Activity
From 04/22/2025 to 04/28/2025
Today
08:15 PMBug #21291 (Open): possible bug with /m regexp and $ based on input
- reported in discord by the.unnamed user:
``` ruby
#!/usr/bin/env -S ruby -vw
# => ruby 3.4.2 (2025-02-15 revisio... 07:33 PMBug #21286: Windows - MSYS2 just updated to GCC 15.1.0, builds failing
- I just updated the code in https://github.com/ruby/setup-msys2-gcc (the repo that assembles Windows build tool archiv...
- Maybe gcc-15 could be added to https://github.com/ruby/ruby/blob/master/.github/workflows/compilers.yml?
We don't ... 02:55 PMBug #21286: Windows - MSYS2 just updated to GCC 15.1.0, builds failing
- I wonder if `-Werror` is being set implicitly?
I'm seeing similar failures up and down my CI pipelines for the gem... 12:42 PMBug #21286: Windows - MSYS2 just updated to GCC 15.1.0, builds failing
- I'm getting the same thing in windows CI runs for a bunch of other first-party gems:
* `date`
* `fiddle`
* `io-con... 06:48 AMBug #21286: Windows - MSYS2 just updated to GCC 15.1.0, builds failing
- I'm seeing the same thing on `ruby/json` GitHub Actions builds.
```
bigdecimal.c:2556:1: note: 'BigDecimal_trunc... - 07:10 PMRevision 6052b12d (git): ZJIT: Replace GetConstantPath with Const if the IC is not empty (#13183)
- * Add rb_zjit_constcache_shareable
* Add rb_zjit_multi_ractor_p
* Replace GetConstantPath with Const if the IC is n... 07:08 PMRevision 1e416685 (git): ZJIT: Drop a duplicated call into optimize (#13196)
- * ZJIT: Drop a duplicated call into optimize
* Update a comment
Co-authored-by: Max Bernstein <tekknolagi@gmail.com... 07:06 PMMisc #21290 (Open): Unable to build ruby extension on Fedora 42 due to possible GCC 15 issues
- At least one Ruby extension (ed25519) fails to build due to incompatibilities(?) with the core library headers on Fed...
05:26 PMBug #21289 (Open): Fix C level backtraces for USE_ELF
- After upgrading GitHub to Ruby 3.4 we noticed that we stopped getting useful C level backtrace information in our cra...
04:23 PMRevision 50b1759b (git): YJIT: Fix potential infinite loop when OOM (GH-13186)
- Avoid generating an infinite loop in the case where:
1. Block `first` is adjacent to block `second`, and the branch f... 03:38 PMRevision 92641117 (git): Support Marshal.{dump,load} for core Set
- This was missed when adding core Set, because it's handled
implicitly for T_OBJECT.
Keep marshal compatibility betwe... - Applied in changeset commit:git|80a1a1bb8ae8435b916ae4f66a483e91ad31356a.
----------
YJIT: Fix potential infinite lo... - Avoid generating an infinite loop in the case where:
1. Block `first` is adjacent to block `second`, and the branch f... 10:55 AMRevision 37db51b4 (git): Suppress error message when gc is cleaned already [ci skip]
10:50 AMBug #21288: StringScanner#named_captures overrides matched captures with unmatched captures
- See https://github.com/ruby/strscan/issues/145 and the linked PR
01:55 AMBug #21288 (Open): StringScanner#named_captures overrides matched captures with unmatched captures
- StringScanner#named_captures operates inconsistently with MatchData#named_captures when there is the same name used i...
08:21 AMBug #20203: `TestEnumerable` test failures with GCC 14
- Was this ever resolved or is the undefined behavior still used?
- Clear the ary variable before setting it to nil. Otherwise, if
the previous ary value was somewhere on the stack, al...
04/27/2025
06:39 PMFeature #21287 (Open): Remove SortedSet autoload and set/sorted_set
- When implementing core Set, I moved the `SortedSet` autoload from `lib/set.rb` to `prelude.rb`. However, since this ...
- MSYS2 just updated to GCC 15.1.0, and the ruby-loco builds failed to compile.
See [MinGW](https://github.com/MSP-G... 01:35 PMBug #21271: Ruby build failure due to invalid probe specifier for dtrace
- @nobu Do you remember why CPPFLAGS is needed for dtrace in this commit https://github.com/ruby/ruby/commit/060ae17c80...
- Now that we have a hash-set implementation we can use that
instead of a hash-table with a static value. 09:35 AMFeature #21284: Request: add `Array#pad` method
- No worries, I thought it was some new syntax I was not aware of!
- Current Ruby doesn't have start-exclusive range.
`^..` is so-called neko-operator in Perl 6, and just this request r... - 06:47 AMRevision 6874b289 (git): Free jump buffers leaked by `cont_restore_thread` in WASI builds
- 06:47 AMRevision 45e814d1 (git): Fix jump buffer leak in WASI builds
- I've had a swing at fixing this in https://github.com/ruby/ruby/pull/13186
01:00 AMBug #21104: Net::HTTP connections failing in Ruby >= 3.4.0 on macOS with Happy Eyeballs enabled
- Thank you for the additional reports.
Based on the information shared so far, it seems that the host machines expe...
04/26/2025
- Do we have examples of the maintenance overhead to just keep date as a default gem?
I would think because it's stable... - In my view, `Baz = Class.new` doesn't really "Start a class or module definition." it creates a class but there is no...
- I think the proposal makes sense and would be worth to try it experimentally (i.e. merge to master and see if it actu...
- I think just removing the warning is not right here (also explicitly testing for chilled strings feels dirty).
I t... - 11:56 AMRevision 687bd837 (git): [ruby/ipaddr] Added IPAddr#+/-
- https://github.com/ruby/ipaddr/commit/78b4f53bf5
11:52 AMBug #21285: Range#overlap? returns false for (nil..nil)
- Yeah, indeed. I was fixed in https://github.com/ruby/ruby/pull/12937. Thank you.
- Now that we have a `set_table` implementation, we can
use it to track const caches and save some memory.
We could ev... - Applied in changeset commit:git|e4f85bfc311a3812de7bc2e9d068934e8b364574.
----------
Implement Set as a core class
... - Set has been an autoloaded standard library since Ruby 3.2.
The standard library Set is less efficient than it could ...
04/25/2025
11:45 PMBug #21104: Net::HTTP connections failing in Ruby >= 3.4.0 on macOS with Happy Eyeballs enabled
- I can repro this on M4 chipset but not M3. `Fix is export RUBY_TCP_NO_FAST_FALLBACK=1`
See https://github.com/ruby... 08:46 PMRevision b1283b45 (git): maybe fix bindgen
- This commit inlines instructions for Class#new. To make this work, we
added a new YARV instructions, `opt_new`. `op... 07:21 PMBug #21285 (Closed): Range#overlap? returns false for (nil..nil)
- I believe it is already fixed.
- I've notices the following edge case:
```ruby
(..1).overlap?(nil..nil)
=> false
```
But I expect `true` - 06:04 PMRevision e0545a02 (git): ZJIT: Bail out of HIR translation if we can't handle a send flag (#13182)
- Bail out of HIR translation if we can't handle a send flag
06:00 PMFeature #21221: Proposal to upstream ZJIT
- Eregon (Benoit Daloze) wrote in #note-6:
> In my reply I was even mostly focusing on the "jumping to other compiled ... - 04:33 PMRevision 871d07a2 (git): ZJIT: Emit CCall if we know the type statically, not just from profiles (#13173)
- Emit CCall if we know the type statically, not just from profiles
09:20 AMBug #21282: `FrozenError` raised by `rb_error_frozen_object` should mention real class of frozen object
- On a second thought, I've come to the conclusion that the benefits in clarity that a fix would provide are not worth ...
- In cb1ea54bbf6cdf49c53f42720fec1a151069810c I added one more
metadata flag, but didn't notice `RB_GC_OBJECT_METADATA_... 07:37 AMRevision f501d114 (git): Use OS label for actions/cache. vcpkg only uses same os cache
- When I try and run that code, e.g.
``` ruby
[1, 2, 3].fill(nil, -1^..10)
```
``` shell
'Integer#^': Range ca... - It may be possible with `Array#fill` and start-exclusive range?
```ruby
ary.fill(object, -1^..padding_length)
``` - A method to pad an array of arbitrary length with objects up to a specified array size does not currently exist.
`... - https://github.com/rubygems/rubygems/commit/62e1bf2d37
- 07:03 AMRevision 24c838f9 (git): Update bundled gems list as of 2025-04-25
- This reverts commit 8e258af5b083a417a8f5e424a2c7d2c4800f8ca2.
- 02:17 AMRevision daef56cf (git): Update default gems list at 6f336d1b154e608d5200bd8de48473 [ci skip]
02:12 AMRevision 6f336d1b (git): [rubygems/rubygems] Remove shellwords autoload
- https://github.com/rubygems/rubygems/commit/2af1646776
- https://github.com/ruby/json/commit/c985e8c6ea
- Fix: https://github.com/ruby/json/issues/788
`multi_json` rely on it, even though it was never documented as
public ... - https://github.com/ruby/json/commit/84443e881d
- This time with explicit deprecation warnings.
https://github.com/ruby/json/commit/0dee9bdad9 - https://github.com/ruby/json/commit/a6949f8656
04/24/2025
- 11:41 PMRevision 9aed9e94 (git): ZJIT: Fold Send into SendWithoutBlockDirect if we know the class statically (#13172)
- Fold Send into SendWithoutBlockDirect if we know the class statically
This applies for constants and also for values... - 06:33 PMRevision 71166f60 (git): Add tests
- 06:33 PMRevision c2677f4d (git): Make type_of use union-find
- Otherwise we might have stale types floating around
- 06:33 PMRevision 4c1f29fd (git): Make PatchPoint(CalleeModifiedLocals) implicit
- We don't have a good model for modeling weakref-like instruction
dependency behavior and I don't think we should add ... - 06:33 PMRevision b437a3ee (git): Strength reduce sends on Fixnums in the optimizer instead of HIR builder
- 06:33 PMRevision f36f7920 (git): Simplify fixnum guarding code
- I (mis)understood side-exit as anything getting out of the current compilation unit, whether that's jumping to other ...
Eregon (Benoit Daloze) wrote in #note-3:
> Isn't this the main reason to move to a method-based JIT?
>
> I th...05:06 PMFeature #20610 (Closed): Float::INFINITY as IO.select timeout argument
- committed.
https://github.com/ruby/ruby/commit/698ef864a59004f468c77534d59e51e862ec2624 - 12:15 PMRevision bbd5d334 (git): Fix tiny typo in syntax/exceptions.rdoc
12:03 PMRevision 18f8c514 (git): Fix memoization for the `/(...){0}/` case (#13169)
- In this case, the previous implementation counted an extra number of
opcodes to cache and the matching was unstable o... 10:09 AMBug #19473: can't be called from trap context (ThreadError) is too limiting
- I agree the current behaviour seems too limiting. If user code does something incorrect, it should fail, or deadlock....
- ```
1) Failure:
TestThread#test_join_argument_conversion [D:/a/ruby/ruby/src/test/ruby/test_thread.rb:249]:
E... - Given that the currently planned ractor local GC implementation
performance will heavilly be influenced by the number... - 07:03 AMRevision 7af5a632 (git): Update bundled gems list as of 2025-04-24
- Probably due to warnings in CP932 encoding.
The warning issue is already reported at: https://developercommunity.v... - The following tests are broken with VS2022 17.14.0 preview 4.0.
```
1) Error:
TestMkmfConvertible#test_typeof_... - Co-authored-by: Nobuyoshi Nakada <nobu@ruby-lang.org>
04/23/2025
- 08:45 PMRevision bf3be272 (git): ZJIT: Fix README example command (#13163)
- Fix README example command
- ```
../string.c:660:38: warning: comparison of integers of different signs: 'rb_atomic_t' (aka 'unsigned int') and 'i... - PR https://github.com/ruby/ruby/pull/13161
- Hello everyone,
I've noticed that the following snippets both raise `FrozenError`, but with different messages:
... - > Since code will not break in this case
That is debatable. It won't raise an exception yes, but the behavior wil... - byroot (Jean Boussier) wrote in #note-5:
> Chilled string as designed should behave like mutable strings, and warn w... - Perhaps it's the jetlag talking, but I think the current behavior is desirable.
Chilled string as designed should ... - byroot (Jean Boussier) wrote in #note-3:
> > But behavior in this case will not change
>
> Maybe I'm missing some... - > But behavior in this case will not change
Maybe I'm missing something, but I think the behavior will change:
```r... - byroot (Jean Boussier) wrote in #note-1:
> Chilled strings should also emit a warning when they are mutated to signa... - > Should this result in a warning (current behavior), or is it safe to avoid changing the encoding of the underlying ...
- * [Feature #15408] Actually deprecate `ObjectSpace._id2ref` (byroot)
* Matz agreed to deprecate `_id2ref` for Ruby ... - # The next dev meeting
**Date: 2025/05/08 13:00-17:00** (JST)
Log: *TBD*
- Dev meeting *IS NOT* a decision-mak... - `github.event.before` for newly pushed branch is all zero and cannot
check out. - This makes the finalizer table fully self contained, so GC no
longer need to delay cleaning the `obj_to_id_tbl`. - I opened a PR to finally deprecated `ObjectSpace._id2ref`: https://github.com/ruby/ruby/pull/13157
05:07 AMRevision 5f3fb35a (git): Add a new spec for inherited
- 04:05 AMRevision 60297819 (git): [DOC] Update for String#split
- Highlight the performance advantages of calling `string.split` with a block, instead of `string.split.each` with the ...
- 01:17 AMRevision c1dbd01c (git): Increase fiber sleep test tolerance. (#13152)
04/22/2025
- `objspace->finalizer_table` must be synchronized,
otherwise concurrent insertion from multiple ractors
will cause a c... - Let me add that the example above is not made up.
I have seen many Rails applications with `require` calls for autol... - > So, Zeitwerk is able to avoid the limitation clarified in this ticket, right?
Maybe, but do you think that any Rub... - fxn (Xavier Noria) wrote in #note-30:
> Correct, Zeitwerk only sets autoloads with absolute paths (that may contain ... - 04:17 PMRevision 9fc5a5db (git): ZJIT: Add option to dump LIR (#13139)
- Now we can dump HIR, optimized HIR, LIR, and assembly.
- 03:50 PMRevision 87d00142 (git): Make yjit team auto-review ZJIT PRs
- https://github.com/ruby/pp/commit/efe5bc878f
- The array allocation was because the keyword splat expression is
not recognized as safe by the compiler. Also avoid ... - Rubygems sets the date of built gems to `DEFAULT_SOURCE_DATE_EPOCH`
now unless `SOURCE_DATE_EPOCH` environment variab... - Only `class_table` values are not quoted and need to terminate %-=
literal. - Also, rename `format_changelog` to `changelog_formatter` since this
method does not format the changelog when called,... - Thanks for your confirmation.
Unfortunately, Ruby 3.2 is already security maintenance status. It means I will not ... 08:10 AMBug #21276: Regression segfault instead of exception `require': cannot load such file
- master as of 22 April 2025 (888820c581065c7d5cbcd4a137af1e66a708979c) is okay and does NOT have the bug.
Another o... - Can you confirm to reproduce this with the current master branch?
- vs2022 17.14.0 preview 3.0 resolved this issue.
- 07:03 AMRevision 888820c5 (git): Update bundled gems list as of 1980-01-02
- https://github.com/ruby/ruby/actions/runs/14585602800/job/40910453142
- 06:34 AMRevision 8e8df8ef (git): Remove the temporary script:
- - The remaining code was not ported over as I felt it wasn't useful
to output the Ruby related version in the conte... - 02:49 AMRevision c3bff185 (git): Update default gems list at 4634a0042e34c022984ef2d3b377ed [ci skip]
- 02:27 AMRevision 067fc410 (git): [ruby/strscan] Bump version
- https://github.com/ruby/strscan/commit/8ff80150c4
- https://github.com/ruby/json/commit/92beca8032
- Because get_push_scope is a method call, Ruby will allocate an array
for *args even though it is not necessary to do ... - 02:27 AMRevision 56c1a15e (git): [rubygems/rubygems] Warn if TLS 1.2 is not supported
- https://github.com/rubygems/rubygems/commit/e4f70a3e4f
- 02:27 AMRevision ff2e0e41 (git): [rubygems/rubygems] Summarize the diagnostic
- https://github.com/rubygems/rubygems/commit/40cf54d256
- 02:27 AMRevision 19477ef2 (git): [rubygems/rubygems] Diagnose the bare net/http connection
- https://github.com/rubygems/rubygems/commit/38a0bdc123
- 02:27 AMRevision 7a10ce8c (git): [rubygems/rubygems] Diagnose the RubyGems connection
- https://github.com/rubygems/rubygems/commit/bf63859e1e
- 02:27 AMRevision cba74080 (git): [rubygems/rubygems] Diagnose the bundler connection
- https://github.com/rubygems/rubygems/commit/0aae094c89
- 02:27 AMRevision ae308ae5 (git): [rubygems/rubygems] Diagnose when OpenSSL can't be loaded.
- https://github.com/rubygems/rubygems/commit/e6aa8aabcd
- 02:27 AMRevision cda29294 (git): [rubygems/rubygems] Add the `bundle doctor subcommand` skeleton:
- - The command can either be run using:
1. `bundle doctor --ssl`
2. `bundle doctor ssl`
The later is most usef... - 02:27 AMRevision c2e58a90 (git): [rubygems/rubygems] Define `bundler doctor` as a subcommand
- - See explanation in previous commit https://github.com/rubygems/rubygems/commit/170890befb4c
https://github.com/rub... - 02:27 AMRevision ef82273c (git): [rubygems/rubygems] Import the SSL Check script:
- - Copy/pasted from https://github.com/rubygems/ruby-ssl-check.
The next commits will break it down for easier revie... - 02:27 AMRevision 89d1af9a (git): Move the doctor command into a subfolder:
- - Adding a new `ssl` option to bundle doctor will make the `Doctor`
command quite bloated. The "diagnose ssl" optio... - `StringIO#set_encoding` changes the underlying string encoding if the string is not frozen, but does not change the u...
Also available in: Atom
