Open Source and Open Data

BOSA.be
November 2020 – AFC Brussels
Bart Hanssens
BOSA DG Digital Transformation
Open Source &
Open Data

Proposed agenda
 Introduction
 Open Source
 Licenses, business models …
 Break
 Open Data
 Challenges, community, …
 Q & A

FPS BOSA DG Digital Transformation
 Formerly: Fedict
 Providing services to other federal public services
 Secure exchange of data between administrations
 Federal Authentication Service (FAS)
 Web hosting platform
 https://dt.bosa.be/en

Setting the scene
 Workshop expectations ?
 Who is already familiar with open source ?
 As a user / as a developer
 Who is already familiar with open data ?
 As a user / as a developer

What is Open Source ?
 Open source / FLOSS
 Versus proprietary
 Open projects: more than just free code
 Questions / support ?
 Dealing with 3rd party contributions
 “Open” development: issue tracker, roadmap…

Picking “healthy” projects
 Is the code itself actively maintained ?
 E.g. releases, changes, code commits
 How active is the community (if there is one) ?
 Activity in mailing list / issue tracker
 How many different contributors ?
 (References to users / other projects) ?

Open source projects are not that different
 Which programming languages / frameworks ?
 Almost any (Java, .NET, Swift, Python, NodeJS, …)
 Open source still needs
 Documentation
 Tests
 Release notes

Differences and similarities
 Not all licenses can be combined !
 What they have in common
 No guarantees, non-exclusive, (mostly) irrevocable
 Right to distribute
 Can be used commercially
 Important differences between licenses
 Patent usage protection or not
 Distribution of source code or not
 License(s) of software combinations

Types of licenses
 “Strong” copy-left licenses
 GPL, EUPL, Affero GPL, …
 “Weak” copy-left
 LGPL, MPL, …
 Permissive licenses
 BSD, Apache, MIT, EPL …
 https://choosealicense.com/appendix/
 https://opensource.org/licenses

Strong copy-left example: GNU/Linux
 GPL kernel + GNU “userland” applications
 Lots of additional kernel features developed
 Different filesystems, firewall, network stack, …
 Proprietary drivers possible, but distributed separately

Weak copy-left example: eID middleware
 Belgian eID card middleware
 https://github.com/fedict/eid-mw
 Lesser GPL (LGPL)
 Can be used by proprietary applications

Permissive example: MacOS / iOS
 Mix of proprietary and open source parts
 https://developer.apple.com/opensource/
 Darwin “core OS”
 BSD licensed parts
 Apple open source license ASPL
 Other parts and applications
 Some Apache licensed parts

Code sharing platforms and communities

Code sharing / versioning
 Focus on sharing code, packaging
 Often include issue tracking, build services
 Free and commercial plans
 Github, Codeberg, SourceForge, …

Foundations and communities
 Focus on process, consistency, license compatibility
 Often NPO with elected board members
 Use or provide own code sharing platform
 Apache Foundation, Eclipse Foundation, Drupal.org…

Communication is key
 Open source projects are international
 Users may not be native (English) speakers
 Different cultures: direct vs indirect communication
 Often driven by volunteers
 They are not (free) employees

Example: RDF4J project board on github

Dual-licensing
 Multiple licenses for the same code
 Only the copyright holder can do this
 E.g. GPL for use in other open source projects…
 … and commercial license for “closed” source
 Examples:
 MySQL Database: GPL and Oracle license
 iText PDF library: AGPL and proprietary license

Additional services or modules
 Training, extra components, integrations …
 Odoo, Alfresco, Pentaho, …
 Providing “stable” combinations, long-term support
 RedHat Enterprise Linux, Oracle JDK, …

Cost savings / synergies
 Sharing cost of development and maintenance
 Companies contributing to tools they use themselves
 Android, Python, Chromium, Eclipse IDE …
 When software is not the core business
 Selling hardware or (more profitable) services
 Network drivers, Kubernetes …

For marketing, fun and/or for the greater good
 Software developed by/for universities, governments
 Paid with taxpayers’ money
 E.g BSD, Decidim, Accumulo …
 Community projects
 For fun and glory
 E.g. GIMP, PostgreSQL, VLC, …
 Some “fun” projects became quite big and profitable
 E.g. Linux, PHP, Drupal, …

Miscellaneous links on open source

Links
 EC Open Source Software Strategy
 https://ec.europa.eu/info/departments/informatics/ope
n-source-software-strategy_en
 “Blue hats” community (French gov community)
 https://www.modernisation.gouv.fr/le-hub-des-
communautes/blue-hats
 Annual FOSDEM conference
 https://fosdem.org

What is open data ?
 Free, easy to reuse, non-sensitive data
 No personal data (GDPR !)
 Download files or webservice / API
 Who publishes open data ?
 Governments (local /regional / federal)
 Companies (voluntary basis)
 Citizens / crowd-sourcing projects (voluntary basis)

Legal framework
 EU Public Sector Information / Open Data Directive
 https://ec.europa.eu/digital-single-market/en/european-
legislation-reuse-public-sector-information
 Transposed into federal law and regional decrees

What kind of data ? Almost any !
 A lot of statistics
 Population, consumer price index, average income …
 Transport and mobility
 Train delays, parking spots, …
 Environment
 Air quality, biodiversity, …
 Geospatial data
 Satellite data, maps, …

Formats
 No single format for each and every type of data
 Depends on application / systems / conventions
 Preference for open, structured file formats
 CSV, GTFS, (Geo)JSON, KML, ODS, sqlite, XLSX, XML …
 (RDF / Linked Data)
 Known / documented APIs
 OGC GIS, OpenAPI/Swagger, SPARQL …
 Availability and meaning often more important

Quick tour of licenses
 Well-known licenses
 Various Creative Commons, Open Data Commons
 https://creativecommons.org/about/cclicenses/
 https://opendatacommons.org/licenses/
 Legal situation in Belgium
 Status: it’s complicated (federated state)
 Quite a few slightly different custom licenses …
 … many of them are very similar to CC-licenses

Differences and similarities
 Not all licenses can be combined !
 What they have in common
 No guarantees, non-exclusive, (mostly) irrevocable
 Right to distribute
 Important differences between licenses
 Can be used commercially or not
 Attribution required or not
 License of combined datasets

Additional services and data
 Combining and transforming data
 Providing Service Level Agreements
 Additional services
 E.g. user-friendly search, visualizations, reports…
 Examples
 Law: https://lex.be , https://jura.kluwer.be
 Companies: https://opencorporates.com

Example: MapTiler with OpenStreetMap data

Where to find open data ?
 Websites and (open data) portals
 https://data.gov.be
 If you can’t find it… just ask !
 opendata@belgium.be
 @BartHanssens
 #data.gov.be:matrix.org

Data.gov.be and the Open Data Task Force
 Open Data Task Force: FPS BOSA DG DT + DAV/ASA
 Portal is “by-product” of full (meta)data export
 See also https://github.com/fedict/dcat
 “No Wrong Door” principle
 Forward questions to Regions / Cities if needed

Combining open source and open data

Open Summer of Code
 Organized by OpenKnowledge Belgium
 + partners from government and private sector
 https://summerofcode.be/
 Students get paid to build open source projects
 Often projects which also use open data
 Mostly IT, but also business, design, marketing, …
 3 x 4 day in July
 With coaches, presentations on soft skills, …

Example: BeST Addresses
 Addresses + geo location from the 3 Regions
 Published and updated weekly by BOSA DG DT
 https://opendata.bosa.be/
 2019 oSoC project
 Python tools developed by students
 Data also available on https://openaddresses.io
 https://github.com/oSoc19/best

Open data communities
 OpenKnowledge Belgium
 Local “chapter” of OpenKnowledge
 https://be.okfn.org
 OSGeo Belgium
 Open data and open source
 https://www.osgeo.org/local-chapters/osgeo-belgium/
 Open Justice
 https://openjustice.be

Data-driven projects
 Mozilla CommonVoice
 https://commonvoice.mozilla.org/en/about
 Telraam traffic counters
 https://telraam.net/en
 Global low-power data transfer / IoT
 https://www.thethingsnetwork.org

Share your open data stories !
 Are you working on a thesis / paper on open data ?
 Did you create a stunning app / service ?
 Are you organizing an open data-driven event ?
 Let us know, so we can:
 Tweet and make some fuzz
 Promote your work
 Convince more organizations to open data

BOSA.be
@BartHanssens
https://data.gov.be
opendata@belgium.be
#data.gov.be:matrix.org
Thank you !

Open Source and Open Data

More Related Content

What's hot(19)

Similar to Open Source and Open Data(20)

More from Bart Hanssens(20)

Recently uploaded(20)

Open Source and Open Data