Activity
From 05/23/2020 to 05/29/2020
05/29/2020
-
11:42 PM Bug #13882: Exception in `ensure` stops threads from exiting
- This bug still seems to be affecting all more recent versions of Ruby up to and including 2.7.1.
I encountered this in the `newrelic_rpm` gem which was intermittently hanging on shutdown in a short-lived script. It turns out that if R... -
10:59 PM Bug #16439 (Closed): Resolv::DNS: timeouts if multiple IPv6 name servers are given and address contains leading zero
- Applied in changeset commit:git|04b5203a031d372b725e407519f10da6deda0e78.
----------
Convert ip addresses to canonical form in Resolv::DNS::Requester::UnconnectedUDP#sender
Otherwise, if the IP address given is not in canonical form, i... -
09:19 PM Bug #16439: Resolv::DNS: timeouts if multiple IPv6 name servers are given and address contains leading zero
- This isn't just a problem for IPv6, it happens for IPv4 as well (e.g. 127.0.0.01). It will happen in any case where the input is not in canonical form, where canonical form is the form that `Addrinfo#ip_address` gives you. I've added a...
-
10:59 PM Revision 9b8825b6 (git): Update NEWS / documentation with GC.start(compact:true)
-
- Otherwise, if the IP address given is not in canonical form, it
won't match, and Resolv will ignore it.
Fixes [Bug #16439] -
- I propose to make `_1 = 1` a SyntaxError, instead of emitting a warning (it emits a warning in 2.7). I've implemented this proposal in a pull request: https://github.com/ruby/ruby/pull/3163
-
- Assigning to the numbered parameter is now a SyntaxError even if it is a local variable in scope, so current master has your expected behavior.
Related to this, accessing the numbered parameter inside a block in master accesses the nu... -
- It's currently expected that Marshal cannot dump objects larger than 2GiB, so this isn't a bug, though arguably RangeError would be more appropriate than TypeError if the data is too large. Supporting the dumping of larger objects does ...
-
- This commit combines the sweep step with moving objects. With this
commit, we can do:
```ruby
GC.start(compact: true)
```
This code will do the following 3 steps:
1. Fully mark the heap
2. Sweep + Move objects
3. Update references
B... -
-
- This crashes inside the pg gem: `https://github.com/ged/ruby-pg/blob/v1.1.4/ext/pg_connection.c#L3221-L3240`. My guess is you are not disconnecting before forking and you end up sharing connections between processes, which can easily ca...
-
- The current RDoc behavior is by design, so this isn't a bug, it's a feature request.
I agree it would be best for RDoc to reflect the reality that Kernel methods are defined in Kernel and not object, because if they were defined in Ob... -
-
-
- Looking at my include directories, I see directories such as:
```
/path/to/include/ruby-2.7/ruby-2.6/x86_64-openbsd
/path/to/include/ruby-2.7/ruby-2.7/x86_64-openbsd
```
but nothing similar for 2.8 even though 2.8 is also instal... -
- Even though this appears to be a bug, it's actually intentional behavior, introduced in commit:5208c431bef3240eb251f5da23723b324431a98e, and there are tests and specs for it. So changing this behavior would be a feature request.
I th... -
08:17 PM Misc #16911: Travis: unstable non-x86_64 jobs
- Related to the unstable non-x86_64 jobs in Travis, I am asking Drone CI to support ppc64le/s390x here.
https://discourse.drone.io/t/supporting-ppc64le-and-s390x/7385/
I understand Ruby project does not use Drone CI, as we need to use... -
- Thanks for the info! I will try the environment variable, and will let you know here.
-
12:22 PM Bug #16921: s390x: random test failures for timeout or segmentation fault
- Try an environment variable `RUBY_TEST_TIMEOUT_SCALE`. If you set `RUBY_TEST_TIMEOUT_SCALE=10`, all timeout tests wait 10 times longer.
-
- I observed the following error related to timeout or segmentation fault on Ruby 2.7.1 s390x in Fedora builds. The tests somtimes fail. It seems the s390x machine used in CI is slow. It's great if we can change the timeout casually if it ...
-
06:53 PM Revision c1f6552b (git): [ci skip] Enhanced Rdoc for Hash (#3155)
- * Enhanced Rdoc for Hash
* Respond to review -
06:19 PM Bug #16922 (Closed): single quote should be escaped in xml
- This is Lisbeth from Taiwan. I'm trying to convert a string to xml by using 'string'.encode(xml: :attr) but single quote doesn't escape as expected
Here's the spec shows which character needs to be escaped: https://www.w3.org/TR/xml/#dt... -
05:52 PM Feature #16597: missing poll()
- rebased
-
05:03 PM Bug #16919 (Assigned): IRB でマジックコメントの : の後に任意の文字を入力するとエラーが発生する
-
09:12 AM
Bug #16919 (Closed): IRB でマジックコメントの : の後に任意の文字を入力するとエラーが発生する
- # 再現手順
``` ruby
irb(main):001:0> #coding:u
```
# 実際に起きた結果
```
irb(main):001:0> #coding:Traceback (most recent call last):
38: from /usr/local/opt/ruby/bin/irb:23:in `<main>'
37: from /usr/local/opt/ruby/bin/irb:23:in `load'
3... -
- I took another look at this and I think we can silence the deprecation warnings with `-W:no-deprecated` even if we can't deduplicate the warnings. Attached is a patch that implements this. I'm not sure if the approach in ruby.h is acce...
-
- > I have never seen the test fail on Debian/Ubuntu aarch64. So use Ubuntu! (joke)
I love Debian. :) It's my pleasure when I spend time with people in Debian community.
> ...
Yeah, it makes sense.
> Can you reproduce the issue w... -
- I have never seen the test fail on Debian/Ubuntu aarch64. So use Ubuntu! (joke)
It is difficult to investigate the issue with no access to the machine...
Can you reproduce the issue with ruby master? Can you try another Fedora+a... -
- I observed the following error on Ruby 2.7.1 building in Fedora aarch64 CI. It was just one time error. It does not always happen.
```
1) Error:
TestThread#test_signal_at_join:
Timeout::Error: execution of assert_separately expir... -
- WSL 2 is officially released. It uses Linux kernel, so almost all specs
for Linux work on WSL, except one: gethostbyaddr. I guess network
resolution in WSL is based on Windows, so the behavior seems a bit
different from normal Linux.
... -
04:15 PM Revision 0d1f0cd9 (git): runnable.rb: ignore broken symlink error on Windows [ci skip]
-
- `Kernel#.require` and `Kernel#.load` do not search also "explicit
relative path" files, not only absolute paths, in the load path. -
03:27 PM Feature #16832: Use #name rather than #inspect to build "uninitialized constant" error messages
- @Eregon good catch, I'll submit a PR with extra specs early next week and tag you.
-
10:13 AM Feature #16832: Use #name rather than #inspect to build "uninitialized constant" error messages
- FWIW it also applies to missing methods and other NameError, which sounds nice but would deserve a spec in `spec/ruby/core/exception/name_error_spec.rb`:
```
$ ruby -e 'c=Class.new { def self.name; "MyClass"; end }; c.foo'
master:
-... - 03:17 PM Revision 2c8821c0 (git): * 2020-05-30 [ci skip]
-
- getgrnam(3) says:
```
ERRORS
0 or ENOENT or ESRCH or EBADF or EPERM or ...
The given name or gid was not found.
```
Process::GID.from_name raises an ArgumentError for 0, but Errno::ESRCH
for ESRCH. Actually, WSL 2 ... -
-
- Applied in changeset commit:git|2ecfb88ee50510955acd3ae9fc94a5f109e7f109.
----------
Correctly remove temporary directory if path yielded is mutated
Another approach would be to freeze the string, but that could
cause backwards compati... -
04:59 AM Bug #16918: Dir.mktmpdir should yield a copy of the dir to protect cleanup
- Makes sense to me what headius wrote, so if there are no side effects (I have too
little experience with Dir.mktmpdir myself) +1 to the suggestion from me.
-
- I agree this is a bug and have added a pull request to fix it: https://github.com/ruby/ruby/pull/3159
-
01:14 AM Bug #16918 (Closed): Dir.mktmpdir should yield a copy of the dir to protect cleanup
- If you modify the dir string passed into the block from `Dir.mktmpdir`, the logic to clean up the temporary directory may fail:
```
$ rvm ruby-2.6.5 do ruby -rtmpdir -e "Dir.mktmpdir('foo') {|dir| dir << 'bar'}"
Traceback (most rece... -
- Another approach would be to freeze the string, but that could
cause backwards compatibility issues.
Fixes [Bug #16918] -
05:45 AM Revision 0bd025ad (git): Add a debug_counter for JIT cancel on leave
-
04:50 AM Revision d300a734 (git): RUBY_ASSERT_NOASSUME
- If __builtin_assume() is enables and RUBY_DEBUG=0, RUBY_ASSERT(expr)
will be compiled to __builtin_assume(expr) and compiler can assume
expr is true and apply aggressive optimizations. However we observed
doubtful behavior because of com... -
03:39 AM Feature #16786: Light-weight scheduler for improved concurrency.
- What about borrowing a little from Crystal? The non-blocking API could be:
``` ruby
spawn do
# non-blocking fiber
end
```
I like how it has a completely different interface without introducing a new term. -
-
-
05/28/2020
-
- I've added a pull request that fixes this issue: https://github.com/ruby/ruby/pull/3157
It's not a perfect fix, but a perfect fix would require disabling a commonly used optimization, resulting in additional array allocations in commo... -
- Extract a function to free the moved list. We'll use this function
later on to compact at the same time as sweep. -
-
- I don't think this is a bug fix, this just makes a different tradeoff. This can help cases where the computer you are connecting to is overloaded by TIME_WAIT sockets. However, by the same token, it causes TIME_WAIT sockets to accumula...
-
- Eregon (Benoit Daloze) wrote in #note-10:
> @jeremyevans0 or @nobu Could you implement it on `master`?
I've implemented basic support, which passes make check: https://github.com/jeremyevans/ruby/commit/672901facca6f88ae40b4ea8ef59ef... -
- Interesting, I never saw the reply above from matz.
Looking at my email I received one but it wasn't attached to the rest of the conversation for this issue for some reason.
Great to hear it's now accepted.
I think we should have ... - 05:45 PM Revision fd7849ce (git): * 2020-05-29 [ci skip]
-
- And fixed typo in compilers.yml.
-
- Applied in changeset commit:git|9a3371be8f5274d1dca64b93c7b9a379e32ea2f5.
----------
Check that BASERUBY is at least Ruby 2.2 in configure
BASERUBY needs to be at least Ruby 2.2 since
46acd0075d80c2f886498f089fde1e9d795d50c4.
I think ... -
- BASERUBY now requires at least Ruby 2.2, so there is no point
trying to support older ruby versions here. -
- BASERUBY needs to be at least Ruby 2.2 since
46acd0075d80c2f886498f089fde1e9d795d50c4.
I think it's better to explicitly fail early as soon as BASERUBY
is used in this case, versus trying to debug later failures.
This modifies things t... -
- Regarding delegation, I agree we need to support at least leading arguments together with `...`.
I filed #16378 6 months ago (before the 2.7 release), it was rejected and then later accepted.
I think we should backport that to 2.7 as i... -
11:06 AM Feature #16891: Restore Positional Argument to Keyword Conversion
- I like the current plan for keyword argument separation. Promotion of hash to keywords is very confusing and cause many problems designing clean interfaces in Ruby. Kudos to @jeremyevans0 and everyone else here for providing calm and rat...
-
-
09:47 AM Bug #16917: "ruby -run -e httpd . " did not work
- works for me as well on fresh 2.7.1 installation (compiled locally)
-
- It works for me. Could you show us the terminal log and the html result (if you can connect to the server)?
-
03:05 AM Bug #16917 (Third Party's Issue): "ruby -run -e httpd . " did not work
ruby -run -e httpd . -p 8080
no file list show
only show HelloWorld .
-
05:41 AM Revision 7473b4c1 (git): Explicitly loading with envutil.rb
-
- Coverage measurement had failed since 7d5da30c9e9c572f6ef0aaccc1ca0e724966e2ee
https://github.com/ruby/actions/runs/708557030?check_suite_focus=true
05/27/2020
-
- byroot (Jean Boussier) wrote in #note-21:
> I'm afraid that's a dead link.
I fixed the link on Redmine. -
- > `$LOAD_PATH.unshift(dir) unless $LOAD_PATH.include?(dir)`
That specific case would degrade properly. Assuming the `include?` returns `false` because that path is now handled by Rubygems/Bundler, it simply mean you'll disable caching... -
04:16 PM Feature #16848: Allow callables in $LOAD_PATH
- > not sure what you mean by that. Like right now you can ` $LOAD_PATH.unshift(Object.new)` and it won't break `$LOAD_PATH.include?(path)`.
If you look in existing gems you'll find lots of stuff like `$LOAD_PATH.unshift(dir) unless $LOAD... -
- I agree with @Eregon that we can probably change the return value of ConditionVariable#wait in Ruby 3. As it always returns true currently, it's worthless and nothing should be relying on it.
I can see that changing `Mutex#sleep` to ... -
- I've looked into this more. The underlying cause is that Resolv expects nameservers to be specified as IP addresses, not as domain names. When using multiple nameservers, Resolv tries to match the IP address of the nameserver respondin...
-
- I think it's worth mentioning nobu's comment from the dev meeting:
> have wanted to deprecate that behavior for years, and made non-nil $/ warned.
So what about finally deprecating this behavior? If the incompatibility is too bad it'... -
01:34 PM Feature #15771: Add `String#split` option to set `split_type string` with a single space separator
- @sawa, for your "use-case", `line.split(/ /, 2)` is far better than `line.split(" ", 2, literal: true)`, I think, no matter what keyword we'd choose.
+ shorter
+ clearer
+ need to rewrite anyway
Matz.
-
04:59 AM Feature #15771: Add `String#split` option to set `split_type string` with a single space separator
- I have a particular use case. I was creating a file that describes UTF-8 characters, which included lines like this:
```
002 !"#$%&'()*+,-./
003 0123456789:;<=>?
004 @ABCDEFGHIJKLMNO
```
The first three characters of each line descri... -
- Dan0042 (Daniel DeLorme) wrote in #note-10:
> That optimization is nice to have, but I think the point of this ticket is that it's currently not possible to have an arbitrary string separator.
I agree.
Dan0042 (Daniel DeLorme) wrote in... -
- This doesn't seem to be a bug, it is by design. Trying to move all syntax errors into the parser is probably too difficult to justify the effort even if it is possible.
If you want to more fully check for valid syntax without having ... -
- Applied in changeset commit:git|fa380208fe4ef36cd6a8535d3ac43e61c7935f1c.
----------
Improve documentation for Net::HTTPHeader#set_form [ci skip]
Fixes [Misc #16916] -
- Fixes [Misc #16916]
-
- From the implementation linked above:
```ruby
case [0, 1, 2]
in [*, a, *]
p a #=> 0; non-greedy match
end
case [0, 1, 2]
in [*, b, 2, *]
p b #=> 1; backtracking
end
```
So yes, it supports "backtracking" in a certai... -
- Accepted. But we have to clarify the term "backtracking" here. It does backtrack but not with a guard, right?
Matz.
-
- Suppose we have the following code:
```
json = {name: "Alice", children: [{name: "Bob", age: 6}, {name: "Chad", age: 4}]}
case json
in {name: "Alice", children: [{name: "Bob", age: age}]}
p age
end
```
It doesn't match because ... -
- Through trial and error, I've determined this is caused by the branchif/branchnil/branchunless optimization, self described as a super nasty hack. I've submitted a pull request to disable this optimization (https://github.com/ruby/ruby/...
-
05:20 PM Feature #12901: Anonymous functions without scope lookup overhead
- Can we visit this again which if we go with 1:) add a new construct we could solve https://bugs.ruby-lang.org/issues/11630 - serializing functions
also, if we add new constructs we can add a multi-method version - https://bugs.ruby-lang... - 03:15 PM Revision 9cfef5b4 (git): * 2020-05-28 [ci skip]
-
- At the time of clean-platform, platform directory may not be empty
because of parallel build. -
-
- * Enhanced Rdoc for Hash
* Respond to review
* Nudge CI testing.
Respond to review -
-
- > We'd rather improve Bootsnap if they need support from the core.
Ok. Then we can close this issue.
That part of Bootsnap doesn't need any support from core, it works fine except for what I explained, it takes the blame for many i... -
- I am negative. Cacheing compiled binary caused a lot of problems in Python, especially with multiple versions installed. We'd rather improve Bootsnap if they need support from the core.
Matz.
-
- and mark it as COLDFUN on JIT, assuming enqueue usually happens on VM.
-
06:48 AM Revision a3f498e4 (git): Fix max, min, minmax documentation (#3131)
- They only need that all objects implement <=>,
but the documentation said it needs Comparable. -
-
-
- to reduce code size and improve locality of hot code.
-
- by calling combined functions specialized for each cancel type.
I'm hoping to improve locality of hot code, but this patch's impact should
be insignificant. -
-
-
- It has never been created by default, and no reason to keep binary
compatibility against extension libraries for ruby 1.8.
1. `LIBRUBY_SO` on Cygwin was named originally.
- 94089e82a12:: `$(RUBY_INSTALL_NAME)-cygwin.dll`
- f0ba57... -
02:26 AM Revision 0fcb2dd5 (git): add static modifier for rb_str_ord func
05/26/2020
-
- I get the point, but we still need a concrete use-case. (Unlike tabs and commas) Space-separated CSV is not common, and consequent spaces are considered as one space usually. I still feel like it's a theoretical concern.
Besides that,... -
- Applied in changeset commit:git|e0b40ef5d8173aff304c81f93516e1246e3c042c.
----------
Don't convert bundler man pages from mdoc to man
These man pages are already in man format and assuming they are
mdoc format breaks things.
Fixes [Bu... -
- I've confirmed this bug and submitted a pull request to fix it: https://github.com/ruby/ruby/pull/3147
-
- These man pages are already in man format and assuming they are
mdoc format breaks things.
Fixes [Bug #16823] -
- This is due to a bug in the optimizer. Adding `return COMPILE_OK;` to the top of `iseq_peephole_optimize` fixes the problem.
Instructions before (optimization/broken):
```
== disasm: #<ISeq:<main>@-e:1 (1,0)-(1,52)> (catch: FALSE... -
10:35 PM Misc #16916 (Closed): [Documentation] Method: Net::HTTPHeader#set_form
- Since this revision / commit
- SVN Revision 65230
- https://bugs.ruby-lang.org/projects/ruby-master/repository/ruby_2_6/revisions/65230
- https://svn.ruby-lang.org/cgi-bin/viewvc.cgi/trunk/lib/net/http.rb?view=log&pathrev=65... -
- I merged this patch upstream: https://github.com/ruby/webrick/pull/42
-
-
-
-
-
- and /dev/urandom intentionally. OpenSSL::PKey::RSA.new opens the two
random generators and keeps the file descriptors.
https://github.com/openssl/openssl/blob/93f99b681ab5a1cf7062053323e09b0cad5ff854/crypto/rand/rand_unix.c#L674
They ... -
- 03:28 PM Revision f8e703e2 (git): * 2020-05-27 [ci skip]
-
- Attempting to invoke a non-existent command led to the leak of fds.
http://rubyci.s3.amazonaws.com/graviton2/ruby-master/log/20200526T140004Z.log.html.gz
```
[ 9101/20195] TestGemExtCmakeBuilder#test_self_build = 0.01 sLeaked file descr... -
- > the "vast majority of the cases" involve something like $LOAD_PATH.include?(path) and string duck-typing doesn't help with that.
@Dan0042 not sure what you mean by that. Like right now you can ` $LOAD_PATH.unshift(Object.new)` and ... -
- > $LOAD_PATH as a "view"
This is what I proposed at comment #16.
Maybe there are two strategies:
(1) extend $LOAD_PATH.
(2) introduce $LOAD_PATH2 and deprecate $LOAD_PATH.
(3) other great idea?
Several years ago I tried this topic wh... -
- > ### String duck-typing
It seems to me that apart from adding a path to $LOAD_PATH (which is a non-issue), the "vast majority of the cases" involve something like `$LOAD_PATH.include?(path)` and string duck-typing doesn't help with t... -
- Sorry I missed your answer from 5 days ago. I wish I'd answered before your second meeting (Redmine no longer send notifications...).
> Could you propose another proposal to care compatibility?
To be honest I kinda expected this co... -
- As I described, the current approach has a compatibility issue.
Could you propose another proposal to care compatibility?
In discussion at devmeeting, there is possibility to introduce a feature to help `$LOAD_PATH` observing which b... -
- Don't mind, I was not clear.
-
11:29 AM Feature #8661: Add option to print backtrace in reverse order (stack frames first and error last)
- mame (Yusuke Endoh) wrote in #note-43:
> The order had been already reverted in #note-31.
Ah, thx and sorry to miss that :blush: -
- The order had been already reverted in #note-31.
-
- Have the commit changed the order or introduced just the option? It seems to that just the later, which is not what was agreed in #note-27
-
- Committed at 39365b46e250162f278cb36aa148bc2a92b1b84a. Thanks!
-
-
-
- Add GH actions with -DRUBY_DEBUG
-
-
09:45 AM Bug #16914 (Closed): Hash.new with `default_proc` returns this `default_proc` instead of calling it sometimes
- Hello,
First of all, I want to apologize because this bug report will not be realized at all in good shape. I encounter this problem on a rails application currently in production, but I am unable to reproduce it. I failed to provide ... -
-
-
-
-
- Applied in changeset commit:git|7d5da30c9e9c572f6ef0aaccc1ca0e724966e2ee.
----------
Test for [Feature #16832] -
-
-
- Applied in changeset commit:git|4e1f2283b432e833bd4e6f7724ba0496760e68e8.
----------
Make Thread#thread_variable? similar to #thread_variable_get
Don't use rb_check_id, which only works for pinned symbols.
Switch inadvertent creation t... -
- That approach didn't work as there are tests for inadvertent symbol creation. To fix that, we need to force ID creation in thread_local_set. I've updated the pull request to use that approach, and made thread_variable_get use the same i...
-
- I can confirm this bug. It's because `thread_variable?` uses different code than `thread_variable_get` and `thread_variable_set`. Removing the `rb_check_id` and switching `ID2SYM` to `rb_to_symbol` fixes it. I've added a pull request ...
-
- Don't use rb_check_id, which only works for pinned symbols.
Switch inadvertent creation test for thread_variable? to
only check for pinned symbols, same as thread_variable_get
and thread_variable_set.
Make key variable name in thread_lo... -
02:59 AM Feature #16824: Follow RubyGems naming conventions for the stdlib
- nobu (Nobuyoshi Nakada) wrote in #note-13:
> Nonsense.
Nobu, I feel deeply offended by your impoliteness and lack of respect.
yuki24 (Yuki Nishijima) wrote in #note-14:
> ...
This is great, well done! -
02:20 AM Feature #16824: Follow RubyGems naming conventions for the stdlib
- While I understand the intention of this request, I'm not sure this would be practically useful for us. As mentioned above, there are definitely gem names that people are already too familiar with, that we probably don't want to rename, ...
-
- Nonsense.
-
- Let's not close this because the original request is not be accepted at 100%
Any volunteer to double-check the list and figure out which gem names are available and which could possibly be reclaimed?
I can help to register the name... -
- Applied in changeset commit:git|faab5cbeb762adec4983d603d71e72add5108249.
----------
Fix documentation for Fiber#raise [ci skip]
Fixes [Bug #16912] -
- Fixes [Bug #16912]
05/25/2020
-
11:33 PM Feature #16913 (Open): Add `ARGF#each_io`
- Add an iterator for each file supplied on the command line, or STDIN. `ARGF#each_io`
## Current Status
Often, we need to do something with individual files ARGF knows about rather than the concatenation of them. We can combine `ARG... -
- 21991e6ca5 enables `__builtin_assume()` for clang and it seems
skip SEGV on rb_class_of() with unexpected value. This test expects
a [BUG] output, so this patch causes [BUG] to show [BUG] message.
https://github.com/ruby/ruby/runs/70708... -
- Applied in changeset commit:git|21991e6ca59274e41a472b5256bd3245f6596c90.
----------
Use RUBY_DEBUG instead of NDEBUG
Assertions in header files slows down an interpreter, so they should be
turned off by default (simple `make`). To ena... - 06:21 PM Revision 7bafc9f4 (git): * 2020-05-26 [ci skip]
-
- Co-authored-by: 卜部昌平 <[email protected]>
-
- Assertions in header files slows down an interpreter, so they should be
turned off by default (simple `make`). To enable them, define a macro
`RUBY_DEBUG=1` (e.g. `make cppflags=-DRUBY_DEBUG` or use `#define` at
the very beggining of the... -
- Fiber#raise has a mismatch between implementation and documentation for "created" fiber.
RDoc:
> Raises an exception in the fiber at the point at which the last Fiber.yield was called, or at the start if neither resume nor rais... -
- I would like to open this ticket to track and fix the issues for Travis non-x86_64 jobs, as I saw the following commits now.
Travis s390x-linux is too unstable
https://github.com/ruby/ruby/commit/f72bec2cea313f7a72a6f907f3dd64f766e91... -
- ```
Failures:
1) The library itself does not contain any warnings
Failure/Error: raise "Invoking #{method}!(#{args.map(&:inspect).join(", ")}) failed:\n#{last_command.stdboth}"
RuntimeError:
Invoking sys_exec!("ruby ... -
- ```
1) bundle version with version outputs the version with build metadata
Failure/Error: expect(out).to match(/\ABundler version #{Regexp.escape(Bundler::VERSION)} \(\d{4}-\d{2}-\d{2} commit #{COMMIT_HASH}\)\z/)
expected ... -
- https://github.com/ruby/actions/runs/703745101?check_suite_focus=true#step:16:27
```
An error occurred in a `before(:suite)` hook.
Failure/Error: contents = File.read(version_file)
Errno::ENOENT:
No such file or directory @ rb_sysopen... -
-
- @odaira Can you confirm/are you okay for your ppcle64 implementation to be released under the MIT License?
-
- Can you check if this is still a problem, I merged my changes which should make this test more reliable. But I did not fix any underlying problems.
-
- 01:19 AM Revision 089dcfbf (git): * 2020-05-25 [ci skip]
-
01:17 AM Revision 06321ea2 (git): RBIMPL_ATTR_DEPRECATED: disable for GCC10
- GCC 10.x seems buggy here. See also https://gcc.gnu.org/bugzilla/show_bug.cgi?id=95302
-
01:09 AM Feature #15277: at_exec
- I would like to add a use-case that I'm encountering right now.
I would like to call prctl(PR_SET_PDEATHSIG, ...) in order to ensure that a spawned process is always killed regardless of how cleanly the parent process exits. I couldn'...
05/24/2020
-
- There was merged this [1] PR, which should resolve the issue even without removal of the `gem` call. I have not verified it myself, but the test case was modeled after my test case in comment #note-10, so I am positive about its function...
-
04:34 PM Misc #16910: BasicObject is resolved in BasicObject
- Ah! I read too fast too :). Appreciate the followup!
-
- Actually I looked too fast, the correct location is
https://github.com/ruby/ruby/blob/b10c9d201222b144df7d63660d1c731af53c4ae2/class.c#L551-L559
So it's defined as a constant of itself (BasicObject) and not Object because Object is n... -
-
- This can be closed (can't do it myself apparently).
-
- Ahhh, that explains it. That code says this is intentional and definitely not a bug.
Thanks a lot Benoit!
-
- As you found, I think it's just a special case that BasicObject has a constant named "BasicObject".
It's defined here in CRuby:
https://github.com/ruby/ruby/blob/b10c9d201222b144df7d63660d1c731af53c4ae2/class.c#L569
And here in TruffleR... -
- Ah, forgot to add that I saw this
```ruby
BasicObject.constants # => [:BasicObject]
```
Could it be the case that `BasicObject` is special-cased somehow? If so, where in the source code? -
- I categorized this as Misc because I do not really know if this is a bug, or perhaps I miss some special rule in the resolution of relative constants.
Basically, I would expect
```ruby
class C < BasicObject
BasicObject
end
``... -
02:47 PM Revision c6c02331 (git): [ruby/rdoc] Check uninitialized instance variable in test
- https://github.com/ruby/rdoc/commit/3dcd5ddbb6
-
02:47 PM Revision 3169ffc0 (git): [ruby/rdoc] Add frozen string literal support
- https://github.com/ruby/rdoc/commit/daac9d0471
-
- https://github.com/ruby/rdoc/commit/9b1da32254
-
- https://github.com/ruby/rdoc/commit/a13d6439da
-
- https://github.com/ruby/rdoc/commit/82ff37a822
-
- Previously, only calls to rb_define_alias were treated as aliases.
This treats calls to rb_define_method with the same C function as
aliases, with the first function defined being the primary method.
This move the dedup code from the C ... -
- The crossref must be linked before tidylink because Klass.method[:sym] will be
processed as a tidylink first and will be broken.
https://github.com/ruby/rdoc/commit/0f47baf6d2 - 02:47 PM Revision c79f9ea6 (git): [ruby/rdoc] Escape method names in HTML
- The following is invalid HTML:
<a href="Array.html#method-i-3C-3C"> -
- Nice, in fact this was already the behavior in TruffleRuby and JRuby.
-
- This patch was merged in commit:adecd43197d5ea2a62a618a5c9be653bcf009c62.
-
- It seems working now, by compiling with gcc 10.1.0.
-
08:08 AM Bug #16905: Ruby required to build Ruby on Haiku?
- > parse.y file has been modified, or its timestamp changed at least.
I didn't edited that file and haven't changed the timestamp, so while your idea makes sense, i don't trust this error message at all, as this is from make, not from ... -
- I've added https://github.com/ruby/ruby/pull/3140 to fix this. It works by ensuring an origin for Enumerable directly after it is created. It does the same for Comparable and Kernel, since I think the issue could affect them.
-
-
-
- I would expect Hash#shift to return either a key-value tuple or nil. Returning the default value is, honestly, incomprehensible.
05/23/2020
- 07:01 PM Revision 0b161774 (git): * 2020-05-24 [ci skip]
-
- It loads `ext/fiddle/lib/fiddle/version.rb`, which causes constant
redefinition warning:
http://rubyci.s3.amazonaws.com/ubuntu2004/ruby-master/log/20200523T153003Z.log.html.gz
```
[ 6317/20193] TestDefaultGems#test_validate_gemspec/home... -
02:16 PM Bug #16787 (Closed): [patch] allow Dir.home to work for non-login procs when $HOME not set
- Applied in changeset commit:git|c15cddd1d515c5bd8dfe8fb2725e3f723aec63b8.
----------
Allow Dir.home to work for non-login procs when $HOME not set
Allow the 'Dir.home' method to reliably locate the user's home directory when
all three ... -
- Allow the 'Dir.home' method to reliably locate the user's home directory when
all three of the following are true at the same time:
1. Ruby is running on a Unix-like OS
2. The $HOME environment variable is not set
3. The pro... -
- Should prefer a prefixed version over the bare pkg-config for
cross compiling. -
- In other words, the current semantics of `return hash.default(nil) if hash.empty?` feel hacky and actually harmful to me.
The user probably never expects to have the default_proc called with a `nil` key in many cases. -
- Maybe `Hash#shift` should not call the `default_proc` or use `Hash#default`?
I.e., it would always return `nil` if `Hash#empty?`.
I think that would be more intuitive and probably compatible enough. -
- ioquatix (Samuel Williams) wrote in #note-2:
> @jeremyevans0 I agree with your assessment, however that hash does not have default_proc so I assume that it would return `nil` after all key-value pairs are `shift`ed out. That's different... -
- @jeremyevans0 I agree with your assessment, however that hash does not have default_proc so I assume that it would return `nil` after all key-value pairs are `shift`ed out. That's different from invoking `default_proc` and returning `[ni...
-
- While your particular example is non-intuitive, there is a simple explanation for it. The first time `hash.shift` is called, `hash` is empty, so it returns the default value (`0`). It gets the default value by calling the default_proc f...
-
- I don't have any strong opinion about this, but I observed the following behaviour which I thought was confusing. Maybe it's okay, or maybe we should change it to be more consistent.
```
hash = Hash.new{|k,v| k[v] = 0}
hash.shift ... -
- This reverts commit 6006ab8cf93a0e06d9d65213eff3117965c9bd0c.
AC_CHECK_TOOL does not have REJECT argument. -
- Applied in changeset commit:git|6006ab8cf93a0e06d9d65213eff3117965c9bd0c.
----------
Search pkg-config by AC_CHECK_TOOL [Bug #16909]
It should use AC_CHECK_TOOL to automatically search for a prefixed
version and not just the bare pkg-c... -
04:28 AM Bug #16909 (Closed): pkg-config search should use AC_CHECK_TOOL instead of AC_CHECK_PROG
- pkg-config is a toolchain variable akin to the compiler/linker/etc... it should use AC_CHECK_TOOL to automatically search for a prefixed version and not just the bare pkg-config.
-
- It should use AC_CHECK_TOOL to automatically search for a prefixed
version and not just the bare pkg-config. -
-
-
-
05:34 AM Revision 3015a7aa (git): [ruby/fiddle] Improve documentation on how to correctly free memory and free memory in tests (#33)
- https://github.com/ruby/fiddle/commit/e59cfd708a
- 05:29 AM Revision 24b615e8 (git): [ruby/fiddle] Export Fiddle::VERSION
- https://github.com/ruby/fiddle/commit/1b93a2d9db
- 05:29 AM Revision 74a4558c (git): [ruby/fiddle] Update file list
- https://github.com/ruby/fiddle/commit/b04cb92d7b
- 05:29 AM Revision d5c0018a (git): [ruby/fiddle] Fix a typo
- https://github.com/ruby/fiddle/commit/445ca6b501
- 05:29 AM Revision a0dcdd64 (git): [ruby/fiddle] Bump version
- https://github.com/ruby/fiddle/commit/f8fb7c4823
- 05:29 AM Revision 13f64333 (git): [ruby/fiddle] Add Fiddle::VERSION
- https://github.com/ruby/fiddle/commit/9dcf64c096
- 05:29 AM Revision 5fb7ca33 (git): [ruby/fiddle] Add missing spec.extensions
- https://github.com/ruby/fiddle/commit/2ce36b1fdc
- 05:29 AM Revision 77cc1394 (git): [ruby/fiddle] Fix assignment to array within struct (#26)
- * Allow access to a struct's underlying memory with `struct[offset, length]`.
https://github.com/ruby/fiddle/commit/24083690a6 - 05:29 AM Revision 4a835621 (git): [ruby/fiddle] Make array access override compatible with base class (#25)
- * Allow access to a struct's underlying memory with `struct[offset, length]`.
* Make accessing a struct's underlying memory more convenient.
* refactor memory access unit tests for improved clarity
https://github.com/ruby/fiddle/comm... - 05:29 AM Revision aa1d3c7d (git): [ruby/fiddle] Initialize memory to 0 when calling Fiddle.malloc(). (#24)
- https://github.com/ruby/fiddle/commit/8414239ca3
-
- If a module has an origin, and that module is included in another
module or class, previously the iclass created for the module had
an origin pointer to the module's origin instead of the iclass's
origin.
Setting the origin pointer corr... -
- https://github.com/ruby/ruby/pull/3137
-
- The following Ruby program fails with VM assertions enabled on OpenBSD (code taken from `test_caller_to_enum` in `test/ruby/test_backtrace.rb`):
```ruby
def foo
return to_enum(__method__) unless block_given?
r... - 12:43 AM Revision fdb31aa7 (git): Time.xmlschema: Clarify error and docs
- It parses a (date)time, not a date.
-