Activity
From 10/02/2020 to 10/08/2020
10/08/2020
-
11:43 PM Revision bca8952f (git): Fix lldb disassembler so it works with core files
- This fixes the lldb disassembler script so that it doesn't need a live
process when disassembling rb_iseq_t. I also added the PC to the output
so you can tell what the VM is executing when it crashed.
For example:
```
(lldb) rbdisasm ... - 08:35 PM Revision c857d0d4 (git): * 2020-10-09 [ci skip]
-
08:35 PM Revision 33776598 (git): Enhanced RDoc for String#insert (#3643)
- * Enhanced RDoc for String#insert
-
10:30 AM Bug #17255 (Closed): Installation failed for ruby version 2.7.1: unknown encoding name: binary (ArgumentError)
- Hi,
I am trying to install version 2.7.1 of ruby. For now, ruby is not installed in my system (I removed it) and I used *ruby-build* to install it.
```
# ruby-build 2.7.1 /usr/local
[...]
```
*Attached log file.*
By lookin... -
10:20 AM Bug #17254 (Closed): ENV.replace may set nil instead of the proper value
- On docs.ruby-lang.org, it uses [snap ruby](https://snapcraft.io/ruby), and it failed to run `rdoc`.
```
Oct 07 13:20:08 docs-2020.ruby-lang.org env[6183]: rdoc --title Documentation for Ruby master --main README.md --output /var/www/... -
09:34 AM Bug #17221: Relax the Fiber#transfer's limitation
- For (2), can we use the error message `attempt to transfer to a yielding fiber (FiberError)` (yielding instead of resuming)? That seems clearer since f2 is not resuming but waiting for a `resume` and inside `Fiber.yield` (= yielding).
-
12:48 AM Bug #17221: Relax the Fiber#transfer's limitation
- I found "(4) can not yield from not-resumed fiber", so I updated the proposal body with (4).
-
07:40 AM Revision 533bca57 (git): Expose assert, assert_respond_to and assert_not_respond_to for default gems.
-
-
- 12:01 AM Revision 0807d05d (git): * 2020-10-08 [ci skip]
-
- The write barrier wasn't being called for this object, so add the
missing WB. Automatic compaction moved the reference because it didn't
know about the relationship (that's how I found the missing WB).
10/07/2020
-
- Note for (2). It can break compatibility, but transferred fibers can not be resumed more, so maybe nobody rely on mixing them.
```ruby
require 'fiber'
f = Fiber.new do
loop do
Fiber.yield :ok
end
end
p f.resume
p f... -
10:31 PM Bug #17221: Relax the Fiber#transfer's limitation
- Great work everyone!
-
- This sounds great to me!
-
- ## Implementation note
Each fiber has "resuming_fiber" and we can check (1).
-
- Using `Fiber#transfer` with `Fiber#resume` for a same Fiber is limited (once `Fiber#transfer` is called for a fiber, the fiber can not be resumed more).
```ruby
require 'fiber'
f1 = nil
f2 = Fiber.new{
f1.transfer
}
f1 = Fiber... -
- The term "concurrency" seems way too abstract to me for this scheduler.
I think the more obvious the better, this schedules Fibers, so I think `Thread#fiber_scheduler{,=}` and `Fiber.schedule {}` fit nicely. -
06:23 PM Bug #17175: Ruby 2.5: OpenSSL related test failures
- jaruga (Jun Aruga) wrote in #note-8:
> Now every test passes on my environment. Could you backport the total 4 patches (commits) on ruby_2_6 branch to ruby_2_5 branch?
Ruby 2.5 is currently in security maintenance mode (see https://w... -
-
-
- Fixed by commit:20ad1017017ea736667d86fa0250dc1a39daefa1.
-
-
- Fixed in commit:f7bd9f075030e5fa876320c1624a80685a636e82
-
- Changing this to a bug, because it can cause unsafe sharing:
```ruby
irb(main):003:0> Encoding::UTF_8.instance_variable_set(:@foo, Object.new)
=> #<Object:0x00005596506b8a28>
irb(main):004:0> Ractor.new { p "b".encoding.instance_vari... -
01:26 PM Bug #17220 (Closed): Rails Active Job integration test fails with Ruby 3.0.0 since 2038cc6cab6ceeffef3ec3a765c70ae684f829ed
- One of the Rails CI, Active Job integration test with sidekiq, against Ruby 3.0.0 has been failing since August 30, 2020.
According to `git bisect` is is triggered by 2038cc6cab6ceeffef3ec3a765c70ae684f829ed . Somehow this issue only ... -
11:24 AM Feature #17219: Make URI#hostname 86-89% and hostname= 55-154% faster
- So I learned that URI is a separate rubygem, therefore I filed PR against https://github.com/ruby/uri/pull/12 and this RM issue is no longer necessary I guess. Please close.
-
- There is a nil exception in the test suite with the initial patch. Also @nobu on Github pointed out that start_with? and end_with? might be faster which is indeed the case. I will update the patch shortly.
-
- Attaching patch. Also available as https://github.com/ruby/ruby/pull/3635
-
- Hello,
URI#hostname extends URI#host with IPv6 support. In URI, IPv6 address must have square brackets (e.g. `http://[2001:db8::1]`), URI#hostname strips these characters out while URI#hostname= adds them if missing. There are three r... -
10:59 AM Bug #17215: Backport for arm64 optimizations that exist for power/x86
- Thanks for explaining the context of "bug" in CRuby.
In the context, I think the commit is to fix the relatively slower performance by the optimization, while there are already x86 and powerpc applying the optimization. So, I change the... -
10:43 AM Bug #17215: Backport for arm64 optimizations that exist for power/x86
- As you know Ruby only backports bugfixes. If you think this is a feature, it won't be backported.
But note that in CRuby context "bug" is defined as fixes which may be backported.
If you believe a fix is a bug, you can request to bac... -
- Hi Ali, thanks for your help!
Please note I think the commit is a new feature, and the ticket with "Tracker: Feature" does not have the backport field to make the backport easy. But if someone likes changing the Tracker type from Feat... -
08:01 AM Bug #17213 (Open): Rails CI against Ruby 3.0.0dev fails likely since a53e2850c572135ed657144bc14e47b29c64fa94
-
- It reproduced again with this newer version of Ruby 3.0.0p-1.
https://buildkite.com/rails/rails/builds/71972#920c6233-8bd8-4af9-8375-8df0ebbf0956
Ruby 3.0.0p-1 (2020-10-07 revision 260322fbc58a225d1c0323d53698ff62a31cf552... -
02:27 AM Revision 62abdbad (git): Revert "Don't export rb_callable_receiver"
- This reverts commit c839168b1141db53bedef771d1bc78908b6ac782.
`rb_callable_receiver` does not need to be exposed under include.
10/06/2020
-
- This iterator uses an st_table, but if objects move the references in
the st table won't be updated. This patch just changes the st table to
an identity hash. - 10:41 PM Revision c839168b (git): Don't export rb_callable_receiver
- 10:41 PM Revision fef52122 (git): Use proc_binding rather than rb_funcall
- FIX
-
- Ruby specs expected this method to be public
- 09:11 PM Revision 1486785a (git): [lib/ostruct] Fix Marshal loading
-
09:06 PM Feature #13683: Add strict Enumerable#single
- For Arrays, pattern matching does seem like a fair solution. Wouldn't that not work for other Enumerable collections?
``` ruby
Set.new([42]) in [one] #!> NoMatchingPatternError (#<Set: {42}>)
```
It seems an Enumerable method would be ... -
08:46 PM Bug #17215: Backport for arm64 optimizations that exist for power/x86
- jaruga (Jun Aruga) wrote:
> Is it possible to backport this commit to ruby_2_7?
> ...
I'm happy to submit a backport. I've tried it and it's pretty trivial. -
04:06 PM Bug #17218 (Assigned): Range#step sometimes behaves unexpectedly with Rational endpoints and increment
-
03:46 PM Bug #17218 (Closed): Range#step sometimes behaves unexpectedly with Rational endpoints and increment
- I am using the latest stable version. The same behavior exists in 2.6.6.
```
11:28:25 $ ruby -v ... - 03:04 PM Revision 260322fb (git): * 2020-10-07 [ci skip]
-
- As debug_counter.c has had a global symbol since cdc614cd0a21,
ranlib should no longer complain that it has no symbols. -
-
11:19 AM Misc #17174: "Error relocating, symbol not found" error when compiling a native extension on Alpine with Ruby >=2.4
- Nakilon (Victor Maslov) wrote:
> docker build - -t temp-ruby2.3.8 --build-arg RUBY_VERSION=2.3.8 --build-arg ALPINE_VERSION=3.8 <Dockerfile
Also tried the same version of Alpine but higher Ruby -- ruby:2.4.6-alpine3.8 -- got the same... -
09:39 AM Misc #17200: DevelopersMeeting20201026Japan
- * [Feature #16746] Endless method definition
* The following was discussed at [DevelopersMeeting20200514Japan](https://github.com/ruby/dev-meeting-log/blob/master/DevelopersMeeting20200514Japan.md)
> Conclusion:
> ...
* [Feature... -
-
-
- This was a workaround for RUBY_ALIGNOF's glitch, which has already been
fixed. See also https://github.com/ruby/ruby/pull/3570 -
- Because `bool` is already used in the header there is no reason to
hesitate. -
- Bit readable to me.
-
- This adds more room for assertions.
-
- cf: https://github.com/ruby/ruby/pull/2991/commits/99add258571bf103c6d942bf0e4d510763b73918
-
03:36 AM Bug #17211: Test failures in ruby2.7.2 and ruby3.0~preview1
- Hello,
aycabta (aycabta .) wrote in #note-6:
> I decided to stop using `system` method for E2E testing because it depends on `ruby` command. I'm sorry for the rough implementation.
> ...
Thank you so much! I can confirm that this in... -
- Okay, considering that this proposal is about concurrency (a subset of asynchronous execution), we could use the following taxonomy:
```
Thread#concurrency_scheduler=
Thread#concurrency_scheduler
Fiber.concurrently do
end
```
To me, t... -
03:18 AM Bug #16674: Build issue of FreeBSD 12
- This could be fixed by using LLVM's LLD after rG88498f44dfe7d9b886f2622335cdeae4dbf2b02a
Reference:
[ELF] -r: allow SHT_X86_64_UNWIND to be merged into SHT_PROGBITS
https://reviews.llvm.org/D85785
-
- Please backport https://github.com/ruby/ruby/pull/3630
`set-env` is deprecated, and it will be disabled in the future.
https://github.blog/changelog/2020-10-01-github-actions-deprecating-set-env-and-add-path-commands/
> We are mon... -
- https://github.blog/changelog/2020-10-01-github-actions-deprecating-set-env-and-add-path-commands/
-
-
-
01:10 AM Revision 29c54b5e (git): Add missing fileutils require
- On my system, the error was being hidden by the presence of a YARD
rubygems plugin that was providing the require and making things work.
10/05/2020
-
05:59 PM Bug #17216 (Closed): Enumerator::Chain doesn't support all Enumerator methods
- Despite the fact that `Enumerator::Chain` (returned by `Enumerator#chain`) is supposed to inherit from `Enumerator`, a lot of `Enumerator` methods do not work on a chain instance throwing the following error:
```
TypeError (wrong arg... - 05:17 PM Revision 49bfd889 (git): * 2020-10-06 [ci skip]
-
- https://github.com/ruby/tempfile/commit/8bac025065
-
- Is it possible to backport this commit to ruby_2_7?
Enable arm64 optimizations that exist for power/x86
https://github.com/ruby/ruby/commit/511b55bcefc81c036294dc9a544d14bd342acd3b
The commit comes from https://github.com/ruby/ruby/... -
02:35 PM Revision 3a3000f5 (git): Fix traditional Ring example in Actor-model
-
- I decided to stop using `system` method for E2E testing because it depends on `ruby` command. I'm sorry for the rough implementation.
https://github.com/ruby/ruby/pull/3628
This patch of this Pull Request resolves the failure of CI. -
- FreeBSD 13 については、https://bugs.ruby-lang.org/issues/16674 にあるように `--disable-dtrace` つけるとどうでしょうか?
-
11:15 AM Bug #17214: BigDecimal exponentiation gives incorrect results
- If I increase the precision of the two arguments ("2222", "3.5"), but leave the operation's precision on default, the result still will be wrong.
``` ruby
(BigDecimal("2222",10000) ** BigDecimal("3.5",10000)).to_i
# => 517135311000
```
T... -
- `BigDecimal` instances hold the precision, and it is not enough in that case.
You need to override the default precision in some operations.
```ruby
BigDecimal(2222).power(3.5, Float::DIG).to_i #=> 517135308457
```
-
- This is an incorrect value:
```ruby
(BigDecimal("2222") ** BigDecimal("3.5")).to_i
# => 517135311000
```
This is the correct value (within Float precision):
```ruby
2222 ** 3.5
# => 517135308457.25256
```
As the Base ge... -
-
- Even if the raw commit log does not end with a newline.
Suggested to use `grep` by znz.
Co-Authored-By: Kazuhiro NISHIYAMA <zn@mbf.nifty.com>
10/04/2020
- 07:32 PM Revision d5ab1979 (git): * 2020-10-05 [ci skip]
-
-
- Okay, I got it. Those tests fail when no executable named `ruby` is available in `PATH`.
https://github.com/ruby/ruby/blob/f8c50109d7d1a067bae80b56e7b0b1923951c068/test/irb/test_history.rb#L131
Note that the build directory is adde... -
- Thanks. I tested `make && make check` on Debian sid x86_64 on Docker on WSL2, but couldn't reproduce the issue. I'm not familiar with the deb packaging process, so can you explain how to reproduce the issue?
-
- utkarsh (Utkarsh Gupta) wrote:
> I am clueless as to what's causing this?
Weirdly enough it looks to me that
```
IRB_HISTORY
stdin.write("5\nexit\n")
end
```
in `test/irb/test_history.rb` (for all these tests) has no impact a... -
- Hey,
mame (Yusuke Endoh) wrote in #note-1:
> Hi, cloud you tell us what environment (OS, distribution, version, CPU, terminal, etc.) you are using? Thanks
Right, sorry for not mentioning this earlier.
I am running Debian Sid. Int... - 02:00 PM Revision f8c50109 (git): * 2020-10-04 [ci skip]
-
- * sync_default_gems.rb (sync_lib): sync from the same directory as
sync_default_gems. -
-
- It has not reproduced since https://buildkite.com/rails/rails/builds/71832#5716101b-5e5b-49f8-90c4-4b359c8b1f96 . The last failed one is https://buildkite.com/rails/rails/builds/71830#2476cfb2-dc3b-4e4e-a891-7a62ecf42ec9
- The last fa...
10/03/2020
-
- Rails CI against Ruby 3.0.0-dev (or 2.8.0dev) had been failing because sprockets and sprockets-rails gem did not support Ruby 3 keyword arguments then I (at least) had overlooked this segmentation fault. Recently, both sprockets and spro...
-
- Rails CI against Ruby 3.0.0dev fails likely since a53e2850c572135ed657144bc14e47b29c64fa94 . This is the oldest failure https://buildkite.com/rails/rails/builds/69461#777c707f-7056-4959-9924-36dac592b737
## Steps to reproduce
1. In... -
- DATA_PTR(ractor) can be NULL just after creation.
-
-
08:14 AM Bug #17178 (Closed): Procs with kw:def/**kw lose elements when called with a single Array
- I think it should be as it is. It would break compatibility from Ruby2.x. It's not that big issue worh breaking.
Matz.
-
- I am still strongly against `Thread#scheduler`. As @eregon stated, we are not going to have other schedulers any time soon. But the concept of scheduler without scheduling target could confuse users. It should be qualified from the start...
-
07:32 AM Misc #17200: DevelopersMeeting20201026Japan
- * [Feature #17187] Add connect_timeout to TCPSocket (glass)
* Could I merge it? -
- This might be a pointer alignment issue / problem with the alloca elision.
After playing around with godbolt compiler explorer, I think this might be one option:
https://github.com/ruby/ruby/pull/3624
However, I wouldn't be surp... -
- Hi, cloud you tell us what environment (OS, distribution, version, CPU, terminal, etc.) you are using? Thanks
-
-
03:17 AM Bug #17212 (Closed): FreeBSDで3.0.0-preview1のビルドが失敗する
- 題名の通り、FreeBSDで3.0.0-preview1のビルドが失敗します。
再現手順は以下の通り
1. tar xfpvJ ruby-3.0.0-preview1.tar.xz
2. cd ruby-3.0.0-preview1
3. ./configure
4. make
FreeBSD 12.1-RELEASE amd64では、以下のようなエラーとなります。
```
building .ext/include/x86_64-fre... -
10/02/2020
-
- Hello,
Whilst building Ruby 2.7.2 and Ruby 3.0~preview1, I got the following test failures:
```
1) Failure:
TestIRB::TestHistory#test_history_save_1 [/<<PKGBUILDDIR>>/test/irb/test_history.rb:31]:
<"exit\n"> expected but was
<"... -
-
- Since Thread.scheduler is removed, I think Thread#scheduler is unambiguous: it's a scheduler per-Thread, so a scheduler of Fibers.
Anyway, I don't imagine a useful thread scheduler or process scheduler written in Ruby anytime soon, th... -
-
-
04:10 PM Feature #17145: Ractor-aware `Object#deep_freeze`
- Maybe we can start with `Ractor.shareable(obj)`?
The major reason why we need it is that there is no way to create a Ractor sharable recursive structure currently.
```ruby
child = {type: :foo}
node = {type: bar, children: [child]... -
08:44 AM Feature #17145: Ractor-aware `Object#deep_freeze`
- I'd really like to support this change for reasons I've describe in https://bugs.ruby-lang.org/issues/17180
We can take something as simple as `URI::parse("http://example.com")` as an example. Right now that can't be used from a Ractor ... -
- https://github.com/ruby/ruby/pull/3622
-
- 9eda6547812cbda23a73ba3b2620520b0de2bdd6 was fixed by
b9488accf9e2cbf5f7c47b42b3eb23469f0aa58d. - 03:00 PM Revision 0d62e320 (git): * 2020-10-03 [ci skip]
-
- RDoc says `ObjectSpace.dump(obj, output: :stdout) # => nil`,
but it returns STDOUT since fbba6bd4e3dff7a61965208fecae908f10c4edbe.
I think it is unintentional change. -
- What about pattern matching?
```ruby
case []; in [a]; p a; end #=> NoMatchingPatternError ([])
case [1]; in [a]; p a; end #=> 1
case [1,2]; in [a]; p a; end #=> NoMatchingPatternError ([1, 2])
``` -
-
09:22 AM Feature #17210: More readable and useful `Set#inspect`
- BTW, as a side note, it would be good to have some "style guide" for inspect behavior in Ruby. Ad-hoc rule seem to be `#<ClassName some representation of content>` most of the time, except for objects that have corresponding literals -- ...
-
- I would like to change `Set#inspect`/`to_s`:
```ruby
# before
puts Set[1,2,3] # => "#<Set: {1, 2, 3}>"
# after
puts Set[1,2,3] # => "Set[1, 2, 3]"
```
This output is shorter, readable, and has the property that it correspond... - 06:04 AM Revision 74aaa8e7 (git): Update marshal.rdoc
-
- Applied in changeset commit:git|89ca842dcce7f05942e2d7be3edc404c9556cafd.
----------
Ensure that the comparison succeeded [Bug #17205] -
- shyouhei (Shyouhei Urabe) wrote in #note-2:
> I like the feature, but isn't "Compact Set" a mathematical concept that refers to something different?
It can, but the idea requires infinite sets, and all elements belonging to a topolog... -
- I like the feature, but isn't "Compact Set" a mathematical concept that refers to something different?
-
- ko1 (Koichi Sasada) wrote in #note-6:
> shyouhei (Shyouhei Urabe) wrote in #note-3:
> ...
The proposed patch is basically `Primitive.cexpr! 'int_size(self)'`. Why it must be faster than `rb_define_method(rb_cInteger, "size", int_size,... -
- shyouhei (Shyouhei Urabe) wrote in #note-3:
> The patch looks good to me. HOWEVER, let me -1 this.
> ...
what is the "extra hustle"? -
12:13 AM Feature #17135: Improve performance of Integer#size method
- `INT2FIX(sizeof(long)` is a constant expression because `INT2FIX` is a constant inline function that only does some arithmetic operations. So, I guess you don't need to keep that value in `rb_cInteger_fix_size` global variable.
-
-
- I forgot to merge it: https://github.com/ruby/ruby/pull/3587
-
- Threads can move, and if they do, their self pointer may go bad. We
need to update it. -
- Range literal is now frozen so we can reuse same Range object if
the begin and the last are Numeric (frozen), such as `(1..2)`.