Activity
From 03/17/2022 to 03/23/2022
03/23/2022
-
11:59 PM Revision 1357b147 (git): Now all extension libraries must consider the ABI header
-
-
10:34 PM Revision d9dd88a6 (git): [rubygems/rubygems] Avoid crash in test teardown
- If an exception happens during test `setup` method, the `teardown`
method will still be run for cleaning up, but if s... -
10:29 PM Bug #18511 (Closed): ruby in infinite loop consuming 100% CPU
-
- I did some testing and can confirm this is still a bug in the master branch. From my testing, this isn't related to ...
-
09:16 PM Bug #18656 (Rejected): Documentation of Comparable Module: clamp method not mentioned in introduction
- From the comparable docs ( https://ruby-doc.org/core-3.1.1/Comparable.html ): "Comparable uses <=> to implement the c...
- 08:32 PM Revision 8b05b5a0 (git): * 2022-03-24 [ci skip]
-
- Completely exclude the full bundler folder. The actual Bundler docs are
excluded anyways by ruby-core (by bundler/lib... -
05:07 PM Bug #14103 (Open): Regexp absense operator has no chance to ^C
- This change degrades the performance of regular expression matching when frequent backtracking occurs.
Before edc8... -
02:56 PM Bug #18635 (Closed): Enumerable#inject without block/symbol will return values or raise LocalJumpError
- Applied in changeset commit:git|8f1c69f27ce6b3f5ed1c1cf8d2aa62aa9701d636.
----------
Raise ArgumentError when callin... -
02:41 PM Bug #18635: Enumerable#inject without block/symbol will return values or raise LocalJumpError
- I think either is fine.
I think some methods raise LocalJumpError early, even if they didn't try to use the block ye... -
- Previously, this would work as expected if the enumerable contained
0 or 1 element, and would raise LocalJumpError ot... -
- mame (Yusuke Endoh) wrote in #note-40:
> @naruse conceived the idea. TBH, I am unsure if it will work well.
@naru... -
- Eregon (Benoit Daloze) wrote in #note-38:
> I think it's not a good idea to have Regexp::TimeoutError < Timeout::Err... -
- This is hiding a real bundler issue as a "network error". It's more
helpful to get a proper bug report, with stack tr... -
01:47 PM Feature #18566 (Closed): Merge `io-wait` and `io-nonblock` gems into core IO
- Closing in favor of [Feature #18655], since `wait_readable` and `wait_writable` are really the two method I personall...
-
- Extracted from [Feature #18566].
The decision was made to consider the methods from `io-wait` and `io-nonblock` on... -
- As this converter seems to convert JIS0201 to JIS0208 as possible, this part would be intentional.
-
- Accessing outside the table is definitely a bug.
Before that, SS2+0xA1 which should be JIS0201 is converted to JIS02... -
- duerst (Martin Dürst) wrote in #note-3:
> Aren't there still regular expressions that RE2 can't handle but that may ... -
02:56 AM Feature #18653: Use RE2 for Regexp
- mame (Yusuke Endoh) wrote in #note-2:
> My original motivation was a security measure against ReDoS. If RE2 worked... -
- vo.x (Vit Ondruch) wrote in #note-1:
> Could you please elaborate what was the motivation for this experiment?
My... -
- Looks like this bug started in Ruby 2.2:
```
$ ruby21 -v t/t55.rb
ruby 2.1.9p490 (2016-03-30 revision 54437) [x8... -
- noraj (Alexandre ZANNI) wrote in #note-3:
> As far as I understand, it's not possible to have a native UTF-16 stri...
03/22/2022
-
- The call-seq for inject does not show usage without an argument or block, so I think it is reasonable to consider thi...
-
08:42 PM Feature #18653: Use RE2 for Regexp
- Could you please elaborate what was the motivation for this experiment? Was it to evaluate compatibility? Or possible...
-
- I have tried to use [RE2](https://github.com/google/re2) as Ruby's regular expression engine. As it turns out, it see...
-
08:26 PM Feature #18654 (Closed): Enhancements to prettyprint
- This issue mirrors the pull request open at https://github.com/ruby/prettyprint/pull/3.
This pull request adds a b... -
07:51 PM Revision 0140e6c4 (git): [DOC] Enhanced RDoc for String (#5685)
- Treats:
#chars
#codepoints
#each_char
#each_codepoint
#each_grapheme_cluster
#grapheme_clust... -
06:33 PM Bug #18643: test-bundler failed on ruby_3_1 CI
- > https://github.com/rubygems/rubygems/pull/5410
The PR was merged in rubygems.
-
- I am working on this PR to fix this test considering the case that the latest Bundler version on RubyGems is newer th...
-
-
03:39 PM Bug #18650: C Extensions Can Segmentation Fault on Ruby 2.7 Compiled with --with-jemalloc and --enabled-shared
- shyouhei (Shyouhei Urabe) wrote in #note-1:
> Setting status to closed. Don't get confused, this is a normal proces... -
12:46 AM Bug #18650 (Closed): C Extensions Can Segmentation Fault on Ruby 2.7 Compiled with --with-jemalloc and --enabled-shared
- Setting status to closed. Don't get confused, this is a normal process. The bug is fixed -- now is the time for bac...
- 03:01 PM Revision 2b01d7f2 (git): * 2022-03-23 [ci skip]
- 03:01 PM Revision 1ff174bf (git): [rubygems/rubygems] Fix a test for `bin/bundle update --bundler` to pass on ruby/ruby.
- Consider the case that the latest Bundler version on RubyGems is higher than
the `system_bundler_version` (= `Bundler... -
01:55 PM Feature #17837: Add support for Regexp timeouts
- mame (Yusuke Endoh) wrote in #note-37:
> BTW, we agreed that we do not introduce `Regexp.backtrack_limit=`. It would... -
- mame (Yusuke Endoh) wrote in #note-37:
> To reuse the code that `rescue`s `Timeout::Error`, `Regexp::TimeoutError` s... -
- I discussed this issue with some committers including @matz, @nobu, @akr, and @naruse. In light of the recent increas...
-
01:43 PM Feature #18634 (Closed): Variable Width Allocation: Arrays
- Applied in changeset commit:git|a51f30c6712798fc07e57f692d0d0e5ccc59acf1.
----------
[Feature #18634] Implement Arra... -
- This commit implements arrays on Variable Width Allocation. This allows
longer arrays to be embedded (i.e. contents d... -
- mame (Yusuke Endoh) wrote in #note-16:
> @akr was against to make io/nonblock built-in because it is a old-style int... -
- Those methods have been in stdlib for a while, I don't think moving them to core is an issue (their name and behavior...
-
- My personal opinion is that `wait_readable` and `wait_writable` should be in core IO. All the rest can stay in the ge...
-
05:23 AM Feature #18566: Merge `io-wait` and `io-nonblock` gems into core IO
- > wait_readable and wait_writable are maybe okay.
> No attendee knows what wait_priority is.
For reading `MSG_O... -
- This issue was discussed at the dev meeting.
The motivating issue of this proposal (the dependency issue of io-wai... -
- Yes, that's correct, and that code would also break when migrating to other form of delegations (if not changing `bar...
-
- As far as I understand, this change will break the following code:
```
def target(a:)
p a
end
# after the ... -
- # The next dev meeting
**Date: 2022/04/21 13:00-17:00** (JST)
Log: https://github.com/ruby/dev-meeting-log/blob/m... -
-
-
12:49 AM Revision bbd29fe0 (git): Fix a link [ci skip]
03/21/2022
-
- As @peterzhu2118 and @duerst have pointed out, putting string method's RDoc into doc/ (which allows non-ASCII in exam...
-
- Sorry if the description is confusing. This is a general timeout for all non-blocking IO operations, specified per IO...
-
- I guess one way to express it is this new IO#timeout would apply to: blocking IO methods on non-blocking IO instances.
-
- I'm still confused, in which cases does this timeout apply?
Only non-blocking IO, right?
The issue title says:
>... -
- > I'm not sure a timeout per IO instance makes sense,
I think it does for various network clients. See for example... -
- Testing Bundler seems to use a Gemfile, so I guess we can just make it point to my branch for testing: https://github...
-
- Eregon (Benoit Daloze) wrote in #note-8:
> @jeremyevans0 There is a good chance this is missing `ruby2_keywords` cal... -
- @jeremyevans0 There is a good chance this is missing `ruby2_keywords` calls in RSpec, could you try with https://gith...
-
- This cannot be merged yet as it currently causes a couple errors in the Bundler specs:
* https://github.com/rubyge... -
- From the [dev meeting log](https://hackmd.io/koJLPz4eRXKzaaDvVqji7w#Bug-18625-ruby2_keywords-does-not-unmark-the-hash...
-
-
- 04:32 PM Revision 35cc7060 (git): * 2022-03-22 [ci skip]
-
- https://github.com/ruby/rdoc/commit/f9f90ef2ff
-
- https://github.com/ruby/rdoc/commit/b7b4cdab6c
-
- https://github.com/ruby/rdoc/commit/c870284163
-
04:03 PM Feature #18619: Reverse the order of GC Compaction cursor movement
- eightbitraptor (Matthew Valentine-House) wrote in #note-3:
> Thanks for the quick feedback folks.
>
> Eregon (Ben... -
- Bump.
Today I just came across a case where I wanted to use this feature.
@nobu according to dev meeting notes you ... -
07:54 AM Bug #17529 (Closed): Ractor Segfaults with GC enabled
- Applied in changeset commit:git|a72b7b898c69a116d754d599e8bb061761015255.
----------
merge revision(s) d0d6227a0da59... -
- ruby_3_0 a72b7b898c69a116d754d599e8bb061761015255 merged revision(s) d0d6227a0da5925acf946a09191f172daf53baf2,fff1edf...
-
06:47 AM Bug #17529: Ractor Segfaults with GC enabled
- I guess that the btest failure of ruby_3_0 branch on icc-x64 env may be fixed by `git cherry-pick d0d6227a0da5925acf9...
-
- I confirmed with 3.0.0 that the issue is reproducible.
According to `git bisect`, it seems to be fixed in fff1edf23b... -
- alen should be actions number on ractor_select()
alen was number of rs, but it should be actions number
... -
-
-
- ruby_3_0 e0146e6cc8f3578b02ad5f228f86bf1aef566d16 merged revision(s) 97426e15d721119738a548ecfa7232b1d027cd34.
-
- [Bug #18627] Fix crash when including module
During lazy sweeping, the iclass could be a dead object that ha... -
- ruby_3_0 branch includes the changeset git|ddd9704ae9bf884d867b6f57a16c095a79393fff.
-
12:51 AM Bug #18651 (Closed): oob access in CP51932 -> CP50220 transcoder
- Hello,
while working on a port of Japanese transcoder, I found a potential out of bounds access:
https://github.com...
03/20/2022
-
02:24 PM Feature #18626 (Rejected): 注釈付き代入演算子 ()= の提案
- Very interesting proposal both syntax-wise and semantic-wise. But I have decided that we are not going to introduce a...
-
- I tried to reproduce the problem at hand, but it seems to have already been fixed in the current master.
According t... -
- https://github.com/ruby/rdoc/commit/1bb0496c53
-
- https://github.com/ruby/rdoc/commit/7736d3a89c
- 05:28 AM Revision 7b3fdcdf (git): * 2022-03-20 [ci skip]
-
- Empty class documents are generated even with `:stopdoc:`.
03/19/2022
-
- Alternatively, the following seems to solve the problem.
```patch
diff --git a/vm_eval.c b/vm_eval.c
index 76e56fa... -
- I shortened the reproduction code.
```
Array.prepend(Module.new)
module UsingRefinements
using(Module.new d... - 01:50 PM Revision 334ca2dc (git): merge revision(s) 496591de96b261b8789332c7f8b2bfbd17658955: [Backport #18578]
- st.c: Do not clear entries_bound when calling Hash#shift for empty
hash
tab->entries_bound is used to check... - 01:45 PM Revision 19fec939 (git): merge revision(s) c79d2e54748f52c5023b0a1ee441561df9826c17: [Backport #18562]
- Fix TAG_THROW through require [Bug #18562]
Previously this was being incorrectly swapped with TAG_RAISE in the n... - 01:41 PM Revision 7eaec9a6 (git): merge revision(s) b555e659c4974acc423083b71b1bd5ec6a926046: [Backport #18388]
- Do not use `fcopyfile` if appending to non-empty file [Bug #18388]
`fcopyfile` appends `src` to `to` and then tr... - 01:34 PM Revision 5da2a3e3 (git): merge revision(s) 7ff1bf317887c0d7b21e91ad548d07b9f05c540c,e89d80702bd98a8276243a7fcaa2a158b3bfb659: [Backport #18516]
- An alias can suppress method redefinition warning
---
test/ruby/test_alias.rb | 11 +++++++++++
1 file ... - 01:27 PM Revision 2d886af0 (git): merge revision(s) 6a6227e0168b059c3ed34c9f0ace2e5dc2364221: [Backport #18517]
- Shifting zero always results in zero [Bug #18517]
---
numeric.c | 2 ++
1 file changed, 2 i... - 01:23 PM Revision 81f5453b (git): merge revision(s) 5c7af72304d0ad33cd3f21b24a4bc44e8acd5b2c,d650b17686d49c2ce8e6a87039861154e93d4621: [Backport #18497]
- Assuming EXIT_SUCCESS equals 0 is not portable
---
test/ruby/test_fiber.rb | 6 +++---
1 file changed, ... -
01:22 PM Bug #18641: UTF-16 surrogate pairs
- Thank you Martin.
I'm actually working on an Unicode study, I was not interested into representing the emoji with ... - 01:15 PM Revision ea9b0fe5 (git): merge revision(s) ae5458f228a5477383e9c00425d85d50a3867817: [Backport #18475]
- thread.c: Convert TAG_BREAK to a normal exception at thread top-level
[Bug #18475]
---
test/ruby/test_e... -
-
- I think the current hard code is a bad habit, and `ldd` should be primary.
03/18/2022
- 10:17 PM Revision 44998f27 (git): * 2022-03-19 [ci skip]
-
- Treats:
#split
#each_line
#lines
#each_byte
#bytes -
09:14 PM Bug #18645 (Third Party's Issue): fiddle.rb assumes wrong file name for glibc shared object on Alpha
- Could you report this to https://github.com/ruby/fiddle/ because ruby/fiddle is the upstream?
-
09:17 AM Bug #18645 (Third Party's Issue): fiddle.rb assumes wrong file name for glibc shared object on Alpha
- On Alpha, the fiddle.rb test fails because the code in helper.rb assumes that the name for the libc shared object fil...
-
06:25 PM Feature #18648: ruby2_keywords and ... name arguments with impossible names
- bkuhlmann (Brooke Kuhlmann) wrote in #note-2:
> Hey Aaron, in regards to `*`, `**`, and `&`, those are passthrough p... -
- jeremyevans0 (Jeremy Evans) wrote in #note-1:
> This isn't a bug, it's currently expected behavior, even if it is a ... -
06:12 PM Feature #18648: ruby2_keywords and ... name arguments with impossible names
- Hey Aaron, in regards to `*`, `**`, and `&`, those are passthrough parameters (or bare parameters as I like to call t...
-
- This isn't a bug, it's currently expected behavior, even if it is a bit inconsistent. I'll switch this to feature, b...
-
- While investigating a break in a library using reflection, I realized that when ... is used or ruby2_keywords is used...
-
- I requested a 2.7 back port here: https://bugs.ruby-lang.org/issues/18650
-
- Greetings,
I've encountered a segmentation fault using the EventMachine gem on Ruby 2.7.5 compiled with configurat... -
- Actually there is another probably related bug on 3.1: #18649
-
-
- @mame Thanks for the repro and explanation.
I've been looking into this and summarized some thoughts in https://githu... -
- ```ruby
def synchronize
yield
end
def execute(task)
success = true
value = reason = nil
end_sync = f... -
- Applied in changeset commit:git|97426e15d721119738a548ecfa7232b1d027cd34.
----------
[Bug #18627] Fix crash when inc... -
- During lazy sweeping, the iclass could be a dead object that has not yet
been swept. However, the chain of superclass... -
- As background, `to_proc` already exists as a coercion protocol, it's what is used for `call(&callable)`, and there is...
-
05:48 AM Feature #18644 (Open): Coerce anything callable to a Proc
- Functional objects are increasingly popular in Ruby. Having objects that respond to `#call` makes them interchangeabl...
-
- Backport was requested in https://bugs.ruby-lang.org/issues/18646, and I adjusted the Backport field for it.
-
- I've set the backport field for https://bugs.ruby-lang.org/issues/17302.
For 2.6 it's unlikely given it's soon EOL.
I... -
11:28 AM Bug #18646: Inconsistent behavior for the targeted TracePoints in ruby 2.6-2.7 and 3+
- Looks like this was fixed in https://bugs.ruby-lang.org/issues/17302
Any plans to backport to 2.6/2.7? -
- Ruby 2.6 introduced the API for targeted tracepoints.
But for some reason, before ruby 3.0, tracepoints looks exclus... -
- I can see that iseq-targeted TracePoints introduced in ruby 2.6 working recursively and there is no API to avoid that...
-
- hsbt (Hiroshi SHIBATA) wrote in #note-1:
> https://github.com/ruby/ruby/pull/5608 will fix this issue.
Unfortunat... -
04:29 AM Bug #18643 (Assigned): test-bundler failed on ruby_3_1 CI
- https://github.com/ruby/ruby/pull/5608 will fix this issue.
-
- I think that this spec has failed since newer bundler released.
https://github.com/ruby/actions/runs/5590419280?ch... -
- This uses similar language to that used in Method#curry.
-
03:01 AM Feature #18598: Add String#bytesplice
- String#bytesplice has been added by commit:git|1107839a7fed31339fc947995b7b45b8eaf4041b.
-
- Applied in changeset commit:git|2fdfd499db489db9eb4046849aa785c3bd382761.
----------
Add a NEWS entry about [Feature... -
01:43 AM Feature #18598: Add String#bytesplice
- Matz said "go ahead".
-
-
-
- from devmeeting yeasterday:
matz: No change.
matz: it would be good to add a dedicated documentation to Proc#curr... -
- `"\uD83D\uDC69"` tries to create an UTF-8 string with surrogates. In UTF-8, surrogates are not allowed, and therefore...
03/17/2022
-
- One of the biggest pain-points working with the existing ripper subclasses is that you have to know what each array i...
-
- test against 3.1.1p18
-
- That Ruby triggers an *invalid Unicode codepoint* error while using surrogate pairs in an UTF-8 string is expected, h...
-
- Applied in changeset commit:git|b6804d62f822237e136e698e006c913df9990ec2.
----------
Make Proc#parameters support la... -
- A lambda: kwarg when it's meant to be used for non-lambda Procs seems confusing.
How about `parameters(from: :source... -
- Hmm, I accept `lambda:` keyword argument (default `false`). There might be a better candidate than `lambda` but it's ...
-
06:56 PM Revision b8e72bd2 (git): Revert "Faster rb_class_superclass"
- This reverts commit 29b68b89a0c0ea7de46c058fab746550398151f0.
-
- This makes it easier to use Proc#parameters to build wrappers.
Implements [Feature #15357] -
- This uses the RCLASS_SUPERCLASSES array to quickly find the next
SUPERCLASS of klass which is a T_CLASS. -
- nobu (Nobuyoshi Nakada) wrote in #note-4:
> How about `letters` and `each_letter`?
I like the general idea, but to m... -
- For the record, "Grapheme" and "Grapheme cluster" are different concepts. If we call them "grapheme", It's kind of li...
-
- How about `letters` and `each_letter`?
-
- > I'm not sure it's valuable to have per-IO timeout.
`async-io` has used it for years successfully as a protection... -
- This part wasn't answered:
> Is there a concrete/practical example (for this new timeout)?
I'm not sure it's va... -
- @matz thanks for your comments.
> time-out may happen from I/O blocking operations or other CPU bound operation. T... - 03:42 PM Revision 58b35517 (git): * 2022-03-18 [ci skip]
-
- Do not "allocate then wrap". It leaks the allocated memory if
failed to create the wrapper. -
-
-
03:06 PM Feature #18576: Rename `ASCII-8BIT` encoding to `BINARY`
- Having solved a lot of encoding issues for co-workers, especially on Windows, I'm with @Eregon. As the programmers be...
-
- I think it's worth changing, the current name is confusing to most Ruby users, and there were only 71 gems out of 170...
-
- The risk of compatibility has been reduced thanks to @byroot's effort, but probably there still are many applications...
-
02:46 PM Feature #18640 (Open): default empty string argument for `String#sub` and `String#sub!`, e.g. `"hello".sub("l")`
- Most of the time I replace a string with an empty string. And I think that's the main use case.
Also, `String#dele... -
- Applied in changeset commit:git|e660b934b98943826f888f2b73f773c6411cd199.
----------
A positional Hash is not keywor... -
- > I'm not against the proposal, but for the record, the change makes Object#extend and Module#include slow.
I thin... -
- I am positive introducing this proposal.
Matz.
-
- OK, I accept introducing `undefined_instance_methods`. It returns a list of undefined methods of the receiver class.
... -
- This inconsistency does not have any actual drawback, so we keep the current behavior.
Matz.
-
01:10 PM Feature #18583: Pattern-matching: API for custom unpacking strategies?
- > This, though, raises a question of several match groups, at which point one starts to want more:
> ```ruby
> case... -
-
- Actually not exactly, while ruby2_keywords ships empty with ruby 3.1, manually installing it does bring a `.rb` file.
-
- I think an empty gem is the strategy the `ruby2_keywords` gem has already followed on Ruby 3.1 and it worked fine (al...
-
- I think the gem could just be empty, and so `require "io/wait"` would just require the stdlib (or noop if a Ruby impl...
-
- > I think we should only add require_ruby_version = "< 3.2" to io-wait gem after merging io-wait into core IO class
... -
- I think we should only add `require_ruby_version = "< 3.2"` to io-wait gem after merging `io-wait` into core `IO` cla...
-
- >Remove io-wait as a dependency of all gems maintained by ruby-core (e.g. net-protocol).
It's already done. -
- Should be fixed. Review the patch and merge it if it's OK. @nobu please?
Matz.
-
08:44 AM Revision e9553a8c (git): spec: skip specs where set TZ environment variable for WASI
- WASI doesn't respect TZ env var for now
-
- WASI doesn't provide a way to spawn a new process
- 08:29 AM Revision df6f2b64 (git): Update default gems list at fadc7a150c90ad44dee90985b19e79 [ci skip]
-
- https://github.com/ruby/logger/commit/c4a4155ab7
-
-
- Applied in changeset commit:git|329c89bb42bb44467588afc1d41d9f99172dfeb5.
----------
Make implicit function declarat... -
- I accept it since it's already 2022. I was persuaded. Go ahead.
Matz.
-
- Enable `-Werror=implicit-function-declaration` by default for
building C extensions for early failures. -
- cap-std, an underlying sandbox implementation of WASI in wasmtime, doesn't
allow to create a symlink to an absolute p... -
- I am in wrong time zone. if you have time, can you check these?
- [Feature #18630] Introduce general `IO#timeout` ... -
- - `www.ruby-lang.org` links to `./www.ruby-lang.org`
- `cgi['field_name']` links to `./'field_name'` -
- @hsbt: I'm very sorry, but this error has resurfaced. Yesterday afternoon, it appeared in about half of the tries of ...
Also available in: Atom