Activity
From 07/03/2025 to 07/09/2025
07/09/2025
- 11:03 PM Revision f5085c70 (git): ZJIT: Mark profiled objects when marking ISEQ (#13784)
- 09:51 PM Revision 1e3d24a0 (git): [ruby/prism] Accept a newline after the defined? keyword [Backport #21197] (#13327)
- * [ruby/prism] Accept a newline after the defined? keyword
[Bug #21197]
https://github.com/ruby/prism/commit/22be955ce9
* Fix a compilation error
Co-authored-by: Stan Lo <stan001212@gmail.com>
---------
Co-authored-by: Kevin Newton... - 09:50 PM Revision 1df94aaf (git): ZJIT: Name side-exit test cases correctly
- 09:50 PM Revision e2a81c73 (git): ZJIT: Optimize `opt_and` and `opt_or` instructions for Fixnum
- 09:50 PM Revision 10b582da (git): ZJIT: Profile `opt_and` and `opt_or` instructions
-
09:01 PM Bug #21505: [Ractor] calling exit from non-main ractor will hang if main ractor waits on it
- I opened https://github.com/ruby/ruby/pull/13838 which fixes this
```
./miniruby -e 'Ractor.new { exit }.join'
-e:1: warning: Ractor is experimental, and the behavior may change in future versions of Ruby! Also there are many implem... -
06:42 PM Revision b146eae3 (git): [DOC] Tweaks for String#clear
-
-
-
06:41 PM Feature #21386: Introduce `Enumerable#join_map`
- I am against this.
Ergonomics: Adding a special "X_Y" method for every common pattern of "X followed by Y" is truly horrible for the ergonomics of the language. It only multiplies the number of useless details the programmer should reme... -
-
- Previously it was possible for two atomic increments of next_shape_id
running concurrently to overflow MAX_SHAPE_ID. For this reason we need
to do the test atomically with the allocation in shape_alloc returning
NULL.
This avoids overfl... -
- When sharing between threads we need both atomic reads and writes. We
probably didn't need to use this in some cases (where we weren't running
in multi-ractor mode) but I think it's best to be consistent. -
- This was using < so subtract one from the last shape id would have us
miss the last inserted shape. I think this is unlikely to have caused
issues because I don't think the newest shape will ever have edges.
We do need to use `- 1` beca... -
04:40 PM Revision fd8a67fc (git): Re-ordered vcpkg related steps. It may be affected with VsDevCmd.bat
-
-
-
-
01:11 PM Revision 08738779 (git): Update to ruby/spec@ed254ba
-
11:18 AM Revision ba246c5a (git): [ruby/io-wait] Revert "Fix dependency for ruby 2.6"
- This reverts commit https://github.com/ruby/io-wait/commit/2eb3841e9c8f.
Ruby 2.6 support has been dropped.
https://github.com/ruby/io-wait/commit/bcc343683e -
10:54 AM Bug #21506 (Closed): Fix timeout in `Addrinfo.getaddrinfo` to actually take effect
- ### Description
This change fixes an issue where the timeout option in `Addrinfo.getaddrinfo` was not functioning as expected.
https://github.com/ruby/ruby/pull/13803/files
It also addresses a related issue where specifying `fast_fa... -
10:40 AM Bug #19473: can't be called from trap context (ThreadError) is too limiting
- > Yes, for example gems like webservers like Puma set trap handlers.
The trap handler code that has already been written and published probably does not use `Mutex#lock`, so it is not related to this problem. Such benign trap handlers... -
08:37 AM Bug #19473: can't be called from trap context (ThreadError) is too limiting
- mame (Yusuke Endoh) wrote in #note-25:
> Indeed, creating a new thread for each signal arrival is extremely simple.
It feels like a hack though and it's a pretty high cost to create a new Thread for every signal.
If this was the off... -
- > Run signal handlers not on the main thread but another thread.
While I don't have a strong opinion on this, I personally feel this approach makes more sense.
However, ko1 said that users can just write this explicitly themselves.... -
- the gem a noop on Rubies older than
2.6"
This reverts commit https://github.com/ruby/io-wait/commit/75fcb74c327f.
The version that does nothing with pre-ruby 2.6 has already been
released, so there is no longer need to consider older r... -
- io-wait became a default gem at ruby 3.0.
Even it can be installed on earlier versions, but the standard library
will be loaded instead of the installed gem.
https://github.com/ruby/io-wait/commit/15b96736cd -
-
- 0a0eb2807ed7 has already replaced most of that code.
-
- Regarding thread-safety, for `add?` the only thing is the return value where the old implementation might potentially return the Set even though the element was added concurrently.
It's still thread-safe in that it doesn't corrupt the S... -
06:35 AM Bug #21396: Set#initialize should call Set#add on items passed in
- knu (Akinori MUSHA) wrote in #note-10:
> Jeremy, thanks for the reply.
> ...
I think that is a reasonable compromise. It keeps the advantages of core Set when not subclassed, and keeps backwards compatibility for subclasses (at least,... -
04:57 AM Bug #21396: Set#initialize should call Set#add on items passed in
- Jeremy, thanks for the reply.
Your point about thread-safety is well taken. It is an important advantage. As a possible compromise, we could keep the C backing and switch the behavior of methods when the class is subclassed in excha... -
- knu (Akinori MUSHA) wrote in #note-8:
> Considering the feedback we've received about compatibility in the new experimental Set implementation, it may be in our best interest to revert to the pure-Ruby version.
> ...
I don't think reve... -
- io-nonblock became a default gem at ruby 3.0.
Even it can be installed on earlier versions, but the standard library
will be loaded instead of the installed gem.
https://github.com/ruby/io-nonblock/commit/c86d0d37af - 04:48 AM Revision 4ed27575 (git): [rubygems/rubygems] Update gemspec based on provided github username when exists
- * Conditionally set changelog_url if gh username passed
and enabled
* conditionally set homepage, source code uri, homepage uri when gh
username passed in
* update documentation to say username will also be used for gemspec file
https... - 04:48 AM Revision 3feba181 (git): [rubygems/rubygems] Let s3_uri_signer accept the HTTP method
- https://github.com/rubygems/rubygems/commit/35fc7f9547
- 04:48 AM Revision 5d880b75 (git): [rubygems/rubygems] Correctly sign S3 HEAD requests
- We sometimes send HEAD requests. The s3_uri_signer.rb code allways assumed GETs.
This lead to consistently getting 403 responses back from S3. Recently, S3
attempted to change the behaviour of how 403s are handled when TCP connections
ar... - 04:48 AM Revision af6012b9 (git): [rubygems/rubygems] Fix date format in S3 URI signer
- %M is minute of the hour.
%m is month of year.
We want the former, not the latter.
https://github.com/rubygems/rubygems/commit/d7ca3fa279 -
04:48 AM Revision c6da0197 (git): [rubygems/rubygems] Add blank line after every question
- To try make output a bit less messy.
https://github.com/rubygems/rubygems/commit/92c8bc6769 -
- If we use long explanations as prompts, sometimes the prompt gets
printed twice due to a (I think) reline/readline bug.
https://github.com/rubygems/rubygems/commit/987e0dfa90 -
- Previous wording assumed explanation was displayed after the question,
not before.
https://github.com/rubygems/rubygems/commit/04eb3430ba - 03:56 AM Revision e9cd3060 (git): ZJIT: Support guarding *Exact types (#13797)
- ZJIT already can generate guard type instructions for *Exact types.
For example:
```
def test(strings)
strings.map do |string|
string.bytesize
end
end
test(["foo", "bar"])
```
```
HIR:
fn block in test:
bb0(v0:BasicObject, v1...
07/08/2025
-
-
-
-
09:12 PM Bug #21505: [Ractor] calling exit from non-main ractor will hang if main ractor waits on it
- Yeah fair enough. It's probably not as complicated to fix as I initially thought.
-
- `Kernel#exit` is basically just `raise SystemExit`, does it mean any exception in a Ractor hangs?
IOW, I don't think it should be disallowed because `exit` is just an Exception. -
- ```ruby
r = Ractor.new do
exit
end
r.join # hangs
```
Should `exit` be allowed in ractors? It would be easiest to disallow, as the code might be tricky if we allow it. -
- ivoanjo (Ivo Anjo) wrote in #note-6:
> I must admit I'm a bit torn on this. As I mentioned above, often the `.c` files are not kept around in the filesystem, whereas in my proposed version I'm showing the exact file (including path) tha... -
- matz (Yukihiro Matsumoto) wrote in #note-5:
> In my opinion, the typical Ruby user is not interested in line numbers in C source code or names of extension libraries.
I think most are or would not mind, because it is always valuable ... -
08:33 AM Feature #21501: Include native filenames in backtraces as sources for native methods
- For me the usefulness is in understanding both what and where something is happening even without needing to go into the source code.
E.g. Let's consider that native extensions didn't exist. If I get an error in production and I see a... -
03:04 AM Feature #21501: Include native filenames in backtraces as sources for native methods
- I didn't quite understand what makes you happy with this proposal. Could you explain a little more how it would be useful? In my opinion, the typical Ruby user is not interested in line numbers in C source code or names of extension libr...
-
- OK, thank you all 3 for clarifying.
I understand that point of view, even though I don't fully agree with it.
One problem is this limitation makes it impossible to support `Timeout.timeout` in `trap`, see https://github.com/ruby/time... -
11:24 AM Bug #19473: can't be called from trap context (ThreadError) is too limiting
- Locking multiple mutex with a fixed order is considered to be deadlock-safe, but permitting mutex in signal_handler, I think fixed order locking is hard or impossible.
The code below looks deadlock-safe, but it's not.
~~~ruby
trap(:... -
- If `Mutex#lock` were permitted inside a trap handler, a program like the following would have a very hard-to-reproduce race condition bug.
```ruby
m = Mutex.new
trap(:INT) { m.synchronize { p :signalled } }
while true
sleep ... -
05:10 AM Bug #19473: can't be called from trap context (ThreadError) is too limiting
- Eregon (Benoit Daloze) wrote in #note-20:
> I think a simple way to look at this issue is to analyze all cases:
> ...
In most cases, (1) occurs, so it is difficult to notice the possibility of (2).
With the current behavior, it is pos... -
07:57 PM Revision c691095f (git): ZJIT: Use BitSet in HIR
-
- 07:51 PM Revision 342ada15 (git): ZJIT: Use nil? optimization to test guard generation against different types
- 07:51 PM Revision 79915e6f (git): ZJIT: Profile `nil?` calls
- This allows ZJIT to profile `nil?` calls and create type guards for
its receiver.
- Add `zjit_profile` to `opt_nil_p` insn
- Start profiling `opt_nil_p` calls
- Use `runtime_exact_ruby_class` instead of `exact_ruby_class` to determine
... -
- The following code hangs when run with `RUBY_MAX_CPU=2 make run`:
Note: `RUBY_MAX_CPU` is set to 2 so that only 1 non-main ractor can run at once.
test.rb:
```ruby
rs = []
2.times do |i|
rs << Ractor.new(i) do |i|
if i =... -
-
-
- 04:28 PM Revision 9e4157a0 (git): ZJIT: Make type definition code more consistent
- 04:28 PM Revision af892c1b (git): ZJIT: More accurately model Class types
- 04:28 PM Revision 6c200828 (git): ZJIT: Support inference of ModuleExact type
-
- I'm leaning toward making Set subclass-friendly again. That's how I've always wanted Set to be (unlike Array and Hash), and the feedback shows that there are real users and use cases that share and rely on this concept, so we shouldn't ...
-
- Considering the feedback we've received about compatibility in the new experimental Set implementation, it may be in our best interest to revert to the pure-Ruby version.
If improving performance and reducing memory footprint remain c... -
02:34 PM Bug #21029: Prism behavior for `defined? (;x)` differs
- @S_H_ I agree, I think the overhead is not worth it in this case and we should pretend the `;` do not exist on both cases.
-
02:23 PM Bug #21029: Prism behavior for `defined? (;x)` differs
- I've fixed that `defined? (x;)` returns an `expression` when using parse.y.
https://github.com/ruby/ruby/pull/13821
However, I have a little skeptical about whether this change is a good idea. If any other good suggestions, I would app... -
- https://github.com/ruby/uri/commit/d79b3f5b94
-
- https://github.com/ruby/etc/commit/9caddede76
-
- https://github.com/ruby/delegate/commit/5ee4189537
-
- I have a different view on this proposal.
If `Array#join` were to accept a block, I would expect its behavior to relate directly to the "join" action itself, not the transformation of each element (which is the responsibility of `map`... - 07:26 AM Revision c913a635 (git): Update default gems list at b9782ab8932f9fb6a24a0522c579ca [ci skip]
-
- https://github.com/ruby/resolv/commit/a28aaed4cb
-
- RFC 1035 specifies the 255-octet maximum name length. This change set
checks the limit.
https://github.com/ruby/resolv/commit/4c2f71b5e8 -
- https://github.com/ruby/tsort/commit/bf2e3a8695
-
- https://github.com/ruby/tsort/commit/ab55dcb7f3
-
- https://github.com/ruby/tsort/commit/1d1711ad23
-
- https://github.com/ruby/dev-meeting-log/blob/master/2025/DevMeeting-2025-06-05.md#feature-21262-proposal-ractorport
> matz: go ahead. I agree to delete Ractor#take.
-
-
- There is already a PR for that: https://github.com/ruby/ruby/pull/7711
Can you take a look? @duerst @naruse -
12:11 AM Feature #20405: Inline comments
- I like Marco's proposal, but I wonder how much code out there could potentially break.
07/07/2025
- 11:46 PM Revision e0841a79 (git): ZJIT: Fix Rust warnings (#13813)
-
11:32 PM Revision 0239809a (git): Remove test/.excludes/_appveyor
- We don't run AppVeyor anymore, so we shouldn't need this.
-
- It seems nice, but I think it would be even more helpful to show lines in C files.
I think this could be achieved by defining `rb_define_method` as a macro using `__FILE__` and `__LINE__` and that should be perfectly portable (and doesn... -
- > I don't think it will need such platform specific code.
> ...
I'm thinking that in this case it would only work with methods defined when the extension gets loaded (e.g. inside `require_internal`), not later.
Having said that, thos... -
- I don't think it will need such platform specific code.
Probably it would be possible by:
* in `require_internal`, store the loading path on `struct rb_ext_config`.
* in `setup_method_cfunc_struct`, copy the path to `rb_method_cfunc_t`.
... -
-
-
06:39 PM Bug #21503: \p{Word} does not match on \p{Join_Control} while docs say it does
- What I mean is that the current implementation of `\p{Word}` does not seem to match `\p{Join_Control}` even though it should and it also says so in the docs
-
- in the [docs](https://ruby-doc.org/3.4.1/Regexp.html#:~:text=/%5Cp%7B-,Word,-%7D/%3A%20A%20member) it is mentioned that `\p{Word}` matches the equivalent of: `[\p{M}\p{Nd}\p{Pc}\p{Alpha}\p{Join_Control}]` as it's also defined in the [uni...
-
06:26 PM Revision 1f024cfd (git): ZJIT: Add opnds macro for Vec<InsnId> to Vec<Opnd> (#13805)
- Along the same lines as the `opnd` macro we already have, but for a
`Vec<InsnId>` instead of a single `InsnId`.
This gets a few for loops and `jit.get_opnd` calls out of the `gen_`
functions. - 03:45 PM Revision c1937480 (git): ZJIT: Add a simple HIR validator (#13780)
- This PR adds a simple validator for ZJIT's HIR.
See issue https://github.com/Shopify/ruby/issues/591 -
01:12 PM Revision 0bb44f29 (git): Rename `ractor_safe_set` into `concurrent_set`
- There's nothing ractor related in them, and the classic terminology
for these sort of data structures is `concurrent-*`, e.g.
concurrent hash. -
-
-
- None of the datastructures involved in the require process are
safe to call on a secondary ractor, however when autoloading
encodings, we do so from the current ractor.
So all sorts of corruption can happen when using an autoloaded
enco... -
- Prior to this commit the debug output for negative offsets would look
like:
```
Mem64[Reg(3) - -8
```
That makes it look like we're adding instead of subtracting. After this
commit we'll print:
```
Mem64[Reg(3) - 8
``` -
- https://github.com/ruby/json/commit/a497c71960
- 02:53 AM Revision e1d09ffe (git): [rubygems/rubygems] Update SPDX license list as of 2025-07-01
- https://github.com/rubygems/rubygems/commit/56b55a198a
-
- https://github.com/rubygems/rubygems/commit/265f718be7
-
- "locally to the installing Ruby installation" felt a bit confusing.
https://github.com/rubygems/rubygems/commit/c950720719 -
- https://github.com/rubygems/rubygems/commit/4b8570ae15
-
- I suspect most experienced users won't like the change in defaults, so
document the setting to toggle back the current default.
https://github.com/rubygems/rubygems/commit/93e2e2bef9 -
- It only affected the `--path` flag which is actually getting removed, so
I don't think it makes sense to make such change. The current behavior
is reasonable and I tried to codify it with a few more specs.
https://github.com/rubygems/ru... -
- And rename the shared examples to "paths dependent on gem name".
https://github.com/rubygems/rubygems/commit/cdcc8ba92a -
- So that they don't run repeatedly.
https://github.com/rubygems/rubygems/commit/1f65e879f4 -
- https://github.com/rubygems/rubygems/commit/eac831a1b7
-
- They were only being run for "edge case" names, yet it tests all kind of
things about generation.
https://github.com/rubygems/rubygems/commit/3e9d805eea -
- https://github.com/rubygems/rubygems/commit/27a4af859e
-
- Originally, all the specs in this file were put inside a shared examples
block, and since then all specs were run only changing the cwd (either
from root, or a subdirectory).
This was in https://github.com/rubygems/rubygems/commit/d7291... -
- The previous default can already be configured with `bundle config
path.system true`.
https://github.com/rubygems/rubygems/commit/cb483b79db -
- This spec now only runs in Bundler 2 mode.
https://github.com/rubygems/rubygems/commit/f52cb240ca -
- There already different ways of toggling off this behavior, like setting
`bundle config clean false`, or configuring Bundler to install to system
gems with `bundle config path.system true`.
https://github.com/rubygems/rubygems/commit/6d...
07/06/2025
-
05:20 PM Bug #21502: Segfault with 3.2.8 with frida gem
- You just beat me to it - thanks Jeremy.
-
- Reading https://github.com/hakivvi/frida-ruby/issues/2#issuecomment-3042173245, it looks like it was a bug in the gem that got fixed.
-
- Last good release running this code was 3.2.4. Segfault observed in 3.2.5 onwards - see my comment on the gem maintainer's page: https://github.com/hakivvi/frida-ruby/issues/2#issuecomment-3041319068
-
- I've just seen a segfault executing the following code in irb with the frida gem:
``` shell
$ rbenv versions
* 3.2.8 (set by /home/me/.rbenv/version)
$ gem install frida
Fetching frida-0.1.1.gem
Building native extensions. This c... - 07:04 AM Revision 4d7e6220 (git): Update bundled gems list as of 2025-07-06
07/05/2025
-
- Consider this example:
```ruby
require 'bigdecimal'
BigDecimal.singleton_class.prepend(
Module.new do
def save_rounding_mode
super
end
end
)
[:example].each do
BigDecimal.save_rounding_mode do
pu... -
05:39 PM Feature #21386: Introduce `Enumerable#join_map`
- nobu (Nobuyoshi Nakada) wrote in #note-4:
> This code would show the difference more clearly.
> ...
Hmm. I missed considering that behavior. In this example though, just the `map` would return wrapped arrays:
```ruby
[[1,2],3].map {|n|[n... -
02:27 PM Feature #21386: Introduce `Enumerable#join_map`
- My expectation is that join_map would behave like map + join
-
- This code would show the difference more clearly.
```ruby
[[1,2],3].map {|n|[n]}.join("|") #=> "1|2|3"
[[1,2],3].join_map("|") {|n|[n]} #=> "[[1, 2]]|[3]"
``` -
- prateekkish@gmail.com (Prateek Choudhary) wrote in #note-2:
> PR: https://github.com/ruby/ruby/pull/13792
This difference is intentional?
```ruby
[1,2,3].map {|n|[n]}.join(",") #=> "1,2,3"
[1,2,3].join_map(",") {|n|[n]} #=> "[1]... -
- PR: https://github.com/ruby/ruby/pull/13792
-
02:03 PM Revision b6817392 (git): [ruby/openssl] pkey/ec: avoid calling SYM2ID() on user-supplied objects
- Compare by the VALUE value instead of ID. Calling SYM2ID() on a dynamic
symbol will pin a permanent ID.
These methods only accept known static symbols, and passing anything
else is an incorrect usage that results in an exception. Noneth... -
- Ruby Parser not used LIKELY macro.
So, can remove it.
07/04/2025
-
- WBCHECK ERROR: Missed write barrier detected!
Parent object: 0x7c4a5f1f66c0 (wb_protected: true)
rb_obj_info_dump: 0x00007c4a5f1f66c0 T_IMEMO/<fields>
Reference counts - snapshot: 2, writebarrier: 0, current: 2, misse... -
-
- Previously this write barrier was using the destination object as the
new parent, rather than the fields object.
Found by wbcheck -
- When moving a "generic IV" object, we need a write barrier to the fields
object.
WBCHECK ERROR: Missed write barrier detected!
Parent object: 0x7c913641d1a0 (wb_protected: true)
rb_obj_info_dump: 0x00007c913641d1a0 T_ARRA... -
- $ RUBY_GC_LIBRARY=wbcheck ./miniruby -e 's = String.new; s.instance_variable_set(:@x, []); Ractor.make_shareable(s, copy: true)'
WBCHECK ERROR: Missed write barrier detected!
Parent object: 0x7ba8162dc890 (wb_protected: true)
... -
09:41 PM Revision 116d1106 (git): Assume that symbol in rb_check_symbol is not garbage
- rb_check_symbol is a public API, so it is always a bug if the user holds
on to a dead object and passes it in. -
- This doesn't reproduce on my machine on 3.5dev, and there have been a number of improvements to Ractors likely to have solved this. Please reopen if that's not the case :)
-
- fxn (Xavier Noria) wrote in #note-13:
> Is constant access synchronized as it is for threads?
> ...
It does! Autoloads are forwarded to the main Ractor, where a new thread is spawned to perform the autoload (or block on the other thread). -
-
- TLS 1.3 renamed the "elliptic_curves" extension to "supported_groups"
to reflect that it now covers more than just ECDH groups. OpenSSL 1.1.1
followed this change by renaming the corresponding API from
SSL_CTX_set1_curves_list() to SSL_C... - 07:05 AM Revision 4d6fac3e (git): Update bundled gems list as of 2025-07-03
-
-
-
-
- This reverts commit cf4d37fbc079116453e69cf08ea8007d0e1c73e6.
-
- This reverts commit dda5a04f2b4835582dba09ba33797258a61efafe.
-
-
-
-
- 12:43 AM Revision 68af1929 (git): Support inference of ClassExact type
- 12:43 AM Revision 0c694b56 (git): Add missed runtime_exact_ruby_class case for Regexp
07/03/2025
- 08:33 PM Revision dda5a04f (git): Make get/set default internal/external encoding lock-free
- Also, make sure autoloading of encodings is safe across ractors.
- 08:33 PM Revision cf4d37fb (git): Add locks around accesses/modifications to global encodings table
- This fixes segfaults and errors of the type "Encoding not found" when
using encoding-related methods and internal encoding c functions across
ractors.
Example of a possible segfault in release mode or assertion error in debug mode:
```... - 08:20 PM Revision 4f4408e9 (git): Get transcoding to work across ractors by locking certain operations
- Ex: `str.encode` and `str.encode!` work across ractors now.
The global table `transcoder_table` needs a lock around each st_lookup/st_insert, and it's a two-level
table so the second level also needs to be locked around insertion/deleti... - 08:09 PM Revision ed3fd94e (git): ZJIT: Panic on BOP redefinition only when needed (#13782)
- 04:40 PM Revision c584cc07 (git): ZJIT: Enable one more btest (#13781)
- 04:30 PM Revision 0abe17da (git): ZJIT: Bail out on register spill (#13773)
- 04:22 PM Revision 57f4460f (git): ZJIT: Skip a hanging ractor test (#13774)
-
- rb_sym2id is a public API, so it is always a bug if the user holds on to
a dead object and passes it in. -
-
- Add missing locks in `rb_intern_str`, `rb_id_attrset` and `rb_intern3`.
-
- Please wait a stable release. master is development branch.
-
01:42 AM Bug #21499 (Closed): make install - power_assert is a runtime_dependency of test-unit
- The below is shown in the `make install` log of master. But, `test-unit` has `power_assert` as a runtime_dependency.
```
Extracting bundle gem power_assert-2.0.5...
Try with hash version of bundled gems instead of gems/power_assert... -
- If we still fit in the existing imemo/fields object we can
update it atomically, saving a reallocation. -
-
-
-
- gcc master is still using 2021 version files.
-
- The only use case is access to `repo.or.cz`, and it redirects HTTP
requests to HTTPS now. -
-
-
-
- commit:b42afa1dbcb Suppress gcc 15 unterminated-string-initialization warnings
commit:f1f0cc14cc7 Separate `__has_attribute` from `defined(__has_attribute)`
commit:543e3a1896a Cast up `int` instruction code to `VALUE`
commit:ed2806117... -
- Seems to still be an issue. Here's an updated reproduction for 3.5dev (Using Ractor::Port instead of take/yield)
``` ruby
class Foo
attr_accessor :x
def initialize_copy(*)
$last = self
super
end
end
o = Foo.new... -
- This performance gap _in the original benchmark_ no longer exists on master. Benchmark from my Ryzen 7 3700X running linux:
```
❯ ruby -v
ruby 3.4.4 (2025-05-14 revision a38531fd3f) +PRISM [x86_64-linux]
❯ RUBY_MAX_CPU=16 ruby test.rb
...