Activity
From 04/09/2022 to 04/15/2022
04/15/2022
-
11:23 PM Revision ccb0572f (git): [rubygems/rubygems] Add modern rubies to the platforms
- https://github.com/rubygems/rubygems/commit/5462322f8f
- 10:08 PM Revision 22f79ae3 (git): [ruby/stringio] Add push task to push built gems
- https://github.com/ruby/stringio/commit/4431eefbad
- 09:35 PM Revision 566fb7fa (git): Update default gems list at 2aca566a104ad42579be03374446fd [ci skip]
- 09:34 PM Revision 2aca566a (git): [ruby/stringio] bump up to 3.0.2.pre1
- https://github.com/ruby/stringio/commit/14ec9bc193
- 06:31 PM Revision 059e389f (git): * 2022-04-16 [ci skip]
-
06:31 PM Revision e021754d (git): [DOC] Enhanced RDoc for Regexp (#5807)
- Treats:
#source
#inspect
#to_s
#casefold?
#options
#names
#named_captures -
03:52 PM Misc #18726: CI Error on c99 and c2x
- Older autoconf versions and `rb_define_method()` both cause warnings.
Having different flags during configure and compilation feels dangerous
to me since it makes the info from the checks potentially untrustworthy.
It feels reasonable... -
10:20 AM Misc #18726: CI Error on c99 and c2x
- Regarding #18615, I think it's different, we still anyway have `-Werror=implicit-function-declaration` for C extensions.
And we still have all warnings for core CRuby code as long as it's not that specific matrix entry.
Also I'd think ... -
- Is there a way we could disable that warning only for configure checks (or even the specific configure checks which need it), i.e., therefore acknowledging it's a workaround for old autoconf and not affect the rest of the codebase?
Or t... -
05:41 AM Misc #18726: CI Error on c99 and c2x
- @alanwu Thank you. I can live with that workaround. But I would also like to hear what @Eregon thinks. The fix ignores some C related warning, which seems somewhat against #18615.
-
03:47 PM Feature #18654: Enhancements to prettyprint
- @mame I have included only a small amount of changes in a smaller PR just to get this started: https://github.com/ruby/prettyprint/pull/6. For now, it only addresses the problem of allowing the user to specify whether or not a Breakable ...
-
- Thank you @mame. I can definitely split it up. I will include examples as well. If it's alright, I'll leave this ticket open and reference it.
For the most part, all of the nodes I've added are aligned with the Wadler paper. The algor... -
08:54 AM Feature #18654: Enhancements to prettyprint
- I spent some time to understand your proposal, but gave up because it was too huge. Can you split your proposal to smaller ones?
You said "There are a couple of limitations with the current approach". I can understand each of the prob... -
02:51 PM Bug #18729: Method#owner and UnboundMethod#owner are incorrect after using Module#public/protected/private
- Eregon (Benoit Daloze) wrote in #note-3:
> > I'm not sure if this is a bug, since the current behavior seems intentional.
> ...
The code deliberates handles ZSUPER methods by looking for the method entry that actually handles the metho... -
- Thank you, that looks great.
And as mentioned in the PR it also fixes `Method#super_method` to actually be the correct one in at least one case.
> I'm not sure if this is a bug, since the current behavior seems intentional.
Why do... -
02:17 PM Feature #10548: remove callcc (Callcc is now going obsoleted. Please use Fiber.)
- I know it's late to weight in on this, but see the ambit[1] and unific[2] gems and the in-progress rulog[2] gem for examples of nontrivial uses of callcc which absolutely do not translate into Fibers.
And on a lighter note, without call... -
11:09 AM Feature #18736: self-p for method chain
- That's exactly I meant. Thanks a lot, I call this topic closed. Sorry for the bad traffic.
-
09:29 AM Feature #18736: self-p for method chain
- It just looks too ugly to me. What do you gain from passing the block to `sp` instead of the method that is originally supposed to handle it? The following conventional way is much more straightforward. (In addition, in your first exampl...
-
- You may want to check object with `p` method at the middle of a method chain.
Here is my recomendation.
```ruby
class Object
def sp(method=nil, *args, &block)
if method
Kernel.p self.public_send(method, *args, &blo... -
11:07 AM Bug #18735: The crash reporter need to be updated for Apple silicon
- No you're right, my patch isn't working as well as I though it was.
> Can you confirm that your PR works well with the latest libunwind?
I'm having trouble compiling it unfortunately. I'll try more later.
But I suppose we can ... -
- I guess [this commit of libunwind](https://github.com/libunwind/libunwind/pull/330) is required to work on M1 mac. Looks like the commit is not released yet.
I couldn't build the master version of libunwind because I have no privilege... -
- I think your patch does not work well for me. Does it work well on your machine?
```
% sw_vers
ProductName: macOS
ProductVersion: 11.6
BuildVersion: 20G165
% ./miniruby -e 'Process.kill "SEGV", $$'
-e:1: [BUG] Segmentatio... -
- C backtrace dumping is only enabled for `x86_64` arch on macOS: `if defined(__APPLE__) && defined(__x86_64__) && defined(HAVE_LIBUNWIND_H)`
I tested `libunwind` on M1 Mac, it seems to work fine. I think we should either remove this `... -
10:30 AM Bug #18734 (Closed): [Win32] `File.sticky?` returns `nil` while `File::Stat.sticky?` returns `false`
- Applied in changeset commit:git|7f81f335478a3ca873f34e3bc0af6927819d3e84.
----------
Return `false` where sticky-bit is not provided [Bug #18734] -
- On the platforms where sticky bit is not provided, `File.sticky?` returns `nil` while `File::Stat.sticky?` returns `false`.
It seems changed accidentally together with `test_s()` at commit:git|86307f52ee1b3c5aa76e2fd6ee118e681dd76905. -
- Applied in changeset commit:git|a0040af6715d85f416f1282588974e151a8164eb.
----------
[Win32] Fix mode of character/pipe device stat [Bug #18732] -
- ```
$ ruby -v -e 'p ["con", "nul"].to_h {|f| [f, [File.chardev?(f), File.stat(f).chardev?]]}'
ruby 3.0.3p157 (2021-11-24 revision 3fb7d2cadc) [x64-mingw32]
{"con"=>[false, false], "nul"=>[false, false]}
``` -
09:13 AM Bug #18733: Heavy GC allocations cause performance issue with Ractor
- One problem I found was that when I ran the same script on different Ruby versions, there was performance gap between them.
``` ruby
require 'benchmark'
def fib(n)
return n if n < 2
fib(n-1) + fib(n-2)
end
puts Benchmark.measure ... -
06:22 AM Bug #18733: Heavy GC allocations cause performance issue with Ractor
- Thanks much! :)
And I change the topic to GC problem instead problem of m1.
Hope the memory management will get improvement. ;)
-
05:42 AM Bug #18733: Heavy GC allocations cause performance issue with Ractor
- Thank you for your report.
On my WSL2 environment with 12 cores, `ruby 3.2.0dev (2022-04-15T04:24:48Z master 92614111c0) [x86_64-linux]` shows worse results.
```
0.003304 0.000000 102.404055 ( 13.083221)
296.139861 262.090810... -
- Code:
```
require 'benchmark'
def fib(n)
return n if [0,1].include?(n)
fib(n-1) + fib(n-2)
end
tp = []
puts Benchmark.measure {
8.times do
tp << fork { fib(37) }
end
tp.each { |t| Process.wait(t) }
}
... -
-
-
-
07:07 AM Revision a0eb4b14 (git): [ruby/nkf] Fix docs
- rdoc parses "Z[0-3]" as a link to "0-3", this commit escapes these so
that they don't become links.
https://github.com/ruby/nkf/commit/269c10061b -
- https://github.com/ruby/optparse/commit/2bea3b38c3
-
- www.ruby-lang.org without the leading https:// will generate an
incorrect link because it will be treated as a relative link.
https://github.com/ruby/rdoc/commit/28f32149b6
04/14/2022
-
- I'm not sure if this is a bug, since the current behavior seems intentional. However, I do agree it would be best to change it. In addition to modifying `#owner`/`#inspect`, we also need to fix `#super_method` so it returns the correct...
-
- FWIW, I've found 2 CRuby tests fail with this change (through implementing `#owner` correctly on TruffleRuby):
```
TestMethod#test_prepended_public_zsuper
TestMethod#test_zsuper_private_override_instance_method
```
They are trivial ... -
- The #owner should be "the class or module that defines the method".
Or in other words, the owner is the module which has the method table containing that method.
This generally holds, and it seems very likely this assumption is relie... -
- Sorry, I pushed a CI fix without finding this ticket first:
https://github.com/ruby/ruby/commit/307cb57ba6d4833cc40d29d218561e98f9a1dcb9
Please feel free to amend it.
By the way, in case someone is looking at the new warnings from c... -
07:15 PM Misc #18726: CI Error on c99 and c2x
- shyouhei (Shyouhei Urabe) wrote in #note-1:
> Is it okay to require 2.71? Or any reasons not to?
2.69 is still default on RHEL8 and similar. -
- The problem is whether we are allowed to do this.
```patch
From b60e5fdfa540a0ddef35f5dc904d004d226dac25 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?=E5=8D=9C=E9=83=A8=E6=98=8C=E5=B9=B3?=
<shyouhei@ruby-lang.org>
Date: Thu, 14 Apr 2... -
- Is it okay to require 2.71? Or any reasons not to?
-
02:43 AM Misc #18726 (Closed): CI Error on c99 and c2x
- https://github.com/ruby/ruby/runs/6013849896?check_suite_focus=true
```
checking char bit... 8
checking size of int... 0
checking size of short... 0
checking size of long... 0
configure: error: in `/__w/ruby/ruby/build':
configure... -
- Maybe not the best idea for CI stability to use development versions of
Clang, but that does give us a preview of what's coming and gives us a
chance to make suggestions upstream. - 07:25 PM Revision b086116c (git): * 2022-04-15 [ci skip]
-
-
- The underlying issue is that `rb_utf8_str_new` and `rb_str_new` will accept a NULL pointer, where `rb_enc_interned_str` crashes on a NULL pointer. I'm not sure that is a bug in Ruby, but at least it is an inconsistency. The json gem wa...
-
- Thank you for the report. The script works fine on Ruby's builtin JSON gem. I can reproduce the crash when I install flori/json. From the stack trace generated, the crash originates from the C extension binary in flori/json, so it's like...
-
11:52 AM Bug #18728 (Third Party's Issue): Segfault since ruby 3.0.4 when parsing JSON
- This is a regression since 3.0.4. When parsing a JSON with an empty string as a key ruby segfaults.
Reproduction steps:
```
require "json"
puts JSON.parse('{ "": "foo" }')
```
leads to
```
RSTRING_PTR is returning NULL!! SI... -
- Here is a quick search of test classes with the same name in different places:
```
class Test_Bignum
class TestDateParse
class TestEmojiBreaks
class TestFileUtils
class TestFuncall
class TestGraphemeBreaksFromFile
class TestMetho... -
- In TruffleRuby I've noticed that some CRuby tests sometimes run or not, non-deterministically.
The TruffleRuby CI currently runs CRuby tests with `-j4`.
Today I investigated and I think I've found the reason for it.
One occurrence... -
02:50 PM Bug #18730 (Closed): Double `return` event handling with different tracepoints
- I'm not sure if this is a bug or intentional behavior, but feels a bit unexpected. Ruby 3.0.x, 3.1.x affected.
Here is the script demonstrating the issue:
```
def bar
42 #bp here
end
tp_line = TracePoint.new(:line) do |tp0|
... -
- Thank you for the bug report. That patch looks good, could you open a Pull Request on GitHub and I can merge.
-
06:18 AM Bug #18727 (Assigned): Make failed on x86_64-cygwin (LoadError)
-
06:08 AM Bug #18727: Make failed on x86_64-cygwin (LoadError)
- ```diff
diff --git a/include/ruby/internal/abi.h b/include/ruby/internal/abi.h
index 98a63927c5..ed779f3558 100644
--- a/include/ruby/internal/abi.h
+++ b/include/ruby/internal/abi.h
@@ -26,7 +26,7 @@
/* Windows does not suppor... -
- Make failed on x86_64-cygwin (LoadError)
```
$ git clone https://github.com/ruby/ruby.git
$ cd ruby
$ autoreconf -fiv && ./configure && make V=1
(snip)
Generating RDoc documentation
/tmp/ruby/.ext/common/pathname.rb:13:in `requi... -
12:52 PM Revision 8751c5c2 (git): [DOC] Enhance documentation for `Module#<` & `Module#>`
-
- The last parameter of `ccan_list_top()` is to acquire the pointer
of the top of element, so `node.ubf` is no problem. But this context
it accesses gvl list, so `node.gvl` is better. -
- This seems to be addressed by [PR#5774](https://github.com/ruby/ruby/pull/5774). Thx @nobu.
For the backport, please note that with this changes, the .gemspec files for expanded gems in tarball are newly located in .bundle/specificati... -
08:14 AM Misc #18687: [ANN] Upgraded bugs.ruby-lang.org to Redmine 5.0
- Oh, `_underscores_` are now back to meaning _italic_! Thank you, it was minor, but constant irritation after previous update (a year ago, I think?)
-
- @nagachika found the regression or behavior change of filter feature.
If you get the expected result like `3.0: REQUIRED` with contained filter, you need to use quote like `"3.0: REQUIRED"`.
-
- Protected characters with `PROTECT_ATTR` should not have special
roles.
https://github.com/ruby/rdoc/commit/c318af0ea2 -
- rdoc uses + for typewriter font rather than backticks.
https://github.com/rubygems/rubygems/commit/be320f1e0c -
-
- I think you can pass build-time-only flags to `make`:
```sh
$ ./autogen.sh
$ CFLAGS='-O2' ./configure
$ make ARCH_FLAG=--specs=/usr/lib/rpm/redhat/redhat-hardened-cc1
```
Then `rbconfig.rb` would not include that flag. -
-
05:51 AM Bug #18658: Need openssl 3 support for Ubuntu 22.04 (Ruby 2.7.x and 3.0.x)
- hsbt (Hiroshi SHIBATA) wrote in #note-5:
> FYI: https://bugs.launchpad.net/ubuntu/+source/ruby2.7/+bug/1946190
> ...
openssl gem v3.0 is compatible with Ruby 2.6 and later, so it's technically possible for ruby_2_7/ruby_3_0 to include ... -
- schneems (Richard Schneeman) wrote in #note-6:
> I think that would maybe not work with people using `bundle exec` to load their program as that gem is not on the load path
Ah indeed.
> ...
Is that so? We need to be sure of that f... -
-
- As underscores are masked to "protect" from the conversion, consider
also `PROTECT_ATTR` as a word character.
https://github.com/ruby/rdoc/commit/db58bb5170 -
-
01:16 AM Revision 3728f83b (git): [DOC] Fix SymbolVariantsRelNotes link
04/13/2022
-
- I'm not sure whether this is a Ruby issue, an OpenBSD issue, or something else, but @ioquatix asked me to post this here. The following program hangs on OpenBSD:
```ruby
require 'io/wait'
rd, wr = IO.pipe
thread_pass = ARGV[0] == ... -
07:51 PM Bug #12312: Make PTY take environment variables like popen does
- FYI, I have updated [PTY.spawn's doc](https://github.com/ruby/ruby/pull/5786) for the optional `env` argument.
- 07:44 PM Revision 59c81274 (git): Use correct capitalization of "NaN" in docs
-
- Passing the optional env hash to PTY.spawn has been supported for years, but it's never documented.
More info: https://bugs.ruby-lang.org/issues/12312 - 07:34 PM Revision dbb227d3 (git): [DOC] add missing size params in fiber scheduler.h (#5441)
-
- Treats:
#[]
#length
#empty?
#upcase
#downcase
#capitalize
#swapcase
#start_with?
#end_with?
#encoding
::all_symbols -
05:41 PM Bug #16672 (Closed): net/http leaves original content-length header intact after inflating response
- Applied in changeset commit:git|0579486f154e80d17521494003dcd2499ef74688.
----------
[ruby/net-http] Update the content-length heading when decoding bodies
Previously, the content-encoding header was removed and the body
was modified, ... -
04:07 PM Bug #18658: Need openssl 3 support for Ubuntu 22.04 (Ruby 2.7.x and 3.0.x)
- > After that, I can use gem install rails or anything as usual.
Thank you for the idea Mame. I think that would maybe not work with people using `bundle exec` to load their program as that gem is not on the load path, but I could cons... - 03:11 PM Revision fde79524 (git): * 2022-04-14 [ci skip]
-
- Previously, the content-encoding header was removed and the body
was modified, but the content-length header was not modified,
resulting in the content-length header not matching the body
length.
Don't delete content-length before yield... -
-
-
- Only valid characters for URLs should be used for generating URLs.
A list of valid characters can be found in sections 2.2 and 2.3 of IETF
RFC 3986 (https://www.ietf.org/rfc/rfc3986.txt).
https://github.com/ruby/rdoc/commit/2bd8fcdd4f -
-
- It's fully expected, `a += b` is just `a = a + b` and so it's N Set instances/copies created vs 1 with <<.
-
08:42 AM Bug #18692 (Rejected): Set += is O(n^2)
$ time ruby26 --disable=gems -r set -e 's = Set.new(); 1.upto(10_000).each { |n| s += [n] }'
real 0m1.346s
user 0m1.290s
sys 0m0.055s
Three times the input size:
$ time ruby26 --disable=gems -...-
- - [Feature #18683] Allow to create hashes with a specific capacity.
- Useful for various parsers and deserializers that know the Hash size in advance, such as Redis protocol, msgpack, etc.
- When creating large hashes of known size... -
-
-
- https://github.com/ruby/ruby/commit/cbb115213c42f15638ef119eb20c4d3106eb8b1a
https://github.com/ruby/ruby/commit/5c1b76a3a55afeb07116bbd3492303c6b6cd890d
04/12/2022
- 10:27 PM Revision af855538 (git): * 2022-04-13 [ci skip]
-
- Treats:
#==
#inspect
#name
#to_s
#to_sym
#to_proc
#succ
#<=>
#casecmp
#casecmp?
#=~
#match
#match? -
04:44 PM Misc #18691 (Feedback): An option to build Ruby with build only flags not propagated to `rbconfig.rb`.
- In a Fedora Ruby RPM packaging, I have a challenge for the current `make rbconfig.rb`. Currently when we build Ruby from source to create the Ruby RPM package, we set compiler flag `CFLAGS` including `--specs=/usr/lib/rpm/redhat/redhat-h...
-
- Allocating a string of length MAXPATHLEN and then shrinking the string
is inefficient when the resulting path is short. Preallocating a large
string is also a problem for Variable Width Allocation since we can't
easily downsize the capac... -
- Applied in changeset commit:git|ebb4378237e572ce2e888136a613c7c051439f95.
----------
[ruby/net-http] Add HTTP#response_body_encoding for setting response body encoding
This allows for the ability to opt-in to a method to set the
encodi... -
- Applied in changeset commit:git|ebb4378237e572ce2e888136a613c7c051439f95.
----------
[ruby/net-http] Add HTTP#response_body_encoding for setting response body encoding
This allows for the ability to opt-in to a method to set the
encodi... -
- https://hackerone.com/reports/1248108
- 11:49 AM Revision 69f9992e (git): Fix dtoa buffer overrun
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_6@67957 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
11:48 AM Revision 3fa771dd (git): Fix dtoa buffer overrun
-
- 11:25 AM Revision c9c2245c (git): Fix dtoa buffer overrun
-
- https://hackerone.com/reports/1220911
-
11:11 AM Revision 4491bb74 (git): teenyup
- 11:03 AM Revision fa7f75dd (git): Backport date-2.0.3 for Reiwa Support [Backport #18514]
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_6@67956 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
-
- 10:52 AM Revision 560a4cdf (git): fix for #16798 to Ruby 2.6 introduced C99 syntax [Backport #18387]
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_6@67955 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
-
- The last example is just:
```ruby
p.call(honyarara, fugafugafuga, hogehogehoge)
```
isn't it? And that's a lot more readable IMHO.
I'm against this proposal, IMHO having multiple local variables for the same thing only increases c... -
- `Kernel#then` passes the receiver to the block as its first positional block parameter.
```ruby
1.5.then{|x| Math.atan(x)}
```
I would like to propose to let `then` take arguments, which would be passed to the block as the other ... -
- I do not understand the logic behind option 2. I would rather expect this logic:
> a beginless-and-endless range has no type, so it imposes no restriction on the type
which would lead us to option 1. Its coverage should return `tru... - 05:07 AM Revision 47c3f31f (git): * 2022-04-12 [ci skip]
-
05:07 AM Revision a6d99089 (git): Fix spec to not touch the network
- And not depend on the state of rack's master branch, in particular, on
their Ruby support range. -
- `RSpec::Mocks::ArgumentListMatcher#args_match?` fails when a
non-keywords hash is passed while a keyword hash is expected.
This is a change in `rspec-mocks` 3.10.3, so we need to adapt to it. -
-
04/11/2022
-
10:14 PM Misc #18689 (Closed): Windows mswin builds, Actions, OpenSSL 3
- Sorry for the long post and the md formatting...
With the release of the Actions windows-2022 image, MSYS2 build tools are no longer pre-installed on the image. Given that updating MSYS's msys and mingw64/ucrt64 tools can take quite ... -
-
-
- Previously, `make test-spec` was not printing the description with +YJIT
even when YJIT was indeed enabled. It was confusing on CI. `fake.rb` was
changing the `RUBY_DESCRIPTION` constant incorrectly.
I suppose `make test-spec` mostly ne... -
- * [Bug #18628] Link contributing is broken (jeremyevans0)
* Also affects #18665. Problem is that ruby uses the --page-dir rdoc option.
* Any documentation generation that does not use the --page-dir rdoc option has broken links.
... - 03:17 PM Revision 492349ab (git): * 2022-04-12 [ci skip]
-
- This allows for the ability to opt-in to a method to set the
encoding of response bodies. By setting the accessor to a String
or Encoding instance, it will use the specified encoding.
Setting the value of true will try to detect the enc... -
- Thanks for the ping. I've had some vacation recently. Looks like I missed the first preview release. I'll be able to start looking at this shortly.
-
- FYI: https://bugs.launchpad.net/ubuntu/+source/ruby2.7/+bug/1946190
Ubuntu backported openssl-3.0.0 for their ruby3.0 package.
-
- @schneems How about manually downloading and installing openssl-3.0.0.gem?
```
wget https://cache.ruby-lang.org/pub/ruby/3.0/ruby-3.0.3.tar.gz
tar xf ruby-3.0.3.tar.gz
cd ruby-3.0.3
./configure --prefix=/path/to ...
make
make install
wg... -
- I will mark this to closed because the basic feature works now.
- 02:31 AM Revision 4bd38e81 (git): * 2022-04-11 [ci skip]
-
04/10/2022
- 01:34 PM Revision abe2bb1d (git): * 2022-04-10 [ci skip]
-
- This covers the first few sections of the class doc for Array.
-
06:26 AM Feature #17849: Fix Timeout.timeout so that it can be used in threaded Web servers
- The handling of timeouts is [documented][1], but that only shows it with a `handle_interrupt` block set up around everything, not under `ensure`. In the examples shown here, can `ensure;begin` or `ensure;Thread.handle_interrupt` be consi...
-
- This is expected behavior and not a bug. As you mentioned, you want to use the block form if you want a new object for each default value.
-
04:38 AM Bug #18688 (Rejected): when array's default value is empty hash adding a hash key value changes all array elements
- ```ruby
ah = Array.new(3, {})
ah[1][:foo] = 'bar'
p ah # [{:foo=>"bar"}, {:foo=>"bar"}, {:foo=>"bar"}]
```
This is not expected result. The expected result is `[{}, {:foo=>"bar"}, {}]`.
Alternatively, below code works as expecte...
04/09/2022
-
-
- I fixed the issue of mailing-list integration.
-
- I found the mailing-list integration is broken now.
-
- Today, I migrated this tracker Redmine 4.2 to 5.0.
It has the following major changes.
* Based on Rails 6.1.
* Allowed username mention like `@hsbt` with autocomplete.
* Support CommonMaker for Markdown formatting.
Release not...