual file data is stored in
pages of virtual memory. The amount of virtual memory available for this
purpose is determined by the SEGMAPSZ parameter.
The NBUF parameter specifies the number of system buffer headers available
for block I/O.
NBUF = default, for example, 100
Adjust: Performance problems (sar –b)
NHBUF
The NHBUF parameter specifies the number of hash table entries that can be
allocated in the system.
NHBUF = 1/5 * NBUF
Adjust: Performance problems
BUFPAGES
The BUFPAGES parameter specifies the number of pages in the file’s system
buffer cache.
BUFPAGES = 1/3 of total internal memory) /4096, (HP)
Adjust: Performance problems
SEGMAPSZ
The SEGMAPSZ parameter specifies the size of the bitmap for file I/O. This
map determines the number of 4K pages and therefore the amount of memory
available for file I/O. If the value is 0, the kernel calculates the value as a
function of the amount of physical memory.
SEGMAPSZ = 0 or 1/x of total internal memory.
Adjust: Performance problems

Technical Manual
11-14
 Shared memory management

BUFHWM
The BUFHWM parameter specifies the maximum amount of memory, in
kilobytes, that can be used by block I/O buffers. If the value of BUFHWM is 0
(the default), the kernel sets the value of the I/O buffers to 25% of available
memory.

Shared memory
SHMALL
The SHMALL parameter specifies the maximum amount of shared-memory text
segments that can be created.
SHMALL = SHMNI
Adjust: error 24
SHMMAX
The SHMMAX parameter specifies the maximum size, in bytes, of a shared-
memory segment that can be created. Several shared-memory segments of this
size can be created in an environment per process.
SHMAX >= SHM_BUFSIZE (in $BSE/lib/shm_param)
SHMMIN
The SHMMIN parameter specifies the maximum size of a shared-memory
segment in bytes.
SHMMNI = SHMSEG * ENVIRONMENTS
Test with: ipcs –m
SHMSEG
The SHMSEG parameter specifies the maximum number of shared-memory
segments that can be used by a process.
SHMSEG = 30
Adjust: error 24

Semaphores
SEMMAP
The SEMMAP specifies the number of entries in the control map used to
manage semaphores. This map is used to keep track of free areas in the system
pool of semaphores.
SEMMAP = SEMNI +2
Adjust: Message “Mfree map overflow”

Technical Manual
11-15
Shared memory management

SEMMNI
The SEMMNI parameter specifies the maximum number of semaphore
identifiers in the kernel. This is the number of unique semaphore sets that can be
active at any given time.
SEMMNI = 32 * users
Adjust: error 28
SEMMNS
The SEMMNS parameter specifies the maximum number of semaphores ipc –sb
permitted in the system.
SEMMNS = SEMMNI
Test with: “ipc –sb”
SEMUME
The SEMUME parameter specifies the maximum number of undo entries per
undo structure. Each undo entry represents a semaphore that has been modified
with the undo option specified in the semop(2) system call.
SEMUME = 10
Adjust: error 22

Message
MSGMAP
The MSGMAP parameter specifies the number of entries in the control map, for
example 100, used to manage message segments. Each entry in this map
represents a free area in the message buffer area.
MSGMAP = default
Adjust: Message “mfree map overflow”
MSGMAX
The MSGMAX parameter specifies the maximum size of a message in bytes. If
the protocol m (message queues) is used in the file $BSSE/lib/ipc_info, the value
must be set to 4096.
MSGMAX = 4096 & MSGMAX <= MSGMNB
MSGMNB
The MSGMNB parameter specifies the maximum length, in bytes, of a message
queue.
MSGMNG = 4096
Check with: ipcs –qo

Technical Manual
11-16
 Shared memory management

MSGMNI
The MSGMNI parameter specifies the maximum number of message queues
that can be used on the system.
MSGMNI=32 *users
Adjust: error 28
MSGSEG
The MSGSEG parameter specifies the number of message segments permitted
on the system, for example 1024. Each message on a message queue consists of
one or more message segments. The size of each segment is specified by the
MSGSSZ parameter.
MSGSEG = default
MSGSSZ
The MSGSSZ parameter specifies the size, in bytes, of a message segment, for
example, 8. Each message consists of contiguous set of message segments large
enough to hold the text of the message.
MSGSSZ = default
MSGTQL
The MSGTQL parameter specifies the maximum number of message headers, in
other words, the maximum number of open messages.
MSGTQL = 32 * users

General kernel parameters

ULIMIT, HFSZLIM, SFSZLIM

SVR3
The ULIMIT parameter specifies the maximum size of a single file on disk.
SCO UNIX and HP use blocks of 512K.
SVR4
The kernel parameter ULIMIT is not used in SVR4. SVR4 uses the parameter
HFSZLIM for the hard limit and SFSZLIM for the soft limit. These parameters
are in bytes.
SVR3 and SVR4 both contain the built-in command ULIMIT through which the
kernel parameter value can be reduced.

Technical Manual
11-17
Shared memory management

NINODE
The NINODE parameter specifies the maximum number of inodes that can be
used in the SV file system. This includes open files, used pipes, and the current
directory. The UFSNINODE parameter specifies the number of inodes for the
UFS file systems.

Example parameter file shm_param

# The default entry (please do not modify)
default:;
{
SHM_BUFSIZE = 512
SHM_MAXMEM = 30
SHM_STEP = 80000
}
SCO_UNIX_386/3.2.2:;
{
SHM_START = 80400000
SHM_MAXMEM = 10
SHM_STEP = 400000
}

Technical Manual
11-18
12 Remote databases

General
The client/server architecture as supported by BaanERP enables the user to work
with several database types. These databases can be distributed over one or
multiple systems. A remote database configuration is one where an application
server (bshell) and the database are not on the same system.
The possible reasons for setting up a remote database configuration are:
n System performance
If all users start the application on the same system, the CPU is primarily
occupied by the Bshell and the display server. You can divide the load by
letting the users work on local systems and keeping the database on a remote
system.
n Lack of disk space on local system
If all tables cannot be stored on the same system due to a lack of disk space,
some of the tables can be installed on a second system.
n Distributed applications
For practical reasons you can decide to install the application tables on
different systems per application.
n Mirrored databases
For security reasons database actions can be carried out on two systems
simultaneously using database mirroring.
This chapter describes how to set up a remote database configuration and also
explains how to use it.

Technical Manual
12-1
Remote databases

Settings for remote database configurations

To work with a remote database configuration, you must predefine a number of
settings. You must know on which systems the tables are located, and also which
types of databases are going to be used. Also, if the communication structure has
not changed, the porting sets of the binaries on the local and remote systems can
be different, otherwise they must be equal. It is recommended to have the same
porting sets. This chapter explains how you can configure a remote database
system.
The tables can be located in various ways in a remote database configuration:
n All tables are located on one or more systems
n A part of the tables are located on the local and another part of the tables are
located on a remote system

Figure 8, All tables on remote system

In figure 8, the user interface and the application logic (Bshell) are both installed
on the local host. From version 6.1 forward, they can be installed on different
systems.
Assuming that all tables are installed on a remote system and the user interface
and application logic (Bshell) are installed locally, the following files must be
available for the system to work normally:

Technical Manual
12-2
 Remote databases

Local host
On the local host, the following files and directories must be present:
$BSE/bin:
audit_srv6.2 hostid6.2
ba6.2 kermit6.2
badmin6.2 licd6.2
Bdbpost6.2 licmon6.2
bdbpre6.2 lp6.2
Bdbreconfig6.2 nlsedit6.2
bic6.2popup6.2 pdaemon6.2
bic_cstub6.2 qcommand6.2
bic_info6.2 qptool6.2
Binperm6.2 refint6.2
binput6.2 repgen6.2
bput6.2 rz6.2
brand6.2 shmmanager6.2
Bshcmd6.2 shmvalues6.2
Bshell6.2 sort6.2
bsp.setperm6.2 srdd_init6.2
client6.2 std_gen6.2
Compress6.2 sum6.2
encrypt6.2 sz6.2
Explode6.2 tsscomp6.2
filter6.2 tsscvt6.2
Gcommand6.2 tssinfo6.2

$BSE/lib:
tabledef6.2
fd6.2.<package_comb> (optional)
ipc_info
/user/r<user>
$BSE/etc:
rc.start
rc.stop
$BSE/tmp
$BSE/log

Technical Manual
12-3
Remote databases

Remote
It is recommended that you install the complete $BSE-environment on the
remote system. If no application is started on the remote system, executable
programs such as Bshell6.2 and ba6.2 do not have to be installed there. It is
practical if one or more users can run the application locally on the remote
system. For this reason they must be run on the remote system. That is why the
entire BSE environment must be installed there.
The remote system must include at least the following files:
$BSE/bin:
fs6.2
ipc_boot6.2
database_servers with associated executable programs, such as ora7_srv6.2
installation and validation programs, such as install6.2, licd6.2, and so on.
$BSE/lib:
tabledef6.2
user/u<user>
ipc_info
driver-specific directory, for example, ora
fd.6.2.<package_comb> (optional)
$BSE/etc – rc.start
rc.stop
If these files are available on the local and the remote system, respectively, you
can adapt the following files. The data dictionary contains sessions where you
can adapt the files. It is better to adapt the files in the sessions in the data
dictionary and not directly in the files themselves.

Local
On the local system the following files are configuration dependent:
n tabledef6.2 Database Definitions (ttaad4510m000) and Tables by Database
(ttaad4111m000))
n u<user> Remote User Data (ttaad2500m000)
n r<user> Remote User Data (ttaad2501m000)

Technical Manual
12-4
 Remote databases

tabledef6.2
Start the application by executing the user interface (UI). The UI then starts the
application logic (Bshell), which reads the file tabledef6.2 to search for the table
location and database type. Tables can be stored on a local or remote system,
which is also indicated in this file.
The file tabledef6.2 is filled using the Database Definitions (ttaad4510m000) and
Tables by Database (ttaad4111m000) sessions on the menu BaanERP ToolsÈ
Database ManagementÈ Database Definitions and Directories. These
sessions indicate whether tables are installed on a remote database.
The file tabledef6.2 is located in the $BSE/lib directory. For more information
about tabledef6.2, refer to the help text for business object Database
Management, on the menu BaanERP ToolsÈ Database ManagementÈ
Database Definitions and Directories.
You can specify whether each table is installed locally or remotely by entering
the remote system name in the System Name field in the Database Definitions
(ttaad4510m000) session. Then use the Tables by Database (ttaad4111m000)
session on the menu BaanERP ToolsÈ Database ManagementÈ Database
Definitions and Directories to assign tables to this table definition.
For example, assume that all tables are installed on the remote system host1. This
means that tabledef6.2 on the local system must contain the line: *:*:host1:N.
If all tables are on a remote system you can set the variable BSE_REM on the
local system. If there is no tabledef6.2 file on the local system, the Bshell
searches for this on the remote system, otherwise it reads the local tabledef6.2
file. The variable BSE_REM indicates the remote system:
BSE_REM=host1

User files
To start an application, every user (logon) must have a user file. The Bshell reads
this user file to test whether the user is authorized to start the application. The
Remote User Data (ttaad2501m000) session creates the file
$BSE/lib/user/r<user>.
If all tables are installed on the remote system, the user files can be installed
there, too, instead of on the local system. In that case you can set the system
variable BSE_REM with the system name of the remote system, for example
BSE_REM=host1. The Bshell looks first in the local $BSE/lib directory. If it
cannot find the needed files there, it uses the BSE environment ($BSE/lib) on the
remote system.

Technical Manual
12-5
Remote databases

The remote system now reads the user file u user. To access tables on a remote
system, you need to log on to the remote system. This is done using a remote
logon user file named r user. You create this file using the Remote User Data
(ttaad2501m000) session. Read session Help for information on how to create
this file.

Variant I
Two variants of r user are possible. If the user’s logon is the same on the local
and the remote system, variant I is sufficient. Define the remote system, the path
on that system, and the password to log on to the remote system.
In variant I, the following information is needed:
n Name of remote host
n BSE path on remote host
n Encrypted password on remote host
EXAMPLE A user “peter” would like to work with BaanERP on his local work station. The
database tables he wants to use, however, are located on a remote system called
host. To access the remote data, he needs a file rpeter in the directory
$BSE/lib/user. If all tables are located on the remote system, he sets the
following system variable:
BSE_REM=host1
This makes sure that everything is read from the remote BSE environment except
the bin directory.
When using Variant I, user peter must have a logon entry on the remote system.
The file r user must contain the name of the remote system, the path of the BSE
on the remote system and the user’s encrypted password for the remote system.

Variant II
A second variant of r user allows you to log on to the remote host using another
logon code. Indicate whether the UNIX and BaanERP permissions must be
inherited from the other user. If so, the user file u user on the remote system
must have the same UNIX logon as on the local system.
In variant II, the following information is needed:
n Name of remote host
n BSE path on remote host
n Logon name on remote host
n Encrypted password on remote host
n SUID inherited logon

Technical Manual
12-6
 Remote databases

The remote user file can contain more than one line.
When using variant II, you can also specify a logon code to access the remote
system. This can be a different logon code than “peter,” for example “john.”
After that you enter the encrypted password for the remote logon (john). The
remote system must still have a user file upeter.
Variant II also allows you to take the UNIX and/or BaanERP authorizations
defined for the logon john. To do so, specify “yes” or “no” for the fields SUID
and Inherit Logon. SUID switches the user ID from peter to john. Inherit Logon
makes sure the BaanERP authorizations defined for john apply also to peter.
To switch the user ID, the file ipc_boot6.2 ($BSE/bin) must have UNIX root
permission root and the s-bit must be set. From a security point of view this is
not advisable.
Here is an example of the two variants:
Variant I:
host1 /usr2/bse NZ+,AHtg&#K-kt"Q{$=Y3-"VLDN’CRtZ
host2 /usr1/t/bse h8i0_-+8yujg@#c67h78i\{}907h67h5R
Variant II:
host1 /usr2/bse john NZ+,AHtg&#K-kt"Q{$=Y3-"VLDN’CRtZ no yes
host2 /usr1/t/bse h8i0_-+8yujg@#c67h78i\{}907h67h5R yes yes

Remote
Whenever a local system connects to the remote system, ipc_boot6.2 is the first
program to start. That process in turn starts fs6.2, which can read tabledef6.2 on
the remote system. It finds an entry in tabledef6.2, showing the database type for
a range of tables to be accessed.
A database server is started, for example, ora7_srv6.2. The path for both fs6.2
and the database server is read from the file $BSE/lib/ipc_info.
Once on a remote system, you cannot call a third system. This means that
tabledef6.2 must always have an entry on the tables.

Technical Manual
12-7
Remote databases

Figure 9, Local and remote tables

Local and remote tables

Tables are located on the local as well as on the remote host.
Another option is to distribute data storage over a local and a remote system. In
that case both the local and the remote system must have the complete BSE
environment installed.
The file tabledef6.2 now contains entries for local as well as remote system
tables. This can also be defined in the Database Definitions (ttaad4510m000) and
Tables by Database (ttaad4111m000) sessions on the menu BaanERP ToolsÈ
Database ManagementÈ Database Definitions and Directories.

Settings for remote menus/forms/objects

Not only databases, but also forms, menus, objects and scripts can be distributed
over one or more systems. The directories for the software components are
defined in the Directories of Software Components (ttadv1115m000) session.
This session fills the file fd6.2.package_combination. In this session you can
specify whether the component is on the local or remote system. See also the
session Help information.

Technical Manual
12-8
13 Multibyte management

General
The currently used 8-bit wide character space is not large enough to fit most
Asian character sets. Asian countries such as Japan, China, and Korea need at
least 16 bits of character space to accommodate local characters. Therefore a 32-
bit wide character space has been implemented in BaanERP Tools to support all
possible character sets.
Be aware that a character is not necessarily the same as a byte and can take up
more than one screen position. A character in the Bshell occupies one or two
screen positions at a size of 4 bytes.

Triton Super Set

Triton Super Set (TSS) is a collection of character sets that can be used in
BaanERP applications. TSS accommodates all multibyte character sets in a
single package. TSS includes both ASCII character sets and multibyte character
sets, such as Japanese, Chinese, Hebrew. To distinguish between, for example,
Japanese and Chinese, TSS consists of planes, each of which holds a 64KB-size
character set, or 256 pages of 256 characters.
This is shown in the following figure. Numbers are according to the hexadecimal
system. For example, the value of 0x21 is 33 in the decimal notation.

Figure 10, BaanERP Super Set (TSS)

Technical Manual
13-1
Multibyte management

The characters 0x00 – 0x20 and 0x7F – 0x9F cannot be included in the sets,
because these ranges contain control and escape codes. This leaves the following
ranges:
0x212121 – 0x7EFFFF
0xA12121 – 0xFFFFFF
To distinguish TSS characters, the character contains the prefix 0x9B. Hence the
complete TSS ranges are:
0x9B212121 – 0x9B7EFFFF
0x9BA12121 – 0x9BFFFFFF
Code Features (CF) and Line Drawing Characters (LDC) have been implemented
in the code range of ISO-8859/1 as follows:
n Line Drawing Characters | 0x80 – 0x8A
n Code Features | 0x8B – 0x9A
n TSS Lead byte | 0x9B
n Reserved | 0x9C – 0x9F
These code ranges do not conflict with the currently implemented EUC character
sets, but they can cause problems when you import these codes. Also these codes
overlap with SHIFT-JIS trailing bytes, causing import conversion errors.
Therefore the code ranges are translated into ASCII characters (0x00-0x7F)
during the export phase and are converted back into the correct LDC and CF
characters during the import phase.
Currently assigned planes are:

Character set Page Range from Range to Remarks

KANJIEUC/SHIFTJIS 0x21 212121 217e7e JISX0208-1983
KANJIEUC/SHIFTJIS 0x23 2321a1 2321df JISX0201-1976
CCDC_HP15 0x24 24a121 24fe7e CCDC
GB2312-80 0x25 25a1a1 25ffff GB2312-80
GB2312-80 0x26 26a120 26a17f Extended GB
BIG5 0x27 27a140 27f9d5
CNS11643 0x28 28a1a1 28fefe CNS11643 Plane 1
CNS11643 0x29 29a121 29fe7e CNS11643 Plane 2
HEBREW 0x30 3021a0 3021fa ISO-8859/8

Technical Manual
13-2
 Multibyte management

Using multibyte character sets

Installation
Entering or displaying multibyte characters is subject to a series of conditions.
The window manager must support multibyte characters and must have been
initialized in the proper native language. If necessary, change the shell variables
LANG and LC_ALL before starting the window manager. This prevents the title
bar of a window from displaying improper characters.
These settings are system dependent.
If necessary, follow the procedure to add a character to TSS, as described in the
BaanERP Super Set.
Enter the proper TSS locale in the Application Configuration (ttaad2100m000)
session. The locale is defined in the Maintain Locale Data (%SEtttss0104m000)
session and contains the definition of the character set.

Printing multibyte characters

To print multibyte characters, make sure the following conditions are met:
n A device created in the Device Data (ttaad3500m000) session must be linked
to a TSS locale that contains the definition of the character set in question. If
the proper locale is not specified, the printer daemon cannot print the report.
n The printer must be able to print the characters from the TSS locale. Check
your printer manuals.

Utilities
This section describes the available programs belonging to BaanERP Super Set.
These programs are present in the directory $BSE/bin.

tsscomp6.2
NAME
tsscomp6.2
This program compiles and checks TSS information file tss6.2. This program is
executed by the Convert TSS Data to Runtime (tttss0103m000) session.
SYNOPSIS
tsscomp6.2 [-svV] [-e error file] [-d debug level] [–i input file] [–o output file]

Technical Manual
13-3
Multibyte management

DESCRIPTION
tsscomp6.2 compiles the file tss6.2, which contains the descriptions of all
supported character sets into a fast-loading binary file tss_c6.2. It also performs
several checks on the entered data of the character sets.
The available options are:
-s Suppress output of errors. Check the exit value to determine if the
program has been successful.
-e Write all warnings and errors to file error file.
-vV Prints version and porting information.
-d Used to provide some debugging information. This is a bit set.
-i Optional input file. Default $BSE/lib/tss6.2.
-o Optional output file. Default $BSE/lib/tss_c6.2.
Return values:
n 0 = success
n > 0 = failure

tsscvt6.2
NAME
tsscvt6.2
TSS conversion filter.
SYNOPSIS
tsscvt6.2 [-vV] [-dow] [-l locale]
DESCRIPTION
This program converts multibyte characters strings into TSS strings or vice versa.
The input/output character set can be set using the –l option, but is read from the
user file (locale resource) if not given.
Use the –o option to convert from TSS to the native character set.
The available options are:
-vV show version and porting information.
-d show some additional debug information.
-w Do not accept any conversion warnings, but exit(1) instead.

Technical Manual
13-4
 Multibyte management

Return values:
n 0 = success
n > 0 = failure

tssinfo6.2
NAME
tssinfo6.2
This program offers information about the current settings such as your TSS
locale, NLS locale, character set information, and so on.
SYNOPSIS
tssinfo6.2 [-vV] [-sf] [-l locale]
DESCRIPTION
The available options are:
-vV show version and porting information.
-s show character set definition of TSS locale.
-f show font assignments.
-l show information about the given locale. This is read from the user file
if not given.
EXAMPLE Command: tssinfo6.2 –sf
User : bsp
TSS locale : KANJIEUC_AIX32
TSS character set : KANJIEUC
NLS locale : ja_JP
Factor :1
Multibyte strings:
Database min. : 0x1
Database max. : 0x9b217e7e
Forms min. : 0x20
Forms max. : 0x9b217e7e
Single byte strings:
Database min. : 0x1
Database max. : 0x7f
Forms min. : 0x20
Forms max. : 0x7a

Technical Manual
13-5
Multibyte management

Character set definition for KANJIEUC

NATIVE TSS
From To from to
1 7f 1 7f
8ea1 8edf 9b2321a1 9b2321df
a1a1 fefe 9b212121 9b217e7e

Font assignments

Font Display Size Font string

0 1 iso8859-1
1 2 jisx0208.1983-0
2 1 jisx0201.1976-0

Technical Manual
13-6
 Multibyte management

Character set description

The file tss6.2 in the directory $BSE/lib contains a description of the character
set. Normally this file is filled using the TSS Character Sets
(%SEtttss0101m000) session.
This file describes all supported character set with a special description language.
The format is as follows:
Set(<number>, "Name", "Description")
{
(
range <number>-<number>
in <expr>
out <expr>
store <number>
bytes <number>
setno <number>
)
[more ranges...]
}
<number> can be hexadecimal: 0xnn, octal: 0nnn or decimal: nnn.
<expr> can be:
1. simple type (fast)
& value binary AND with value
| value binary OR with value
+ value ADD value
- value SUBTRACT value
2. complex type (slow)
"expression".
See the "expr" function.
Range Defines the input range (native data).
In/out Convert the character data for input (conversion to TSS) and output
conversion to native character set).
Store Defines the place in the character set where the data is stored in TSS
after the in conversion expression (add 0x9b000000 to see where it is
stored, but not in the conversion entries).
Bytes Gives information about the number of bytes per native multibyte
character.
setno Refers to the X-character set-number as defined in
$BSE/lib/locale/LOCALE for display purposes. LOCALE is the current
locale as defined in the user file.

Technical Manual
13-7
Multibyte management

A character set usually consists of several subsets, such as ASCII, KANJI-EUC,

and KATAKANA in some Japanese environments.
NOTE Never change the character set data in an active environment. This can have
catastrophic results, and the changed conversion tables can corrupt your data
beyond repair.
EXAMPLE KANJI-EUC
Set(3, "KANJIEUC", "Japanese EUC")
{
# ASCII
(
range 0x1-0x7f
store 0x0
bytes 1
setno 1
)
# KATAKANA (JISX0208)
(
range 0x8ea1-0x8edf
in &0x00ff
out |0x8e00
store 0x232100
bytes 2
setno 3
)
# KANJI (JISX0201)
(
range 0xa1a1-0xfefe
in &0x7f7f
out |0x8080
store 0x210000
bytes 2
setno 2
)
}

Technical Manual
13-8
 Multibyte management

Locale information
The file tss_locale6.2 in the directory $BSE/lib contains all possible TSS locales
that can be defined in the user file or specified with special commands, such as
tsscvt6.2 –l locale. This file is filled by the Maintain Locale Data
%SEtttss0104m000) session. The format is as follows:
Locale, Charset, NLS_locale, MBdbLow, MBdbHigh, MBFormLow, \
MBFormHigh, SBdbLow, SBdbHigh, SBFormLow, SBFormHigh
Locale As specified in the user file.
Charset Any of the character sets defined in tss6.2.
NLS_locale The NLS locale. Mainly used for character collation. May be
NULL.
MbdbLow The lowest possible multibyte character value for database
storage.
MBdbHigh The highest possible multibyte character value for database
storage.
MBFormLow The lowest possible multibyte character value for forms (see
the expr function).
MBFormHigh The highest possible multibyte character value for forms (see
the expr function).
SBdbLow The lowest possible single byte character value for database
storage.
SbdbHigh The highest possible single byte character value for database
storage.
SBFormLow The lowest possible single byte character value for forms (see
the expr function).
SBFormHigh The highest possible single byte character value for forms (see
the expr function).
EXAMPLE # Japanese – HPUX 9.0
KANJIEUC_HPUX: KANJIEUC, Japanese.euc, 1, 65278, 32, \ 65278, 1, 127, 32, 122
# \ is used here to indicate that the line following this character must actually be
in the same line.
The directory $BSE/lib/locale contains files having names identical to the locale
names, with character set information used to display the character set ranges. A
set is chosen by the setno entry in the tss6.2 character set description file. These
files are managed by the Fonts (tttss0106m000) session.

Technical Manual
13-9
Multibyte management

EXAMPLE Taken from KANJIEUC_AIX

# character set display width
iso8859-1 1
jisx0208.1983-0 2
jisx0201.1976-0 1
The sequence number of these character sets is the value that is used as setno
entry in the tss6.2 file and in the TSS Character Sets (%SEtttss0101m000)
session.

Storage of multibyte characters

Normally the characters are stored as TSS characters. However, you can also
store the characters as native characters. To do that, the file tss_mbstore6.2 must
be created in the $BSE/lib directory. That file contains the database fields, whose
value or text must be stored as native instead of TSS characters.
native character : 8ea1
TSS character : 9b238ea1
The file is processed in the following order. The asterisk (*) refers to all tables.
*
ppmmm###
ppmmm###.field
ppmmm###ccc
ppmmm###ccc.field
(p:package, m:module, #:table number, c:company number)

EXAMPLE # entire table:

ttaad320
# only field stat from ttaad320000
ttaad320000.stat
Do not store native data as SHIFT-JIS. This conflicts with the LDC (line drawing
characters) and CF (code features) characters. It is better to store native data as
KANJI-EUC. You can use the SHIFT-JIS character set for the user interface
BX/BW. These programs have been designed to handle this conflict, by adding
special flags to all stored characters.
For other parts of the BaanERP application it is problematic to store the SHIFT-
JIS characters combined with LCD and CF characters. A good approach is to
choose tables for which data must be stored as SHIFT-JIS. These tables can be
manually added to the file tss_mbstore6.x.

Technical Manual
13-10
14 Audit management

General
This chapter introduces the audit management facility provided with BaanERP
Tools from Version 6.0 onward. It describes the audit server, the audit
management utility, and where audit data is stored. It also describes the structure
of the audit files. Finally, this chapter explains how to give users permissions to
print, maintain, or clean audit data.

Storage of audit data

For each table in each company number, the audit server uses a set of files to
store audit data. You can read audit files using the 4GL functions or the
dictionary programs. This section describes the types of file that are used.

Sequence files
The audit data is stored in a set of files called sequence files. There can be
multiple sequence files per table per company number.
Apart from storing audit data and transaction information, a sequence file also
stores information about audited columns of the table, the audit data dictionary.
The name of the sequence file is built using the table name and company
number. The format of the sequence file name, where prefix a indicates an audit
file, just as t indicates a table itself, is as follows:
a<mmmnnn><company number>.<sequence file number>
mmm represents the module code
nnn represents the file number
For example for table ttadv999 and company 000 the sequence files are:
aadv999000.000
aadv999000.001
aadv999000.002
aadv999000.003
..........
For detailed information of the sequence file, refer to the section “Format of
audit files.”

Technical Manual
14-1
Audit management

Information file
There is one information file per table per company number, which:
n Stores information about all sequence files used for this table/company
number combination
n Contains controlling information used to create and maintain sequence files
The name of the information file is based on the table name and the company
number. The format of the information file name is as follows, where mmm
represents the module code and nnn represents the file number:
a<mmmnnn><company number>.inf
For example, for table ttadv999 and company 0, the information file is
aadv999000.inf.
NOTE There is one set of files per table name/company number combination. This
means that any associated information, sequence file, or operation is identified
by the table name/company number combination. For the sake of readability this
document uses the word “table” to refer to a table name/company number
combination.

Location of audit files

The file $BSE/lib/auditdef6.X is used to specify the location of the sequence and
information files for the tables. Under the directory specified in auditdef6.X, a
subdirectory is created for each module. Audit files for all tables in this module
for all company numbers are stored in this directory, unless a different directory
is specified for different company numbers. The format of the directory name is:
a<Package Code><Module Code>
Here is an example of an entry in auditdef6.X:
ttadv:*:/usr/adv/AUDIT
*:*:/usr1/AUDIT
In this example the audit files for all ttadv tables are stored in the directory
/usr/adv/AUDIT/attadv. For all other tables, respective directories are created
under /usr1/AUDIT. For example, atfacr is created for the BaanERP Finance acr
module, atiitm for the BaanERP Industry itm module, and so on.

Technical Manual
14-2
 Audit management

There can also be an entry such as:

ttadv:*:/usr/adv/AUDIT/#
In this case all ttadv tables are stored under specific respective company
numbers. For example, for company number 000 the files are stored under
directory /usr/adv/AUDIT/000/attadv.

Other parameters of audit files

The location of the audit files is controlled as explained in the section “Location
of audit files.”
The user can control other parameters. These are defined in a file named
$BSE/lib/audit_spec. For more information on this file, and specifications of
parameters in it, refer to Format of Audit Files.
Information file
The tables for which audit trail is enabled can contain sensitive data such as
company financial information. This means that the audit files also contain
sensitive data. Audit management contains a security mechanism by which
access to the audit data can be controlled according to user requirement.
A security level is defined per table/company number combination. Depending
on the security level, users can or cannot access or modify the audit data. The
audit server reads the security level from the audit_spec file and stores it in the
information file of the table. The central audit management utility (see the Audit
Management business object), uses this security level to control audit data access
and modification.
Sequence File
The start and end sequence numbers for a table are specified in the audit_spec
file. For example, if for table ppbdb000 within company 000, the start sequence
is 0 and the end sequence is 999, the table has the following sequence files:
abdb000000.000
abdb000000.001
....
abdb000000.998
abdb000000.999

Technical Manual
14-3
Audit management

For the same table, if for company 100 the start sequence is 1, and the end
sequence is 50, the files are:
abdb000100.001
abdb000100.002
....
abdb000100.049
abdb000100.050
In principle, all audit data for a table/company number combination can be
written to a single sequence file. But because this file becomes too big to manage
properly, the audit data is split across multiple sequence files.
The maximum size of a sequence file for a table is defined in the audit_spec file.
If the limit is exceeded, the audit server tries to use the next sequence file.

Audit management
The audit management version 6.2 incorporates the following features in
BaanERP audit management utility:
n Audit data is stored in a set of files, instead of a single file. Multiple files
allow for better management than a single large file. See the section “Storage
of audit data.”
n Each transaction is identified by a transaction header, which logs useful
information about the transaction, such as the corresponding user and time.
Also, the data is organized per transaction. Each transaction has a header
followed by all audit data, which enables easy tracking of the changes that
users have made.
n When the audit data is being written, information of audited columns in
tables is also stored in audit files. If the table data dictionary is later changed,
you can still interpret old audit data by using the stored column information.
n Table level commands affecting table data such as create table, clear table,
and drop table, are also logged by the audit server.
n Audit management such as printing audit files and display audit sequences is
done by using the sessions belonging to the Audit Management business
object.

Technical Manual
14-4
 Audit management

Table data dictionary as seen by audit server

When a table data dictionary is created in a BaanERP environment, the audit trail
can be enabled for each column. This allows a user to audit a single column or all
the columns on the table.
The audit server is concerned only with audited columns of a table. It logs the
information only for the audited columns. If a column is not audited, no
information is logged for it. If the table has no audited columns, no information
is logged for the table. The file tabledef6.2 specifies the tables and columns for
which audit is enabled. See also the section “Accessing tables” in the chapter on
database management.
For the rest of this chapter, it is assumed that all columns are to be audited unless
stated otherwise.
The audit data dictionary (the information of audited columns) is stored in the
sequence header. This information is stored in a particular order of table
columns. First, information for all audited primary key columns is stored in the
order in which they form the primary key. Next, information of remaining
audited columns is stored in the order in which they occur in the table.
For example, suppose a table has columns c1, c2, c3, c4, c5, c6 with primary key
(c4, c2) and all columns except c5 are audited. The audit data dictionary contains
information for the columns in this order: c4, c2, c1, c3, c6. For each column, the
audit data dictionary stores the column name without the table name prefix, field
type, size, and depth. The format of a column entry in the audit data dictionary is
as follows:

Name Type Depth Size

Name
Stores column name without table name prefix.
Type
Single byte indicating field type. The defines in BDB layer for field type are
used, because 4GL also uses the same values. For example, for the CHAR field,
the BDB type is BDB_CHAR, and the 4GL counterpart DB.CHAR has the same
define as BDB_CHAR.
Depth
Single byte, indicating depth of the field.
Size
2 bytes, indicating field size. For array fields this is the size total of all fields in
array.

Technical Manual
14-5
Audit management

Commands logged by audit server

The audit server tracks all commands that affect the table data, that is, insert,
update, and delete commands. It also tracks certain table level commands such as
create table, drop table, and clear table, which affect all rows in a table.
For each of these commands, the audit server puts certain information in the
audit files. This information can be considered as Audit Row in the audit file.
The audit server does not track select commands (with or without lock), or other
table level commands such as change order, count rows, or create or drop index,
because they do not change the table data in any way.

Data stored by audit server

Row operations
For row operations, the audit server keeps the values of audited fields.
Insert actions
The new values of all audited fields to be inserted are stored in the audit file.
Delete actions
The field values of the record to be deleted are stored in the audit file.
Update actions
To comply with an application user requirement, the old and new values of the
fields that are changed in an update are stored. Primary key field values are also
stored, even if they have not been changed. In addition, if any other audited
column changes, its old and new values are stored. If an audited field is not
changed, and it is not a primary key field, its value is not stored in the audit row.
In addition to field values, a one-character code indicating the type of row
operation is also stored.

Table operations
For table operations such as create table, drop table, and clear table, it is
sufficient to store only an indicator. The audit server stores a one-byte indicator
for table operations.

Technical Manual
14-6
 Audit management

Application data
In accordance with the application user requirement, the audit server keeps 4
bytes extra space reserved in each audit row, to be used by application programs.
Application programs are free to use this space for their own usage. This 4-byte
space is called application data. The audit server does not interpret the
application data in any way.

Format of audit row

The format of the audit row in the sequence file is as follows:
Length Appl Data Action Data...

Length
A two-byte field that indicates the length of this audit row (excluding this field
itself).
Appl Data
A four-byte field reserved for application program usage. The audit server, when
writing a row, initializes this to nulls.
Action
A one-byte character code indicating the action done on the table. The following
codes are available:
-C create table
-R drop table
-L clear table
-I insert row
-D delete row
-U update row
Data
The previous fields fulfill the audit requirement for table-level actions. No Data
field is included for table-level actions.

Technical Manual
14-7
Audit management

For row level actions, the important values of audited fields are stored as data, as
explained in the following sections.
Insert/Delete actions
Values of all audited fields are stored for insert and delete actions. They are
stored in the same field order as that of the audit data dictionary of a sequence
header. If the audit data dictionary of a table has columns in the order c4, c2, c1,
c3, c6, the values are stored in the same order.
Update actions
For update actions, the field number precedes the field values in the audit row.
Field numbers used here are related to the sequence of fields in the audit data
dictionary and not to the original table data dictionary. Suppose a table has
columns c1, c2, c3, c4, c5, c6, with primary key (c4, c2), and all except c5 are
audited. The audit data dictionary contains information for these columns in
order c4, c2, c1, c3, c6. Thus the field number 0 refers to field c4 and not to field
c1. Similarly field 3 refers to c1 and not to c3. For update actions, an audit row
looks as follows:
Length Appl Data ’U’ Multiple field entries

A single field entry is as follows:

Field number Changed? Old value New value

Field number
A two-byte entry that indicates the field in the audit data dictionary.
Changed
A single-byte flag, used to indicate whether the field is updated or not. The field
value for primary key fields must be stored, whether it is changed or not. When
the field value is changed, the corresponding flag is set to Y, otherwise it is set to
N. The values for all other audited fields are stored in the audit row only if they
are changed, so for these fields, the flag is always set to Y.
Old Value
The old value of the field that is being updated. For a primary key column that is
not changed, this is the only value stored.
New Value
If an audited field value is changed, this represents the new value. For primary
key audited fields that are not changed, this field is not present in the audit row.

Technical Manual
14-8
 Audit management

Using this logic, the following is possible:

The primary key field (audited) is not changed. In this case the field entry is:
Field number N Old value

The primary key field or any other audited fields are changed. In this case the
field entry is as follows:
Field number Y Old value New value

Here the new value follows the old value.

Examples of the audit row for various operations

Suppose you have a table with the columns c1, c2, c3, c4, c5, c6 with primary
key (c4, c2). If all columns except c5 are audited, the audit data dictionary
contains information for these columns in order c4, c2, c1, c3, c6. The field
number 0 then refers to field c4, field number 1 is c2, and so on.
Create Table
Length Appl Data ’C’

Here the length is the same for all tables, 5.

Drop Table
Length Appl Data ’R’

Here the length is the same for all tables, 5.

Clear Table
Length Appl Data ’L’

Here the length is the same for all tables, 5.

Insert Row
Length Appl Data ‘I’ c4 c2 c1 c3 c6

Here c4, c2, and so on, indicate the values of the corresponding fields. Each field
entry size is the size of the field itself.

Technical Manual
14-9
Audit management

Delete Row
Length Appl Data ‘D’ c4 c2 c1 c3 c6

Update Row
Assume that c2 (primary key column) and c6 are changed.
Length Appl 'U' 0 N c4.o 1 Y c2.o c2.n 4 Y c6.o c6.n

Here, the entry for field 0 (c4) is present, although it is not changed, because the
field is a primary key field. Its only value stored is c4.o, the old value. Because
c2 (field 1) and c6 (field 4) are changed, their old and new values are stored as .o
and .n respectively.

Limit on the size of audit data

All audit data generated in a transaction for a table is always put in a single
sequence file. The transaction data is never split up across sequence files. If the
current sequence does not have enough space to write the transaction data (that is
if its size exceeds the maximum sequence size), the audit server tries to write the
transaction data in the next sequence file. If the transaction data cannot be
accommodated in the entire sequence file, the audit server cancels the
transaction. If this happens, the application programmer must either reduce the
scope of the transaction or increase the maximum file size for the sequence file.

Audit server
Long transactions and overflow file
The audit server buffers audit data in its memory buffers until the end of the
transaction. For example, if during a transaction 100 rows are inserted in a table
and then a commit is done, all the audit rows for the insert are buffered by the
audit server. This data is written to the audit files only at the time of commit.
The audit server uses individual memory buffers for each table. The size of the
buffers is set to enable them to hold large amounts of data. However, the memory
buffers might not be able to contain the results of a very large transaction. In this
case, the audit server uses an overflow file to buffer the remaining data.
The overflow file is created in $BSE/tmp. There is one overflow file per session
per server. Its name is made using the session ID and process ID (pid) of the
audit server, as follows:
ao.{Session Id>.<Process Id of the Audit Server>

Technical Manual
14-10
 Audit management

A single overflow file is used to buffer data for all the tables in the session for
which the memory buffers were insufficient. At the end of the transaction, if any
table has data in the overflow file. The overflow file is read and put in the
corresponding sequence file. After all data has been read and copied from the
overflow file, the overflow file is deleted.
If a transaction is canceled, the overflow file is deleted.
NOTE The memory buffers of the audit server can hold a large amount of data. If the
audit server still has to use the overflow file, the transaction is too big. This can
be a flaw in application design or coding. If possible, such transactions must be
avoided.

Sequence termination
The audit server uses a set of sequence files to store audit data for a table. The
audit server switches from one sequence file to the next under the following
circumstances.

Sequence file maximum size reached

Audit data generated in a transaction for a table is written to a single sequence
file. If the size of the transaction data exceeds the maximum sequence-file size
limit for the current sequence, the audit server marks the current sequence file as
terminated and writes the transaction data to the next sequence file.
Before the audit server switches to the next sequence, it ensures that the
maximum possible space is available. If the transaction data for the table still
cannot be accommodated, then it cannot be put in any sequence file. This is an
error condition and the audit server cancels the transaction. Note that in this case
the current sequence is not terminated. When this error occurs, you can either
increase the maximum size of the sequence file or reduce the scope of the
transaction.

Table audit data dictionary changed

When the audit server starts using a sequence, it writes the audit data dictionary
to the sequence headers of the sequence file and the information file. Audit data
in the sequence file depends on the audit data dictionary of that sequence. For
example, for an insert/delete action, all audited field values are put in the
sequence in the order in which they occur in the audit data dictionary. For
updates, the field values are preceded by field numbers, which are offsets in the
audit data dictionary. The reading of audit data by the 4GL interface is also
dependent on the audit data dictionary of the sequence.

Technical Manual
14-11
Audit management

If the audit data dictionary changes after the system writes some audit data to a
sequence file, you cannot continue using the same sequence. In this case, the
audit server stops using the current sequence and marks it closed while the audit
data dictionary has been changed. It starts using the next sequence.
Note that here the switch is made only if the audit data dictionary changes. This
is different from changing the table data dictionary. Changing table data
dictionary does not mean that the audit data dictionary is also changed.
Events that lead to changes in the audit data dictionary are in the following
example.
Suppose you have a table with columns c1, c2, c3, c4, c5, c6 with primary key
(c4, c2) in which all except c5 are audited. The audit data dictionary contains
information for these columns in order c4, c2, c1, c3, c6.
Consider the following changes:
Number of audited columns changed
If the audit trail status for any field is changed, the audit data dictionary changes.
For example, if audit trail is enabled for c5, the audit data dictionary must have
column information in the order c4, c2, c1, c3, c5, c6. This is different from the
audit data dictionary of the current sequence, so a sequence change occurs.
Similarly, if auditing is disabled for c1, the new audit data dictionary is c4, c2, c3
and c6, which is different from the current audit data dictionary. As a result, a
sequence change occurs.
Order of audited primary key columns changed
Initially the primary key was (c4, c2). If its definition is changed to (c2, c4), the
new audit data dictionary is c2, c4, c1, c3, c6. Because this is different from the
audit data dictionary of the current sequence, the audit server closes the current
sequence and uses the next sequence.
Order of other audited columns changed
The audit data dictionary changes if the order changes in which the columns
(other than the primary key columns) occur in the table definition. For example,
if the table is now defined as c1, c2, c4, c5, c6, c3, the new audit data dictionary
is c4, c2, c1, c6, c3. In this case the sequence order is changed.

Technical Manual
14-12
 Audit management

Number of audited primary key columns changed

You can change the number of audited primary key columns without affecting
any of these conditions. For example, if the primary key definition is changed to
(c4, c2, c1), the new audit data dictionary is the same as the current audit data
dictionary. But because c1 is now a primary key field, its value must be stored in
an update action even if it is not changed. In earlier updates in the current
sequence, this would not have been done. To make the sequence file consistent,
the audit server switches to the next sequence.

Sequence terminated by user

If current security allows them to, application users can terminate the current
sequence by using the central audit management utility (see the Audit
Management business object). Application programs can also terminate the
sequence. In this case the terminating program sets the sequence end date and
time and the sequence status.

Transaction termination
When a transaction is started, the audit server buffers audit data until the
transaction ends. Depending on the manner in which the transaction is ended,
two types of behavior are possible: simple commit/abort or two-phase commit.

Simple commit and abort

When doing simple commit, the buffered data for all tables in the transaction is
written to the corresponding sequence file, one table at a time. For example,
suppose a transaction involves four tables t1, t2, t3, and t4. When doing the
commit, first the audit data for t1 is written, then the data for t2, and so on. When
the data for all tables is written successfully, the audit server sends a success
code back to the client.
If writing data for any table fails, the audit server returns the E_BDB_ABORT
error to the client indicating that the transaction is canceled. If this happens and
the data is already written for some tables, it must be marked invalid. In this
example, if the audit server fails to write data for table t3 after writing data for
tables t1 and t2, the server changes the status of transactions for tables t1 and t2
to ABORT_DONE.
If the client decides to cancel the transaction, the audit server just discards the
buffered data for the tables, if any. No data needs to be written for any table.

Technical Manual
14-13
Audit management

Two-phase commit protocol

Beginning with Tools release 6.1c, BaanERP supports the two-phase commit
functionality for audit servers only. Two-phase commit is not supported for
database servers.
A three-part process takes place when an application issues a commit:
n Prepare audit drivers (first phase)
n Simple commit to all database drivers
n Commit audit drivers (second phase)
Prepare audit drivers (first phase)
To prepare for a commit, the system writes audit data to the audit files and sets
the PREPARE_DONE flag in the status field of the transaction header in the
sequence files of the tables involved.
If the preparation for a commit fails for any one table during the update of
multiple audited tables, the PREPARE_DONE is reversed. That is, the
ABORT_DONE flag is set for the tables on which the PREPARE_DONE was
set. PREPARE_DONE is not set for the remaining tables. The audit server issues
an E_BDB_ABORT error to the client to indicate that the transaction has
aborted.
Simple commit to all database drivers
Next a commit is issued to all the database drivers involved in the transaction. If
the commit fails for any of the database servers, the system issues an ABORT on
the remaining database servers. Because BaanERP does not support two-phase
commit for database servers, the system cannot issue an ABORT on the database
servers for which the commit was already issued.
Commit audit drivers (second phase)
After all database drivers have successfully committed, all audit servers involved
in the transaction are committed. A COMMIT_DONE flag is set in the
transaction status field of the transaction header for all the audited tables
involved in the transaction.
If the commit fails for any table, the transaction status field is set to
ABORT_DONE for all audited tables involved in the transaction, including
tables for which a COMMIT_DONE was set. When all the tables involved in the
transaction have both their PREPARE_DONE and COMMIT_DONE flags set,
the transaction has been successfully completed and successfully audited.

Technical Manual
14-14
 Audit management

How to use and set two-phase commit

If any of the audited tables involved in the transaction is specified in the file
$BSE/lib/xchdef, the transaction performs a two-phase commit. Otherwise, the
decision is based on the resource variable dbcinit. If dbcinit is set to 01, two-
phase audit commit is used for all audit transactions regardless of whether the
table is specified in the file $BSE/lib/xchdef.
If none of the audited tables involved in the transaction is specified in the file
xchdef, and if dbcinit is not set to 01, a simple audit commit is performed for that
transaction.
Performing a simple commit is fast. But performing a two-phase commit
guarantees that, if both PREPARE_DONE and COMMIT_DONE flags are set, a
full commit is done.
For Drop Table and Create Table actions, two-phase commit is always used
regardless of whether the audited table is specified in xchdef file or dbcinit is set.

Format of $BSE/lib/xchdef6
The format of an entry in the xchdef6 file is as follows:
table_name:company_number:{Y|N}
EXAMPLE ttadv:*:Y
tiitm001:*:Y
ttaad410:000:Y
*:*:N
In this example, two-phase commit is used for the following tables:
n all tables in the module ttadv for all companies
n table tiitm001 for all companies
n table ttaad410 for company 000 only
For other tables the decision is based on the setting of the dbcinit resource
variable. If the xchdef6 file is not present, the system behaves as though the file
were present with the entry:
*:*:N
In other words, the decision is based on the dbcinit resource variable only.

Technical Manual
14-15
Audit management

Reusing sequence files

Suppose you have a table with start sequence 0 and end sequence 49, for a total
of 50 audit files. In the first run, each sequence file is created. After all 50
sequences are created and used, the audit server normally uses sequence 0 again,
overwriting the existing file.
If you do not want this to happen, you can disable the reuse feature. Instead of
overwriting the file, the audit server then sends an error message saying that the
file must be deleted before the server can reuse the files. Thus when the reuse
feature is disabled, an existing sequence file is not overwritten. You have to
delete the file by using the Purge Audit Files (ttaad4161m000) session before
the audit server can continue. If the reuse feature is enabled, the audit server
overwrites any existing sequence file.
NOTE If you want to receive a message before the data in audit files is destroyed, you
must disable the reuse feature. This can be done by not specifying a SEQREUSE
keyword in audit_spec file. If you want to use the reuse feature, include the
keyword SEQREUSE in the audit_spec file.
Always delete the sequence files using the Purge Audit Files (ttaad4261m000)
session rather than by using commands such as rm or del. This is because the
status of the sequence file is maintained in the sequence header in the
information file. When this session deletes the sequence file, it also sets the
status of this sequence in the information file as deleted. If the file is removed by
any other means, the status is not set and this results in inconsistency.

File locking
When writing the data for a transaction, the audit server can close the current
sequence and start using the next sequence. Thus it can change the current
sequence field in the information file header. It also changes the sequence status
and sequence end time for the current sequence, and changes the status and start
date/time of the next sequence.
Even if the audit server does not switch sequences, it needs to change the current
offset in the information header and the number of transactions in the sequence
header. This means that it always changes the information header and the
sequence header. Moreover, because the sequence header is maintained in the
sequence file and information file, both of these must be changed.

Technical Manual
14-16
 Audit management

At the start of a transaction termination command for each table and before
writing buffered data to the audit file, the audit server locks the information file
header and sequence header in the information file and sequence file. The audit
server puts write locks on, so that no other read or write lock can succeed. After
the data for all tables is written, the locks on all tables are released. If there is an
error on any table, the status of transactions for tables already written is changed
to abort and the locks on all tables are released.

Limit on open files

When writing audit data for a table, the audit server has to open the information
file and one sequence file for that table. Once these are opened, they are kept
open for subsequent use even after the end of a transaction. These open files can
be used in all subsequent transactions involving this table.
Having so many files open, however, means that the server can reach the
OPERATING SYSTEM limit of maximum open files per process. For example,
if a transaction involves 15 tables, 30 files are opened at the end of that
transaction. If the next transaction has 10 different tables, 20 more files are
opened. At the end of the second transaction 50 files are open, which is over the
maximum limit.
When this limit is reached, the server closes two open files: the information file
and the sequence file for a table. The server uses the least recently used (LRU)
criteria to choose the table for which the files are closed, as explained in the
following paragraph.
The audit server uses a data structure, called a handler, to keep track of the file
descriptors for the information and sequence files on each table. When these files
are opened, the file descriptors are stored in the handler. The handlers of all
tables are kept in a linked list in an LRU manner. For example, when files are
first opened, the handler is allocated and is put in most recently used (MRU)
position. As other tables are opened, this handler moves toward the LRU
position. Later on, if this table is involved in another transaction, the same
handler is reused and is again moved to the MRU position. Also, when a file is
open, all data is written to the file before the data goes to the next table. Thus,
until all data is written for a table in the current transaction, the handler for that
table is guaranteed to stay in MRU position and in no way can it reach LRU
position.
If the audit server reaches the open file limit when opening a new information
file, sequence file, or overflow file, it picks up the LRU handler in the handler
list and closes its information file and sequence file. The assumption is that
because the table has not recently been used, it is the best choice as a victim for
closing files.

Technical Manual
14-17
Audit management

Opening large number of tables in single session

A session can update a large number of tables. On systems that have a small
maximum number of open files, this can create a problem. Suppose a session
with a limit of 30 open files starts to update 50 tables. After opening the
information and sequence files for the first 30 tables, the system reaches the
maximum limit of open files. When files for table31 are opened, the LRU logic is
activated and files for table1 are closed. But table1 is itself in transaction and its
files are locked. When the audit server closes the files, the locks are released and
other processes can alter the files. Now the same audit server may again need to
operate on the audit files of table1.
This means that the audit files for table1 need to be opened again. But if the files
were changed in the meantime, the status is changed to ABORT_DONE. In this
case, the audit server gives an error and cancels the transaction.

Multisession support
The audit server supports multiple sessions, which means that the same audit
server is used for multiple sessions within the client.
The audit data dictionary of a table is kept globally. The same table/company
number in multiple sessions shares a single audit data dictionary.
The handlers (one used per table/company number combination) are kept in
global lists. If multiple sessions operate on the same table/company number
combination, they share a single handler.

File security for various audit management files

It is important that the audit_spec file has read permission for all users, and that
write permission is granted only to the Administrator user. Otherwise users other
than Administrator can change the security parameter in the file, which would
allow them illegal access.

Technical Manual
14-18
 Audit management

Audit server debugging options

For debugging purposes, a code is put in the audit server at appropriate stages to
print useful information. In normal usage, this printing is not enabled. You can
enable it by setting an environment variable AUDITLOG. You can print various
types of information by setting this variable to different values. The values must
be set as octals.
0001 General information
0002 Print audit rows generated
0004 Print audit data dictionary for table
0010 Print data structures such as information header, sequence header, and
transaction header
The information is put in a log file. The name of the file is audit.log.pid of the
server. This file is created in the current directory of the audit server.

Audit errors
The audit server uses the following error messages. If you receive one of these
error messages, refer to $BSE/log/log.audit for more information regarding the
cause of the error.
E_AUD_SETUP 251 The audit setup is not correct.
This can occur if the specification in $BSE/lib/audit_spec file is not in proper
format.
TOSEQ 6 SECURITY 28 MAXSEQSIZE 25K or
No table/company number combination
aa:00a:TOSEQ 6 SECURITY 28 MAXSEQSIZE 25K
Incorrect company number.
*:*:TOSEQ 6 SECURITY 28 MAXSEQSIZE 5K
The maximum sequence size limit is less than the specified limits. Refer to the
section “Format of audit files” for details about the format of this file. It can also
mean that the current sequence file was illegally removed or that the sequence
files were removed without using the 4GL interface. Audit files used during the
first run can also cause this error.
E_AUD_CORRUPT 252 An audit file is corrupt.
Some corruption of the information file has occurred. There is a mismatch of
sequence headers in information file and sequence file.

Technical Manual
14-19
Audit management

E_AUD_LOCKED 253 Another user has locked the audit file.

Multiple users are trying to use the same audit file at the same time. The audit
file is locked before it can be modified, and the next user receives an error
message.
E_AUD_ABORT 254 Commit transaction has failed in audit server.
The audit file is changed when the audit data is written to the audit file. When
too many files are opened and the maximum limit is reached, the LRU file is
closed. This error can refer to the file used in the transaction. When the file is
reopened, the starting date time of the sequence header is checked with memory
data structure. If they are not the same, the E_AUD_ABORT error is returned.

Authorizations
The Administrator user grants audit security permissions to users using the Audit
Authorizations (ttaad4562m000) session, on the menu BaanERP Tools, Audit
Management. These permissions determine whether the user is allowed to use
the sessions in the Audit Management business object.
Security permissions assigned to users are checked against previously existing
records to see if they conflict.
You can assign security permissions for all packages, modules, table numbers,
and companies. You can give some permissions for all packages, and disable
these permissions by giving no permissions to a range of packages.
You can use the Specified command to give particular permissions to a specified
package. By using Specified, the highest priority is given to the security
permissions assigned for that particular package, module, table and company.
You cannot assign security permissions for a range within a range. Suppose you
assign some permissions using the All command, then disable permissions for a
range between tt—vv. You cannot then enable permissions within this range or
an overlapping range. However, you can use the Specified command to enable
the permissions for a specific package.
The conflicting records are not stored in the table.
The security permissions for users are stored in a table ttaad462, which you can
access accessed using the sessions within the Audit Management business object.
The security checking is done at two layers, one at the user level and another for
reading audit information. The user can have permissions at the user level but not
for reading the audit information. At audit level, the security permissions are
stored in the audit information header of the audit information file.

Technical Manual
14-20
 Audit management

User level permissions Audit level permissions

None Print
All Clean
Print Maintain
Clean Application-defined
Maintain

At user level, the Administrator user can grant permissions in different

combinations. The Administrator user assigns its own permissions to print and
clean audit information through the Audit Authorizations (ttaad4562m000)
session, on the menu BaanERP Tools, Audit Management. The only session
the Administrator user is allowed to execute without being assigned any security
permissions, is the maintain session. For example:
None No permissions
All Print, clean, and maintain permissions
Print Only print permissions
Print/Clean Print and clean permissions
Print/Clean/Maintain Same as All
NOTE None and All are mutually exclusive.
At Audit level the permissions can be granted in combinations except
Application defined. Only Maintain can be combined with Application-defined.
Print Only print permissions
Print/Clean Print and clean permissions
Print/Clean/Maintain Same as all permissions
Application defined Access to some sessions restricted
Maintain/Application Only Maintain permission, no print and clean
defined permissions
The user is not allowed to print or clean up any audit information. To print,
clean, or maintain the audit information from the audit files, the permissions
must exist at both levels.
Permission at user level Print or print/clean
Permission at audit level Application defined

Technical Manual
14-21
Audit management

The user is not allowed to Print or Clean up any audit information in the audit
files using the sessions in the Audit Management business object. It is possible,
however, to run the Audit Information File (ttaad4160s000) session, because
the user has Maintain permission only at audit file level.
Permission at user level Maintain or Print/Maintain or
Clean/Maintain
Permission at audit level Maintain/Application defined
If the user is not Administrator and only Application-defined exists at the audit
file level, no session of the Audit Management business object can be executed.
The Administrator user can use the maintain session even if no permissions are
assigned to the Administrator user. However, the Administrator user cannot use
print and clean sessions unless security permissions are assigned to
Administrator.
If user bsp has Print permission at user level, a further check is made to see the
security permissions at audit level. User bsp is only allowed to continue if print
permission is available at the audit level. Otherwise the session is canceled and
an appropriate message is displayed.

Format of audit files

auditdef6.2
This file is used to specify the directories in which audit files are created. There
is one such file for each BSE environment, and it is located in $BSE/lib.
This file has following structure:
<Table Name>:<Company Number>:<Directory Name>
The following section lists the meaning of each field.
Table Name
This indicates the table. This can be in the following formats:
"*" – This means all tables in all packages.
Package – This means all tables in this package. For example tf, td etc.
Package & Module code – All tables in the module. For example, tccom,
ttadv.
Package, Module and table number – for example ttadv999, tccom010.

Technical Manual
14-22
 Audit management

Company Number
Tables can be chosen based on company number. For example:
* – All companies.
100 – Company 100 only.
100,000 – Company 100 and 000.
Directory
Directory in which the audit files are stored. Under this directory, another
directory is created, called:
aPackage + Module Code

Information file
There is one information file per table per company number. You can read the
information file by using the 4GL functions or the dictionary programs.
The information file contains the controlling information for the sequence files,
called the information header, followed by the sequence headers of all the
sequence files created.
The format of the information file name, where mmm means module code and
nnn means table number, is as follows:
a mmmnnn company number.inf
For example, the name for table ttadv999, company 0, is aadv999000.inf.
The structure of the information file is a header called information header,
followed by multiple sequence headers, one for each sequence created.
Information header
n Version Information
Used to store the version information. This must never be modified by the
application by any means.
Length: 4 bytes
n Table name
The table for which audit information is stored.
Length: 8 bytes.
n Company Number
The company of the table.
Length: 2 bytes.

Technical Manual
14-23
Audit management

n Status
Specifies whether sequence reuse is allowed or not.
Length: 1 byte.
n From Sequence Number
Audit data is stored in sequence files specified from this value.
Length: 2 bytes.
n To Sequence Number
The last sequence file after which sequence reuse is started, or the user must
take appropriate action, such as deleting the existing file before reusing them.
Length: 2 bytes.
n End Sequence Number
Specifies the last sequence written in the information file. This field is for
internal use by Audit Server only! The user must never change this field,
otherwise the audit server will not work correctly.
Length: 2 bytes.
n Max. Size of 1 audit file in Bytes
Specifies the maximum file size of a sequence file.
Length: 4 bytes.
n Security Level
This lists the users who have access to the information in the audit files. The
security permissions regarding access to audit information are stored here.
Length: 2 bytes.
n Current Sequence Number
Determines the current sequence number that is used to write the audit data in
the sequence file.
Length: 2 bytes.
n Offset in current sequence file.
Specifies the offset in the current sequence file where the audit data is
written.
Length: 4 bytes.
n Reserved space
Space that is reserved for future use. Application programs must never use
this space because it might be required for a future release of the audit server.
Length: 8 bytes
n Length of following sequence header
Length: 2 bytes

Technical Manual
14-24
 Audit management

Sequence Header
Each sequence file also begins with a sequence header. The information file
contains copies of sequence headers for all sequences created.
n Sequence Number
This specifies the sequence number of the sequence file.
Length: 2 bytes.
n Creation Date
This specifies the creation date of the sequence file.
Length: 8 bytes, format: YYYYMMDD.
n Creation Time
This specifies the creation time of the sequence file.
Length: 6 bytes, format: HHMMSS.
n Termination Date
This specifies the termination date of the sequence file.
Length: 8 bytes, format: YYYYMMDD.
n Termination Time
This specifies the termination time of the sequence file.
Length: 6 bytes, format: HHMMSS.
n Status
The status gives information to the user whether the sequence file was user
terminated, closed because of a data dictionary change, or closed because
transaction data could not be accommodated in the sequence file.
Length: 2 bytes.
n Number of transactions in the sequence
This specifies the data of number of transactions present in the sequence file.
Length: 2 bytes.
n Application Information – 1
Application Information – 2
Application Information – 3
Application Information – 4
These bytes are reserved for storing data specific to the application. If the
application needs to store any information in audit files, these fields are used
to store that information.
Length: 4 bytes, per application information field.
n Number of fields being audited
The total number of fields that are being audited.
Length: 2 bytes.

Technical Manual
14-25
Audit management

n Number of fields in primary index

Indicates how many fields are in primary key.
Length: 1 byte.
n Field Name (Length: 8 bytes), Field Type (Length: 1 byte), Field Depth
(Length: 1 byte), and Field Length (Length: 2 bytes)
Field information for all fields that are being audited. First, the fields that are
part of the primary index are written in the same sequence as in the primary
index. Next, all remaining fields are written in the order in which they occur
in the table data dictionary.
EXAMPLE For example, if fields f1, f2, f3, f4 are being audited and the primary index is
on f4, f2, then field information is written first for f4, then for f2, and then for
all remaining fields.
n Reserved space
This space is reserved for future use.
Length: 8 bytes.
n Length of following sequence header
Length: 2 bytes.

Sequence file
Multiple sequence files are used to store audit data for each table/company
number combination. You can read the sequence files by using the 4GL
functions or the dictionary programs. The format of the sequence file name,
where mmm refers to the module code and nnn means table number, is as
follows:
a mmmnnn company number.sequence file number
For example, these sequence files can exist for table ttadv999, company number
0:
aadv999000.000
aadv999000.001
aadv999000.002
aadv999000.003
....
The sequence file contains the table name and company number, followed by the
sequence header. This is followed by transaction data of multiple transactions.
Transaction data is followed by a transaction header and contains multiple audit
rows generated for the transaction.

Technical Manual
14-26
 Audit management

Sequence File Layout

n Table Name
Length: 8 bytes
n Company Number
Length: 2 bytes
n Length of sequence header
Length: 2 bytes
n Sequence Header
n Transaction Header
n Transaction Data (Audit rows)
n Transaction Header
n Transaction Data (Audit rows)
n Transaction Header
n Transaction Data (Audit rows)
n ....
Transaction Header Format
You need to log both the operating system user ID and the BaanERP user name.
This is because a user can change the USER environment variable and use
different BaanERP user names. Also, different BaanERP users can have different
table privileges in SQL databases. Programs can find the operating system logon
name by using the operating system user ID print.
EXAMPLE Operating system user id 4
Baan User Name 8
Date ( YYYYMMDD ) 8
Time ( HHMMSS ) 6
Session Name 15
The transaction status can have one of the following values:
COMMIT_DONE
ABORT_DONE
From Baan IVc onward, the status can also have the following flags. The values
for these flags can be found in the audit include file.
PREPARE_DONE
GATI_UTC_FLDS

Technical Manual
14-27
Audit management

Transaction Status 1
Number of audit rows in the transaction 2
Also since Baan IVc, the following fields are added at this position.
{
GATI 0 4
GATI 1 4
Commit Date ( YYYYMMDD ) 8
Commit Time ( HHMMSS ) 6
Reserved space 8
}
Total bytes of data of this transaction 4
You need to log both the operating system user ID and the BaanERP user name.
This is because a user can change the USER environment variable and use
different BaanERP user names. Also, different BaanERP users can have different
table privileges in SQL databases. Programs can find the operating system logon
name by using the operating system user ID print.
n Operating system user ID
Length: 4 bytes
n BaanERP user name
Length: 8 bytes
n Date (YYYYMMDD)
Length: 8 bytes
n Time (HHMMSS)
Length: 6 bytes
n Session name
Length: 15 bytes
n Transaction status
Length: 1 byte. The transaction status can have the following values:
COMMIT_DONE
ABORT_DONE
n Number of audit rows in the transaction
Length: 2 bytes
n Total bytes of data of this transaction
Length: 4 bytes

Technical Manual
14-28
 Audit management

Transaction Data Format

The format of the audit row is as follows:
{
Length of the row 2
{
SNAT 4
}
Application Data 4
Type of operation 1
Row Data ...
}
n Length of the row (Length: 2 bytes)
n SNAT (Sequence Number of action in Audit Transaction) (Length: 4 bytes)
n Application data (Length: 4 bytes)
n Type of operation (Length: 1 byte)
n Row data
Four bytes of space are kept in every audit row to store application data. This
data is stored and manipulated by 4GL applications and not by the audit server.
The audit server keeps the 4 bytes of extra space to be used later. The type of
operation is a single character field, indicating the database action performed on
the table.
The following characters are used to indicate audited database actions.
n C – Create table
n L – Clear table
n R – Drop (remove) table
n I – Insert row
n D – Delete row
n U – Update row
In the audit row, the row data is present only for row level actions. The field
values, when stored, are stored in the same format as their data type. This means
that a value for a long field is stored as a long data type in the audit row. The
length of each field entry in the audit row is the same as the length of the column.
The row data for an insert consists of the field values to be inserted of all audited
fields, in the order in which they occur in the audit data dictionary.
The row data for a delete consists of the field values of the row to be deleted and
of all audited fields, in the order in which they occur in the audit data dictionary.
Not all fields in the table are updated for an update. Apart from storing field
values, an identifier is required to specify the field to which this data belongs.

Technical Manual
14-29
Audit management

Field numbers are used for identifying the fields. The field number is its position
in the AUDIT data dictionary, starting with field 0.
An entry for a field contains the following information:
n Field number – 2 bytes
n Is the value changed – 1 byte (Y or N).
n Old value
n New value
If a field is changed here, both old and new values are stored. If an audited
primary key field is not changed, only the old value is stored. If other audited
fields are not changed, nothing is stored. For all changes made to all audited
fields, entries are made to the audit data.
n Transaction status
Length: 1 byte. Transaction status can have the following values:
− COMMIT_DONE
− ABORT_DONE
n Number of audit rows in the transaction
Length: 2 bytes
n Total bytes of data of this transaction
Length: 4 bytes
Transaction Data Format:
n Length of the row (Length: 2 bytes)
n Application data (Length: 4 bytes)
n Type of operation (Length: 1 byte)
n Row data

Four bytes of space is kept in every audit row to store application data. This data
is stored and manipulated by 4GL applications and not by the audit server. The
audit server keeps the 4 bytes of extra space to be used later. The type of
operation is a single character field, indicating the database action done on the
table.
The following characters are used to indicate audited database actions.
n C – Create table
n L – Clear table
n R – Drop (remove) table
n I – Insert row
n D – Delete row
n U – Update row

Technical Manual
14-30
 Audit management

In the audit row, the row data is present only for row level actions. The field
values, when stored, are stored in the same format as their data type. This means
that a value for a long field is stored as a long data type in the audit row. The
length of each field entry in the audit row is the same as the length of the column.
The row data for an insert consists of the field values to be inserted of all audited
fields, in the order in which they occur in the audit data dictionary.
The row data for a delete consists of the field values of the row to be deleted and
of all audited fields, in the order in which they occur in the audit data dictionary.
Not all fields in the table are updated for an update. Apart from storing field
values, an identifier is required to specify the field to which this data belongs.
Field numbers are used for identifying the fields. The field number is its position
in the AUDIT data dictionary, starting with field 0.
An entry for a field contains the following information:
n Field number – 2 bytes
n Is the value changed – 1 byte (Y or N).
n Old value
n New value
If a field is changed here, both old and new values are stored. If an audited
primary key field is not changed, only the old value is stored. If other audited
fields are not changed, nothing is stored. For all changes made to all audited
fields, entries are made to the audit data.

audit_spec file
The audit_spec file is used to specify audit file parameters. There is one such file
for each BSE environment, and it is located in the $BSE/lib directory.
The format of the file is
Table Name:Company Number:Audit Specifications
Following is the meaning of each field.
n Table Name
Indicates the table. The table name can have the following formats:
*
All tables in all packages.
Package
All tables in this package, for example tf, td, and so on.

Technical Manual
14-31
Audit management

Package & module code

All tables in the module, for example tccom, ttadv.
Package, module and table number
For example ttadv999, tccom010.
n Company
This can be in one of following formats.
*
All companies.
100
Single company for example, company 100 only.
100,000
Comma-separated list of companies, for example, company 100 and 000.
n Specifications
The following specifications can be put in the file.
Start sequence
This is specified by the keyword FROMSEQ or fromseq. The start sequence
is always 000. The user cannot change this.
To sequence
This is specified by the keyword TOSEQ or toseq followed by a sequence
number. Note that there must be a space between the keyword and the
sequence number. For example:
TOSEQ 999 or toseq 999
If the end sequence is not specified for a table/company combination, the
default is 999.
Maximum sequence file size
This is specified by the keyword MAXSEQSIZE or MAXSEQSIZE
followed by a size specification. Note that there must be a space between the
keyword and the size specification. The size can be specified in:
− bytes for example, MAXSEQSIZE 10000 (10000 bytes)
− Kilobytes for example, MAXSEQSIZE 10K (10KB or 10*1024 bytes)
− Megabytes, for example, MAXSEQSIZE 2M (2MB or 2*1024*1024
bytes)
Note that for kilobyte and megabyte specifications, the characters K and M
must be uppercase only and there cannot be a space between the number and
the specification character.

Technical Manual
14-32
 Audit management

If the maximum sequence size is not specified for a table/company

combination, the default is 512K.
Whether sequences can be reused or not.
If existing sequences can be reused without deleting them first, the keyword
SEQREUSE or seqreuse must be put in the specifications. If you do not
want this, do not put the keyword in the specifications. If this is not specified
for a table, the sequence reused is not allowed.
Security Assigned to current table/company number audit files.
This is specified as keyword SECURITY or security followed by a security
number. Note that there must be a space between the keyword and the
security specifications. For example:
SECURITY 4 or security 4
The different security levels and their codes are as follows:

Security Value
Print 4
Clean 8
Maintain 16
Application defined 32

The combination of these is explained in the section “Authorizations.”

To specify Clean and Maintain, security is specified as security 24
(the sum of 8 + 16).
Application-defined cannot be combined with Print and Clean. It can be
combined with Maintain permissions only. Note that these specifications can be
put in any order.
CAUTION The audit_spec file must have a default entry covering all tables/company
numbers. This can be done with an asterisk (*) for table and company number
fields. This must be the last line of the file.
A sample audit_spec file can look as follows:
ttadv999:000:TOSEQ 99 FROMSEQ 000 MAXSEQSIZE 2M
tt:000:seqreuse toseq 49 maxseqsize 1M
tf:000:toseq 49 seqreuse maxseqsize 20K
*:*:FROMSEQ 000 TOSEQ 499 MAXSEQSIZE 100K

NOTE The audit_spec file must have read permission for all users and write permission
for the Administrator user only. Otherwise users other than Administrator can
change the security parameter in the file, which would allow them illegal access.

Technical Manual
14-33
Audit management

Technical Manual
14-34
15 OLE Automation

General
OLE Automation is an industry standard that applications use to expose their
OLE objects to development tools, macro languages, and other applications that
support OLE Automation. For example, a spreadsheet application can expose a
worksheet, chart, cell, or range of cells—all as different types of objects. A word
processor can expose objects such as application, document, paragraph, sentence,
or selection.
When an application supports OLE Automation, the objects it exposes can be
accessed by Visual Basic. You can use Visual Basic to manipulate these objects
by invoking methods on the object or by getting or setting the object’s properties.
For example, if you create an OLE Automation object named MyObj, you can
write the following code to manipulate the object:
Dim MyObj As Object ’declaration

Sub PrintWordDoc()
Set MyObj = CreateObject("Word.Basic") ’start MS Word
MyObj.FileNewDefault ’open a new file
MyObj.Bold 1
MyObj.Insert "Hello, world." ’insert new text
MyObj.FilePrint ’print current file
MyObj.FileSaveAs "C:\WORDPROCDOCS\TESTOBJ.DOC"
MyObj.AppClose ’close MS Word
Set MyObj = Nothing
Exit Sub
End Sub

This example runs Microsoft Word to create and print a document with the text,
“Hello, world.”
The application that exposes objects is called an OLE Automation server, the
application using those objects is called an OLE Automation client.

Technical Manual
15-1
OLE Automation

In fact the OLE Automation interface can be compared to the C Interface of

BaanERP. Where the C Interface offers an interface from other C programs to
BaanERP, OLE Automation offers an interface from other MS-Windows
applications to BaanERP. The implementation is based on the BaanERP 4GL
function parse_and_exec_function in the same way as the C Interface.

BaanERP as OLE Automation server

In combination with the BW user interface, BaanERP is also an OLE
Automation server. The BaanERP application exposes an object through which
clients can call BaanERP dynamic link library (DLL) functions. This section
describes how you can create and use this BaanERP object.
The BaanERP application exposes just one object, which is the application
object. In Visual Basic this object can be created with:
Set BaanObj = CreateObject("Baan.Application")

This starts BaanERP with the BW user interface invisibly, in other words only
the Option Dialog icon is displayed. OLE Automation objects, such as BaanObj,
can have methods and properties. A method is a function that can be started to
perform a certain action. A property is an attribute that has a value. Property
values can be set and/or retrieved.
The BaanERP application object has the following methods:
n ParseExecFunction “dllname”, “functioncall”
n Quit
The BaanERP application object has the following properties:

Name Type Value can be

Timeout long Set
FunctionCall string Set
Error long Retrieved
ReturnValue string Retrieved
ReturnCall string Retrieved
Binary boolean Set

Technical Manual
15-2
 OLE Automation

These methods and properties allow Visual Basic programmers to call any
function in any DLL. For example:
BaanObj.Timeout = 10 ’ 10 seconds timeout
BaanObj.ParseExecFunction "mydll", "myfunction(arg1, arg2)"
If BaanObj.Error <> 0 Then MsgBox("An error occurred")
MsgBox("The return value is " + BaanObj.ReturnValue)

This calls the function myfunction in the DLL named mydll. Note how
arguments are passed to myfunction. If arguments are passed by reference they
can be changed by the function call, therefore the property ReturnCall contains
the modified arguments after the call.
The property Timeout is used to prevent blocking on erroneous calls. The
property Error contains the return status of the ParseExecFunction call, and the
property ReturnValue is the actual return value of the DLL function. If the DLL
function returns a long value, it is converted to a string in ReturnValue.
To return binary data in ReturnCall arguments, the Binary property must be set to
true. ReturnCall can contain null characters.
Possible values for the property Error are:
n 0 success
n -1 unknown DLL name
n -2 unknown function name
n -3 syntax error in function call
n -10 timer expired
n -11fatal error in function
To end an automation session you can quit a BaanERP application as in this
example:
Set BaanObj = Nothing

Technical Manual
15-3
OLE Automation

Restrictions
When you use the OLE Automation interface, you must consider the following
restrictions:
n Due to the 3GL function parse_and_exec_function, you cannot supply
arrays as arguments to a DLL function.
n DLL functions with a variable number of arguments or optional arguments
are not supported.
n When a string is passed by reference, the result cannot become longer than
the input string. When a string must be filled make sure it is initialized with
the same number of spaces as the maximum length that can be returned. The
maximum length is defined in the domain of the variable.
n The maximum size of the FunctionCall string is 4KB (Bshell limit).
n The maximum size of the ReturnValue string is 4KB (Bshell limit).
n Two client applications cannot simultaneously use the same BW to make
DLL function calls.

Example: Import BaanERP users

This example fills an MS-Excel spreadsheet with the names of all BaanERP
users. To do this, you must:
n Implement DLL functions to obtain the names of the users from the database.
n Implement some Visual Basic code in MS-Excel.

Technical Manual
15-4
 OLE Automation

DLL function example

The following DLL, called demo.dll, contains two functions that retrieve the
names of the users from the database. One function initiates the query and gets
the first name, and the other gets the next name. When the last record has been
returned, an empty string is returned.
table tttaad200
long sql_id
function extern string get_first_user()
{
string sql_query(512)
string user(512)

switch.to.company(0)
sql_query = "select ttaad200.user from ttaad200"
sql_id = sql.parse(sql_query)
sql.exec(sql_id)
e = sql.fetch(sql_id)

if e = 0 then
user = ttaad200.user
else
user = ""
endif

return(user)
}
function extern string get_next_user()
{
string user(512)

e = sql.fetch(sql_id)
if e = 0 then
user = ttaad200.user
else
user = ""
endif

return(user)
}

Technical Manual
15-5
OLE Automation

Visual Basic example

MS-Excel contains Visual Basic for Applications (VBA), which allows for
scripting to control both the Excel application and other applications through
OLE Automation. The following is the Visual Basic source that calls the DLL
functions and displays the results in a spreadsheet. The subroutine
GetBaanERPUsers is an MS-Excel macro that can be invoked through buttons or
menus on the spreadsheet.
Dim user As String
Dim BaanObject As Object

Sub GetBaanERPUsers()
On Error GoTo CannotConnectToBaanERP

’ run BaanERP
Set BaanObject = CreateObject("Baan.Application")

On Error GoTo BaanAutomationError

’ get first user

BaanObject.ParseExecFunction "demodll",
"get_first_user()"
If (BaanObject.Error <> 0) Then GoTo
BaanAutomationError
user = BaanObject.ReturnValue

Row = 1
Column = 1
While (user <> "")
’ fill the spreadsheet
Worksheets("Main Sheet").Cells(Row, Column)
= user
Row = Row + 1
If Row > 15 Then
Row = 1
Column = Column + 2
End If
’ get next user
BaanObject.ParseExecFunction "demodll"
"get_next_user()"
If (BaanObject.Error <> 0) Then GoTo
BaanAutomationError

Technical Manual
15-6
 OLE Automation

user = BaanObject.ReturnValue
Wend

’ exit BaanERP
Set BaanObject = Nothing
Exit Sub

’ error handling
CannotConnectToBaanERP:
MsgBox "Unable to connect to BaanERP"
Exit Sub

BaanAutomationError:
MsgBox "A BaanERP automation error occurred"
Set BaanObject = Nothing
Exit Sub

End Sub

Example: Using BaanERP SQL

The following example fills an MS-Excel spreadsheet using BaanERP SQL. To
implement this, a DLL called ottdllsql_query has been developed to parse and
execute the query. To receive more information about this library, enter the
following command:
*> bic_info6.2 –eu ottdllsql_query

DLL information
The DLL ottdllsql_query contains:
n Functions to parse, execute, and stop a query:
− olesql_parse
− olesql_fetch
− olesql_break
− olesql_close
n Functions to import query results into Visual Basic:
− olesql_getstring
− olesql_getint
− etc

Technical Manual
15-7
OLE Automation

n A function to convert an Easy SQL query to a string:

− easysql_to_olesql
n Declaration of external variables, such as:
− olesql_count
− olesql_float0, olesql_float1 ... olesql_float9 (double variables)
− olesql_int0, olesql_int1 ... olesql_int9 (long variables)
An MS-Excel example is installed in the SAMPLES directory of BW
(BAAN.XLS). This spreadsheet contains several macros.
The first macro, GetBaanUsers, shows a number of BaanERP users on the
spreadsheet Worksheet Users. To use it, press the button on the Worksheet Users
that starts this macro. For more information refer to the Visual Basic code in
Worksheet Module 1.
The second macro, GetItemCount, shows all item groups in BaanERP with the
numbers of items per item group on the spreadsheet Worksheet Itemcount. Press
the button on the Worksheet Itemcount to start this macro. To run the macro you
have to insert an Easy SQL query in the BaanERP Application. To insert the
query, start the Query Data (ttadv3580m000) session, insert a record named
item and choose Maintain Query by Text Manager. In the text editor enter the
following query:
select tiitm001.citg, | Item group
count(tiitm001.item) | Item
from tiitm001 | Item data
group by tiitm001.citg
order by tiitm001.citg

For more information refer to the Visual Basic code.

Technical Manual
15-8