Activity
From 08/28/2019 to 09/03/2019
09/03/2019
-
11:04 PM Bug #16135 (Feedback): Segmentation fault in Ruby 2.6.1
- From the C-level backtrace, it looks like this may be caused by AMD Kysy. Please try uninstalling AMD Kysy and see if the problem still occurs. If it still occurs, you should also try upgrading to 2.6.4 (or at least 2.6.3) to see if the...
-
05:56 AM Bug #16135 (Third Party's Issue): Segmentation fault in Ruby 2.6.1
- I am facing Segmentation fault issue while running test cases for my application in Ruby 2.6.1 .Same test cases work fine in Ruby 2.2.6 .
Below is the output of test case ran using Ruby 2.6.1.
```
C:/Code/KYSY-3153/Source/test/unit... -
- 77596fb7a91cc119b25ac9e19b3c8682709765b4 should be backported to 2.5 and 2.6, as k0kubun mentioned in https://github.com/ruby/ruby/pull/2423.
-
09:02 PM Revision 77596fb7 (git): Do not turn on keyword_init for Struct subclass if keyword hash is empty
- This was accidentally turned on because there was no checking for
Qundef.
Also, since only a single keyword is currently supported, simplify
the rb_get_kwargs call. -
- Add rb_keyword_given_p to the C-API
- 06:31 PM Revision f702cd6a (git): * 2019-09-04 [ci skip]
-
- Previously, passing to_enum/enum_for a method that was defined in
Lazy itself returned wrong results:
[1,2,3].to_enum(:map).to_a
# => [1, 2, 3]
[1,2,3].lazy.to_enum(:map).to_a
# => []
I'm not sure why methods that are designed ... -
- This spec should not be checking where methods are defined, only
that the method works as expected (returns a Lazy instance). -
- Previously, Enumerator::Lazy#with_index was not defined, so it
picked up the default implementation from Enumerator, which was
not lazy.
Based on earlier patch from nobu.
Fixes [Bug #7877] -
04:42 PM Feature #16029: Expose fstring related APIs to C-extensions
- @ko1 and @nobu discussed the issue. The memo says:
> ko1, nobu: now it is difficult to expose them because of current implementation.
I didn't follow the discussion, so I don't know the detail. @ko1 , @nobu , could you explain them? -
03:32 PM Feature #16029: Expose fstring related APIs to C-extensions
- @mame this feature was added by @nobu in the last developer's meeting issue [#15996], but I suppose it wasn't discussed by lack of time.
Should I just add it back to the next developer's meeting issue once it is created? -
03:56 PM Feature #16122: Data: simple immutable value object
- zverok (Victor Shepelev) wrote:
> **Why not a gem?**
> ...
If a concept is popular and there is a well-designed gem that implements it then people use it. For example, a lot of people use `dry-initializer`, which is also dead-simple an... -
02:50 PM Feature #16137: Add === to UnboundMethod
- Since I love case/when statements in ruby, and I also like the idea to decouple methods
at "runtime", I agree with okuramasafumi.
I can not add much to potential use cases or what osyo asked, but personally I just like
the idea. :) -
12:58 PM Feature #16137: Add === to UnboundMethod
- hi.
How about making `# bind_call` an alias of` # === `?
```ruby
class UnboundMethod
alias_method :===, :bind_call
end
```
see: https://bugs.ruby-lang.org/issues/15955#note-10
Also, you should consider the difference be... -
12:28 PM Feature #16137 (Closed): Add === to UnboundMethod
- # Abstract
`UnboundMethod` class should have `=== ` so that it can be used in case statement.
# Background
`Method` class has `===` method so that we can do something like:
``` ruby
require 'prime'
case 11
when Prime.method(... -
12:49 PM Revision fb67d4fc (git): Move an unstable bootstraptest to pending
- This has been unstable on AppVeyor mswin since the introduction
3fd83cb6fcc483d2eac0795bc139c521a3a59bd2.
https://ci.appveyor.com/project/ruby/ruby/builds/27103307/job/j7xwjmsos2k22cck
Let's have it in pending.rb to be fixed. -
-
- for simplicity and consistency.
Now SUPPORT_JOKE needs to be prefixed with OPT_ to make the config
visible in `RubyVM::VmOptsH`, and the inconsistency was introduced.
As it has never been available for override in configure (no #ifndef... -
- to deal with random failures:
https://github.com/ruby/ruby/runs/210617845 -
- hoping to stabilize:
https://app.wercker.com/ruby/ruby/runs/mjit-test1/5d6df8a8a952c20008acf75b?step=5d6df90e4971a6000714c627 -
11:24 AM Bug #16136 (Closed): String corruption in 2.6.4
- When trying to upgrade activeadmin's tests to use ruby 2.6.4, I got some very weird failures where some strings would end up containing random content. See https://circleci.com/gh/activeadmin/activeadmin/20329 for an example of those fai...
-
09:56 AM Bug #16007: Process.clock_getres matches the clock in practice for Process::CLOCK_{PROCESS,THREAD}_CPUTIME_ID FAILED fails on armv7hl
- It should be possible to get some hardware here:
https://fedoraproject.org/wiki/Test_Machine_Resources_For_Package_Maintainers
I am afraid I don't have any knowledge about `Process::CLOCK_{PROCESS,THREAD}_CPUTIME_ID` -
09:36 AM Revision 36a0c668 (git): Remove SKIP_DOXYGEN because job.env does not set in `if:`
-
08:02 AM Feature #15901: Enumerator::Lazy#eager
- Thanks, I see your point. It could be handy, but it seems to me that adding Enumerable#eager is like encouraging users to consider calling `enum.eager` on every given enumerable object and supporting Lazy objects to be passed around; th...
-
04:03 AM Feature #15901: Enumerator::Lazy#eager
- knu (Akinori MUSHA) wrote:
> > I'd also alias Enumerator#eager to #itself.
> ...
Yes, I was thinking of a method accepting a potentially lazy enumerator. It would be to avoid having to write `enum = enum.eager if enum.respond_to?(:eage... -
07:51 AM Revision 83889292 (git): Examine TestIO#test_select_exceptfds on Solaris with 1 byte data
- On Solaris, it seems that the select(3C) in this test works only
when sending 1 byte out-of-band data, though I cannot investigate
the cause. The behavior is observed on a Solaris 10 server in
addition to Solaris 11 on which the test had... -
- Yes, I think these are resolved.
-
05:43 AM Revision e913fa94 (git): delete ruby_eval_string_from_file_protect
- Not used from anywhere.
-
-
03:59 AM Revision 355ccdea (git): Disable method definition type checks on Cygwin too
- [Bug #16134]
-
- The test consistently fails on OpenBSD.
https://rubyci.org/logs/rubyci.s3.amazonaws.com/openbsd-current/ruby-master/log/20190903T010009Z.fail.html.gz
```
1) Failure:
TestFiber#test_fork_from_fiber [/home/chkbuild/chkbuild/tmp/build/20... -
03:28 AM Revision e6c4ea22 (git): Touch tar file itself to make gz file hash stable
-
-
-
-
-
- Compilation error on x86_64-cygwin.
```
make[2]: Entering directory '/tmp/ruby-git/ruby/ext/-test-/arith_seq/extract'
linking shared-object -test-/arith_seq/extract.so
extract.o: In function `Init_extract':
/tmp/ruby-git/ruby/ext/-t... -
- Delegate now handles `eql?` after changes from #12684. However, it always returns true if the object is the same:
```ruby
nan = 0/0.0
nan.eql? nan # false
s = SimpleDelegator.new(nan)
s.eql? s # true
```
I'm not sure it is wo...
09/02/2019
-
- Fixed by https://github.com/ruby/ruby/commit/1994adf938afcdc562f87497156e6d4900f3f06b.
-
06:17 AM Bug #7768: Inherited Array class missing
- `Array#uniq` should be fixed.
Matz.
-
- Dan0042 (Daniel DeLorme) wrote:
> So far the only edge case I can think of is related to keyword extension:
> ...
There are probably many edge cases with this approach. Here's an example of one such edge case:
```ruby
def foo(*arg... -
- Previously, Array#uniq would return subclass instance if the
length of the array were 2 or greater, and would return Array
instance if the length of the array were 0 or 1.
Fixes [Bug #7768] -
- consistent with the definition
-
-
-
-
-
- In 91aa8bfff8, my understanding of the branch was inverted.
-
- For some reason, the Travis osx environment has been really unstable.
It failed on today's cron too:
https://travis-ci.org/ruby/ruby/builds/579843163
As we have almost the same test environment (including OpenSSL version)
in GitHub Acti... -
- travis-ci user does not live in #ruby-ja. Therefore the notification
isn't working anymore. -
- We don't have topic branches on ruby.git anymore.
-
-
- jeremyevans0 (Jeremy Evans) wrote:
> I think the only case that is questionable still is:
> ...
Recent changes to the master branch have fixed this issue. Keyword splats of empty hashes in arrays no longer add an empty hash to the arra... -
- With recent commits to the master branch, keyword splats of empty hashes do not pass positional arguments to methods or cause an empty hash to be added to an array.
-
- Ignore empty keyword splats in arrays
- 03:08 PM Revision be865914 (git): * 2019-09-03 [ci skip]
-
- It looks like a flag which cannot be enabled on configure.
-
-
- This seems to have been broken since 4e15be8bade.
-
-
- I don't like `--tail-call-optimization-off`. I will not specify the option, see an omitted backtrace, and then I must re-run my code with the option. It is not an easy-to-debug language for me.
And I have another concern. If tail c... -
08:37 AM Feature #12543: explicit tail call syntax: foo() then return
- mame (Yusuke Endoh) wrote:
> I'm strongly against "ON by default". It makes the backtrace difficult to understand. Consider the following program:
> ...
This should be changed to something like:
```
Traceback (most recent call ... -
- I'm strongly against "ON by default". It makes the backtrace difficult to understand. Consider the following program:
```
1: def foo
2: raise
3: end
4:
5: def bar
6: foo
7: end
8:
9: bar
```
If tail-call optimizatio... -
- I don't think tail call optimization should be a feature that is switched on or off by the programmer at each location. I think it should be an option used on execution, and it should be ON by default.
We want programs to be fast, and... -
03:29 AM Feature #12543 (Assigned): explicit tail call syntax: foo() then return
- Another idea: `tailcall return foo()`
* Background: `return` is void expression and any `method(return)` is not prohibited on current compiler. So nobody use it == no-incompatibility. Any keywords are acceptable (example: `goto return... -
- The test seems to have a race condition, which fails on very slow
machine like Solaris10s. So skip it.
In addition, this change restores timeout guard that was removed at
0660d7cb538cf5284d50f66adfcbd78609839715. This is because the t... -
09:47 AM Bug #16133: 2.5 --> 2.6: backwards incompatible change in MalformedCSVError
- I'll do that.
-
- Please report it at https://github.com/ruby/csv.
-
- In Ruby 2.5, CSV::MalformedCSVError simply inherited from RuntimeError. In 2.6, it defines its own #new, taking 2 arguments (as opposed to the single argument of RuntimeError).
Code explicitly raising new errors of this class outside ... -
- znz (Kazuhiro NISHIYAMA) wrote:
> ```
> ...
Seems not for us:
```php
var_export(pathinfo(".profile", PATHINFO_EXTENSION)); //=> 'profile'
``` -
- In some other languages:
```
% python
Python 2.7.16 (default, Mar 4 2019, 09:01:38)
[GCC 4.2.1 Compatible Apple LLVM 10.0.0 (clang-1000.11.45.5)] on darwin
Type "help", "copyright", "credits" or "license" for more information.
>... -
- `.git` is not an extension but a filename. `.a.jpg` has `.jpg` extension. I wonder how we can parse `..jpg`. It's undefined behavior for me.
Any reasoning?
Matz. -
- [Bug #15558]
-
-
- I don't think `overlap?` is a good name. Besides that, `Array#overlap?` creates an internal hash table anyway, so that it wouldn't solve the bottleneck. In this case, creating a hash table explicitly would perform better.
Matz.
-
06:50 AM Bug #16007 (Feedback): Process.clock_getres matches the clock in practice for Process::CLOCK_{PROCESS,THREAD}_CPUTIME_ID FAILED fails on armv7hl
- Patch is welcome
-
- We have no device to debug it.
Could you make a patch?
-
- Implicit assignment from named capture is kind of confusing. Unless strongly persuaded, I am not going to enhance the feature.
Although we have no plan to remove the feature.
Matz. -
- I am not positive for `Array#pluck`. ActiveSupport may add the method.
Matz. -
- ActiveSupport has [Enumerable#pluck](https://api.rubyonrails.org/classes/Enumerable.html#method-i-pluck), so I don't think we want to diverge from that by adding a method with the same name in Array.
-
- At the developers' meeting, no one was against this feature. It would be great if we can print a warning per a pair of callsite and callee (to suppress duplicated redundant warnings), but the implementation is not trivial (exploiting me...
-
- I prefer `caller_location(n=1)` or `direct_caller`. For former, it might be confusing with `caller_locations`. The latter is straightforward but it is restricted to the direct caller (`n=1`).
Matz. -
-
- > I'm personally +1 for eager.
Thanks! This proposal has been accepted by Matz.
> ...
Could you elaborate on why? Are you thinking of situations where lazy and eager enumerators are mixed?
-
- I feel a bit reluctant to have an implementation in our source tree and keep maintaining it when there's almost always a better implementation available in OpenSSL::Digest when you need one.
Here are some examples of Digest gems I wro... -
- @headius @eregon @brixen
Do you have any opinion about this as developers of other Ruby implementations? -
05:43 AM Revision 633ae327 (git): Add Net::FTP#features and Net::FTP#option
- Patch by darkphnx (Dan Wentworth) . Thanks!
[Feature #15964] -
-
- This reverts commit 83498854eb5a824f1f83c31fac18c9279f9ee10d.
This didn't pass rubyspec. -
-
- I don't see enough demand for `compact` where we have `reject(&:nil?)`. Any additional use-case?
Matz. -
- Previously, Enumerator::Lazy#with_index was not defined, so it
picked up the default implementation from Enumerator, which was
not lazy.
Based on earlier patch from nobu.
Fixes [Bug #7877] -
- Here's an updated patch that uses `#bind_call` and tests successful delegation to a BasicObject target.
-
- rb_reg_match expects its first argument to be a Regexp instance.
Should check that. -
-
- ----
trunk: ruby 2.6.0dev (2018-09-18 trunk 64767) [x86_64-darwin15]
ours: ruby 2.6.0dev (2018-09-18 opt_regexpmatch 64775) [x86_64-darwin15]
last_commit=opt_regexpmatch1 is actually making things slower.
Calculating -------------------... -
- The changes in #14183 solve this issue. You will now get warnings:
```ruby
my_func(Test.new)
# (irb):101: warning: The last argument is used as the keyword parameter
# (irb):92: warning: for `my_func' defined here
# ArgumentError... -
- vm2_regexp was for opt_regexpmatch1.
-
- With the acceptance of #14183, non-symbols are allowed in keyword hash splats:
```ruby
m(k: 1, **{2 => 3})
# [{:k=>1, 2=>3}]
m(**{4 => 5})
# [{4=>5}]
``` -
- The first example should fail, because you are passing an unsupported keyword to a method that accepts a keyword.
With the acceptance of #14183, the second example now submits a warning:
```ruby
do_something({ user: { id: 1 } })
... -
- With the acceptance of #14183, double splatting an empty hash no longer sends an empty positional hash to the method.
-
- With the acceptance of #14183, you now get:
```ruby
foo(h)
# => {"a"=>1}
foo(h).equal?(h)
# => true
foo(h, y: 1).equal?(h)
# => true
``` -
- With the acceptance of #14183, you now get a warning as the behavior will change in Ruby 3:
```ruby
Foo.bar({ b: 1 })
# (irb):67: warning: The last argument is used as the keyword parameter
# (irb):58: warning: for `method_missing'... -
- With the acceptance of #14183, this code now emits warnings as the behavior will change in Ruby 3:
```ruby
foo({b: 1})
# (irb):53: warning: The last argument is used as the keyword parameter
# (irb):51: warning: for `foo' defined h... -
- With the acceptance of #14183, double splatting an empty hash when calling a method no longer passes an empty positional hash to the method.
-
- With the acceptance of #14183, this can be closed. The behavior in Ruby 3 will be that `**args` in a method call will set keyword arguments if the method accepts keyword arguments, and will be a positional hash argument if the method do...
-
- With the master branch, you now get:
```ruby
without_parameters(*array)
# => nil
without_parameters(**empty_hash)
# => nil
with_parameters(**empty_hash)
# => []
# This doesn't raise ArgumentError, because the method does ... -
- instead of waiting 2 seconds, if possible.
-
- https://github.com/ruby/ruby/pull/2419
-
- Will this behavior be removed in 2.7?
-
- Fix following error on `utun*`:
```
1) Error:
TestSocket#test_udp_server:
Errno::ECONNREFUSED: Connection refused - recvmsg(2)
``` -
- Pointed out by ko1.
-
09/01/2019
-
- vm_call_method_missing was dropping VM_CALL_KW_SPLAT, so this just
makes it not drop it, to get the same behavior as calling the method
directly. -
09:37 PM Revision 85dc89c9 (git): Treat Japanese IME on/off code correctly
- 08:10 PM Revision 6285b235 (git): Bump version to 2.5.7
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_5@67800 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
04:13 PM Bug #16127: Delegates to BasicObject do not work
- It might be useful to have a test where the delegation happens successfully.
-
-
-
12:49 PM Feature #10118: Double splat for non-symbol keys
- I think this feature has already been implemented for Ruby 2.7. If I am correct, please close this issue.
- 03:42 PM Revision b0d885da (git): * 2019-09-02 [ci skip]
-
-
-
- This reverts commit 3be3948870f6589343c4aecb541e22fae7751b47.
The Solaris environment couldn't lookup the hostname itself by a wrong
setting. Now it is fixed, so try again. -
-
-
-
-
-
-
- This reverts commit 84dca8eff0cbcb1c23623b47fb78b0daf5c76e35.
"exceution expired" occurred on Solaris.
https://rubyci.org/logs/rubyci.s3.amazonaws.com/solaris11s-sunc/ruby-master/log/20190901T072504Z.fail.html.gz -
07:39 AM Revision 94d6ec1d (git): Make pattern matching support **nil syntax
-
- which are usually optimized away by -O3.
This CI can detect missing exports like
ea84a680755b5a7fa700618cbe78e3b2fc7be01d which was needed for
761346a9604ca2c79777d1d67fb5dcc3c30dbf69. -
-
- I released IRB 1.1.0.pre.3 and Reline 0.0.2 several hours ago. Please try below at Ruby 2.6.4 or latest version of Git:
gem install irb --pre
gem install reline -
- to suppress redefinition warnings.
-
- This reverts commit 8adefd4cf29288f6e43f20efbdd44b215ae16c7a.
I couldn't see any failure on Solaris if the guard is removed.
Give it a try. -
- vm_call_opt_send was dropping VM_CALL_KW_SPLAT, so this just makes
it not drop it, to get the same behavior as calling the method
directly. -
- Since 50f5a0a8d6e7ad89d6caff695a08dbd38edb7a6e, some C++ extensions
needed fixes like https://github.com/eagletmt/faml/pull/49 to make their
build succeed.
Thus it's worth noting that C API declarations are changed. -
- This mirrors earlier changes in keyword argument separation for
calling Ruby methods and calling procs/lambdas, so that behavior
is kept the same. -
-
-
-
-
01:36 AM Bug #14415: Empty keyword hashes get assigned to ordinal args.
- jeremyevans0 (Jeremy Evans) wrote:
> The keyword argument separation changes just made to the master branch did not affect this code, since it isn't a method call. This behavior has been present since Ruby 2.2. I think it would be a g...
08/31/2019
-
11:56 PM Revision 1756ec20 (git): Ignore the all of warnings with inline_spec.rb of bundler examples.
-
09:08 PM Bug #16132: Ruby 2.7 IRB crashes after pressing Insert and Delete
- IRB also crashes when you write: a = 5 or any line of code, and press the Delete key.
The error goes like this:
`irb(main):001:0> puts ?? Traceback (most recent call last):
35: from /home/sourav/.rvm/rubies/ruby-2.7.0-preview1/bin/... -
- Hello I am using the rvm 2.7 preview version of Ruby.
[I also tried the git version yesterday, which has the same problem mentioned below].
I am unable to use keys like End, Home, Insert etc.
The IRB crashes (with TypeError) when yo... -
- Sure, it would be fine to use `#bind_call` for this. At the time this patch was submitted, `#bind_call` wasn't available. I know I have a patch in #7833 where `#bind_call` could also be applied. We should probably do a sweep for other...
-
12:21 PM Bug #16127: Delegates to BasicObject do not work
- With `UnboundMethod#bind_call` added in https://bugs.ruby-lang.org/projects/ruby-trunk/repository/git/revisions/83c6a1ef454c51ad1c0ca58e8a95fd67a033f710, we should be able to save some allocations by assigning `::Kernel.instance_method(:...
- 04:56 PM Revision a060629e (git): * 2019-09-01 [ci skip]
-
-
-
- 8897fa60fdbc83f9cce7df384893e902c17c1bf2 worked.
But we want to retry that part and this commit is trying to fix the
problem. -
- 58a373e9ed54556e18a13825dfe72d3f7a86b833 and
3868df83330bb897be996604df3862639abbf002
didn't work. -
- https://travis-ci.org/ruby/ruby/jobs/579116510
-
- https://travis-ci.org/ruby/ruby/jobs/579108692
-
- With recent changes to the master branch, you now the get the following results:
```ruby
kws = {}
->{}.call **kws # => nil
->{}.call **{} # => nil
->{}.call **({}) # => nil
->{}.call **({};) # => nil
->{}.call ... -
-
- For insatnce
https://ci.appveyor.com/project/ruby/ruby/builds/27086475/job/mb9whkiygemlfy93 -
-
- This wraps iseq_location and should fix the leaked global test.
-
- This shows locations in places it didn't before, such as for
proc calls, and fixes the location for super calls.
This requires making iseq_location non-static and MJIT exported,
which I hope will not cause problems. -
- I was trying to think of a way to have better backward compatibility for generic forwarding, and I think I managed to stumble on a relatively clean way of doing it.
Basically the idea is to use a subclass of Hash for `**kw`
```ruby... -
- With https://github.com/ruby/ruby/commit/3463e83192215c36bdcebad8be907eaa09593a41, you now get warnings for calls where the behavior will change in Ruby 3:
```
No source
SOURCE: {}, OPTS: {:length=>2000}
file.rb:12: warning: The ke... -
- Fixed by #14183:
```
$ ruby -e 'def foo(*args); p args; end; foo(**{}); foo(**Hash.new) '
[]
[]
``` -
- Fixed by #14183:
```
$ ruby -e 'def foo(*args); p args; end; foo(**{}); foo(**Hash.new) '
[]
[]
``` -
- Fixed by #14183:
```
$ ruby -e 'def foo(*args); p args; end; foo(**{}); foo(**Hash.new) '
[]
[]
``` -
- Previously, there was no warning in this case, even though we will
be changing the behavior in Ruby 3.
Fixes [Bug #14130] -
- This warns about a case that we will continue to support.
08/30/2019
-
- @jeremyevans0 Completely agreed.
Matz says that it is a bug to split keywords depending upon whether the key is Symbol or non-Symbol. (The behavior was introduced without matz's confirmation.) So I think it is okay that the behavior... -
- With the changes in #14183, keyword splats support non-symbol keys, so with the master branch you now get:
```
No source
SOURCE: {}, OPTS: {:length=>2000}
SOURCE: {:length=>2000}, OPTS: {}
Source is mixed hash
SOURCE: {}, OPTS:... -
-
- A follow up for 16c6984bb9..b5b3afadfa. [Feature #14183]
-
- In all released versions of Ruby I tested, `zoo(hashie)` raises `ArgumentError`.
With the changes in #14183, in the master branch, you also get a warning notifying you of the upcoming behavior change:
```ruby
zoo(hashie)
# file.r... -
- With the changes in #14183:
```ruby
Foo.options({})
# (irb):21: warning: The last argument for `options' (defined at (irb):6) is used as the keyword parameter
# nil
# {}
Foo.kwarg({})
# (irb):22: warning: The last argument for... -
- With the the changes in #14183, keyword splats can support arbitrary keys, and the second example now passes:
```ruby
test(:ads, :d => 6, :f => :gah, {a:3,b:4} => 3)
# ads
# {:d=>6, :f=>:gah, {:a=>3, :b=>4}=>3}
``` -
- Matz approved this feature at the developer meeting a couple days ago, and it was merged earlier today (https://github.com/ruby/ruby/compare/b0a291f6f6a5834fd84807eb48be906ade429871...b5b3afadfab4072f55320075ccac6afe333a140c)
- 07:39 PM Revision 38e08db4 (git): * 2019-08-31 [ci skip]
-
- There are more issues than this, but hopefully this is enough
to get make test-bundler passing in CI. -
-
- This reverts the changes to parse.y in
a5b37262524ac39d2af13eea174486370a581c23 as they are not actually
needed and cause the warning for duplicate hash keys to not be
emitted. -
- Also handle some warnings for behavior that will change in 3.0.
-
-
- For methods that accept keyword arguments but do not accept a
keyword splat, if a keyword splat is passed, or keywords are
used with a non-symbol key, check the hash. If the hash contains
all symbols, keep the same behavior as before. ... -
-
- Use a [:nokey] entry in this case.
-
- Use false instead of nil for the keyword and keyword rest values
in that case. -
- The on_params hook will use :nil as the keyword rest argument.
There is a new on_nokw_param hook as well.
This fixes a type issue in the previous code, where an ID was
passed where a VALUE was the declared type. The symbol :nil is
pass... -
- This syntax means the method should be treated as a method that
uses keyword arguments, but no specific keyword arguments are
supported, and therefore calling the method with keyword arguments
will raise an ArgumentError. It is still al... -
- This fixes MJIT after rb_hash_stlike_foreach used vm_args.c.
-
- This restores compatibility with Ruby 2.6, splitting the last
positional hash into positional and keyword arguments if it
contains both symbol and non-symbol keys. However, in this case
it will warn, as the behavior in Ruby 3 will be to... -
- Now that keyword splats accept non-Symbols, the type of exception
changes.
Previously, a TypeError (hash key "k1" is not a Symbol) was raised
for this test, but now an ArgumentError (unknown keyword: "k1") is
raised. -
- Now that keyword splats accept non-Symbols, the inspect value of
the keyword is used instead of the string value. -
- If all keys are not symbols, then the non-symbol keys would not
be treated as keywords in previous versions. It doesn't make
sense to treat these hashes as keywords to break compatibility and
warn about behavior changes in Ruby 2.7 when... -
-
- Mostly requires adding ** in either calls or method definitions.
-
- Previously, it only applied if the call had more positional
arguments than the method it was calling accepted. -
- Treat the ** syntax as passing a copy of the hash as the last
positional argument. If the hash being double splatted is empty, do
not add a positional argument.
Remove rb_no_keyword_hash, no longer needed. -
- And, allow non-symbol keys as a keyword arugment
-
- Ok I see what you meant. BTW Struct#values_at follows the Array rather than Hash API, because Struct also thinks of itself as a tuple :-/
Struct.new(:x).new(42).values_at(0) #=> [42]
Struct.new(:x).new(42).values_at(:x) #=... -
02:55 PM Feature #16122: Data: simple immutable value object
- @Dan0042
> Question: you say "Doesn't think of itself as almost hash" but at the same time you say it should have `to_h`. Isn't that a contradiction?
Nope. An object that has `to_h` is not "something hash-like", it is just "someth... -
- Question: you say "Doesn't think of itself as almost hash" but at the same time you say it should have `to_h`. Isn't that a contradiction? What exactly are you looking for?
Naming suggestion: BasicStruct (in parallel to Object and Bas... -
- @naruse Of course, there _are_ several good gems with more-or-less similar functionality. But, from the hard experience, large codebases tend to look with great caution on the "small utility" gems to avoid dependency bloat and tend to de...
-
- > I believe that concept is that simple, that nobody will even try to use a gem for representing it with, unless the framework/library used already provides one.
I'm using immutable_struct.gem. -
- jeremyevans0 (Jeremy Evans) wrote:
> For `tainted?`/`trusted?` to work correctly, we would need to continue to support taint tracking at least in some state. We could reduce the scope of the taint flag, though. For example, we could m... -
- By no-op, I meant they would make no changes and return self. I didn't mention `tainted?` or `trusted?` earlier, but I think it may make sense to remove them earlier than `taint`/`trust`/`untaint`/`untrust`. Maybe a non-verbose warning...
-
- @jeremyevans0, by "no-op" did you mean only in the context of $SAFE mode, or did you mean that `tainted?` and `trusted?` would always return false? In the second case I think it's better to just remove the method, at least that's an obvi...
-
- I must admit to using taint sometimes in my code, as a way to keep track of dirty/modified status on an object (mea culpa)
```ruby
hash.taint[key] = newvalue
...
save(hash.untaint) if hash.tainted?
```
It's probably not common ... -
- I'm also +1 for jeremy's proposal.
I often got the test fails related `$SAFE` on rubygems. I'm happy to leave them with this proposal. -
- +1 for the removal, and I agree with Jeremy's plan for 2.7 and 3.0.
For 3.2 and 3.3, I think we may keep all the methods as no-op because old not-maintained-well scripts may break, though I'm not so strongly against the removal.
(Anywa... -
01:41 PM Bug #14716: SecureRandom throwing an error in Ruby 2.5.1
- I got this when running a ruby script that openened a large number of TCP socket (and thus files). In fact I had to increase the file limit of the shell using ulimit to be able to open enough sockets.
-
- to drop MJIT_DLDFLAGS from compile_c_to_o for OpenBSD failure
https://rubyci.org/logs/rubyci.s3.amazonaws.com/openbsd65/ruby-master/log/20190830T110008Z.fail.html.gz.
8c7f4e8f8b7f9e74195ea0eb51f39014fec342c0 did not work for i686-linux
... -
- I think this did not work for some environments, but this seems to fix
OpenBSD RubyCI failure:
https://rubyci.org/logs/rubyci.s3.amazonaws.com/openbsd65/ruby-master/log/20190830T110008Z.fail.html.gz
Let me check RubyCI impact by this. -
09:24 AM Revision c45dd4d4 (git): Make the dot-colon method reference frozen
- [Feature #16103]
Close: https://github.com/ruby/ruby/pull/2267 -
08:07 AM Feature #15553: Addrinfo.getaddrinfo supports timeout
- Thank you for reviewing. I will commit it.
-
- Reviewed the patch and found no issues. So LGTM.
-
- nobu (Nobuyoshi Nakada) wrote:
> I thought UTF-16LE and CP932 as the main purpose however, I'm bit surprised that these texts have been extinct on Windows already. :tada:
They are not yet extinct, unfortunately :-(. In Japan, there m... -
- I thought UTF-16LE and CP932 as the main purpose however, I'm bit surprised that these texts have been extinct on Windows already. :tada:
-
- of lib/bundler/vendor/molinillo/lib/molinillo/resolution.rb.
This reverts commit 500149709b92ccb36396589a0c39afb3ff77bfb6 partially. -
- It seems to be fixed by b870ca58791e576f5dbb263f54ae433a7f6d65ee.
-
- osyo (manga osyo) wrote:
> If you use https://bugs.ruby-lang.org/issues/15955, you can write as follows.
> ...
Thank you for the suggestion, but that looks a little too long. It can also be written as:
```ruby
:product.to_proc.(*ar... -
- Try to fix failure at https://github.com/ruby/ruby/runs/207580232#step:10:382
```
undefined method `name' for "Gemfile.lock":String
/home/runner/work/ruby/ruby/lib/bundler/resolver/spec_group.rb:65:in `eql?'
``` -
-
-
- Applied in changeset commit:git|83c6a1ef454c51ad1c0ca58e8a95fd67a033f710.
----------
proc.c: Add UnboundMethod#bind_call
`umethod.bind_call(obj, ...)` is semantically equivalent to
`umethod.bind(obj).call(...)`. This idiom is used in ... -
- Matz approved the feature. The name "apply" was arguable in some terms:
* We may want to use the name "apply" for other purpose in future.
* This API will be used not so frequently. Only some fundamental libraries (like pp, sorbet-... -
-
-
-
-
-
-
-
- This changeset make it possible to use very old clang.
-
- Related to [Feature #15955].
-
- For UnboundMethod#bind_call [Feature #15955] introduced in
002e592e0d67bb0271d16314a32380ad947c9ae9. -
- `umethod.bind_call(obj, ...)` is semantically equivalent to
`umethod.bind(obj).call(...)`. This idiom is used in some libraries to
call a method that is overridden. The added method does the same
without allocation of intermediate Meth... -
- `Kernel#pp` has wrapper, but `PP` does not.
https://github.com/ruby/ruby/runs/207405029#step:10:141
`NameError: uninitialized constant Bundler::Molinillo::Resolver::Resolution::PP`
08/29/2019
-
09:58 PM Revision 0a9d74f8 (git): Move pread + pwrite tests out of RUBY_ENGINE block
- These tests were guarded by a RUBY_ENGINE of "ruby" even though they test an official Ruby feature (pread/pwrite added in Ruby 2.5). This commit moves them to the top level of the test case so they will run on other implementations.
-
09:51 PM Revision bdc36094 (git): Add some NODE information for lldb
- Just adds a conditional in the lldb scripts so we can more easily debug
NODE objects. -
- @mame, @matz, I updated the description, tried to include a proper rationale for every design decision made.
-
- @mame I understand your concerns. I'll update the description today or tomorrow to include all the terminology and detailed rationale behind the proposal.
-
- I couldn't understand what is "value object", and I found: https://martinfowler.com/bliki/ValueObject.html
Please do not assume that everybody knows such an essay ;-)
No one pointed out the article during the developer's meeting, so we c... -
- @Dan0042 you are (probably) missing `#inspect`, `#==`, `#eql?`, `#hash`, `#to_h` and a bunch of other methods that are pretty trivial, but also important for the "value object".
-
- If I understand correctly, the idea is to have `X=Struct::Value.new(:x,:y,:z)` which is strictly equivalent to
```ruby
class X
def initialize(x=nil, y=nil, z=nil)
@x,@y,@z = x,y,z
end
attr_reader :x, :y, :z
#and othe... -
- @matz Sorry for not sharing more detailed reasoning which led to the current proposal (I explained the "final reasons" in its text, but it is too terse).
So, it went as following:
1\. First, I really wanted just `Struct.new(..., im... -
- The typical solution is `Struct.new(...).freeze`. This doesn't require any enhancement. The other option is `Struct.new(..., immutable: false)`. It looks simpler than the proposed `Struct::Value`.
Matz.
-
- byroot (Jean Boussier) wrote:
> > 3.2 `taint/trust/untaint/untrust` non-verbose warning when called
> ...
No. Verbose warning means a warning only printed in verbose mode (`ruby -w`, or `$VERBOSE = true`). Non-verbose warning means a... -
- > 3.2 `taint/trust/untaint/untrust` non-verbose warning when called
Maybe you meant verbose here?
Other than that I agree with the proposed timeline, and as soon as these methods are noop, their cost become mostly null.
Making t... -
- I agree with the removal of `$SAFE` and the taint tracking. Proposed timeline:
2.7:
* Remove taint tracking/mechanism.
* Non-verbose warning on setting/access of `$SAFE`
* `taint`/`trust`/`untaint`/`untrust` become no-ops, verbos... -
- Ruby had Taint checking which is originally introduced in Perl.
https://en.wikipedia.org/wiki/Taint_checking
It was intended to provide a useful tool for handle objects which are come from outside.
Input data is set as tainted by de... -
06:57 PM Revision 94b79bff (git): Add word "Euler's number" to Math::E docs
- When searching for this constant, I landed on the correct page https://ruby-doc.org/core-2.6.4/Math.html however I was using CMD+f to search for "Euler" and did not find it. If we add the full name for this constant then it will be easie...
-
- e.g. `GZIP=-9`
-
- Applied in changeset commit:git|e51dca2596db9567bd4d698b18b4d300575d3881.
----------
Disallow use of attrset symbols as Struct members
Fixes [Bug #11326] -
- Accepted.
Matz. - 03:23 PM Revision c52d38b7 (git): * 2019-08-30 [ci skip]
-
- Fixes [Bug #11326]
-
-
-
-
-
- typedef struct _COORD {
SHORT X;
SHORT Y; // I wanted to take this...
} COORD, *PCOORD;
typedef struct _CONSOLE_SCREEN_BUFFER_INFO {
COORD dwSize;
COORD dwCursorPosition; // ...of this one
WORD wAttributes; ... -
- typedef struct _CONSOLE_SCREEN_BUFFER_INFO {
COORD dwSize; // 4(SHORT X, Y)
COORD dwCursorPosition; // 4
WORD wAttributes; // 2
SMALL_RECT srWindow; // 8(SHORT Left, Top, Right, ... -
11:50 AM Revision 13b69220 (git): bump up teeny version to 2.6.5.
- git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_6@67799 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
-
- Applied in changeset commit:git|a4a19b114ba94b8f28d5a91aee5d595a516006d5.
----------
Allow non-finalizable objects in ObjectSpace::WeakMap
[feature #16035]
This goes one step farther than what nobu did in [feature #13498]
With this p... -
- I think this is the required behavior for `WeakMap` to implement a cache for example. Accepted.
Matz.
-
- sorry, value is collectable (my misunderstand).
-
-
- `42` never be collected. Is it intentional?
-
- [feature #16035]
This goes one step farther than what nobu did in [feature #13498]
With this patch, special objects such as static symbols, integers, etc can be used as either key or values inside WeakMap. They simply don't have a fina... -
- I just wrote a draft implementation of Enumerator.produce. I just didn't notice @zverok kindly worked on writing a patch, so I'll review and integrate it later!
-
- I prefer `produce` to `iterate`, `generate` or `from`. Accepted (at least for the experiment).
Matz. -
- @matz Of course. What prompted me to open this feature request is a feature I implemented in Rails where I use this exact pattern: https://github.com/rails/rails/pull/35891 (more specifically https://github.com/rails/rails/blob/48ae1ba36...
-
- I am not sure if the proposal has real-world use-case. Can you elaborate?
Matz.
-
- It seems the combination fails at the moment. Don't know exactly
why but I suspect there can be issues in resolving weak references.
Let's rule them out for now.
https://ci.appveyor.com/project/ruby/ruby/builds/27036383/job/x3c5d54839a... -
- We can check the function pointer passed to
rb_define_singleton_method like how we do so in rb_define_method.
Doing so revealed many arity mismatches. -
- We can check the function pointer passed to rb_define_private_method
like how we do so in rb_define_method. Doing so revealed some
problematic usages of rb_obj_dummy. They had to be split according
to their arity. -
- We can check the function pointer passed to
rb_define_protected_method like how we do so in rb_define_method.
This changeset revealed no prototypes mismatches. -
- We can check the function pointer passed to rb_define_method_id
like how we do so in rb_define_method. This method is relatively
rarely used so there are less problems found than the other APIs. -
- To properly generate documents.
-
- We can check the function pointer passed to rb_define_global_function
like we do so in rb_define_method. It turns out that almost anybody
is misunderstanding the API. -
- We can check the function pointer passed to rb_define_module_function
like how we do so in rb_define_method. The difference is that this
changeset reveales lots of atiry mismatches. -
- The rb_define_method function takes a pointer to ANYARGS-ed functions,
which in fact varies 18 different prototypes. We still need to
preserve ANYARGS for storages but why not check the consistencies if
possible.
Q&As:
Q: Where did th... -
- This is just a trivial mistake introduced in
0f36e8fc03a5c6433972d6bb5f177d5f6e106bac. -
- This is just a trivial mistake introduced in
6c56dae4b23c5c50e351758538141ca26b9aba40. -
- Applied in changeset commit:git|abe12d8b96be3c12618811f22ca01788366f99f8.
----------
Freeze method reference operator object
[Feature #16103] -
- Matz accepted it :tada:
-
- [Feature #16103]
-
- If you use https://bugs.ruby-lang.org/issues/15955, you can write as follows.
```ruby
class UnboundMethod
# apply or other name
def apply(receiver, *args)
bind(receiver).call(*args)
end
end
arrays = [["a", "b"], ["c... -
- I like the OP's idea. It's up to @marcandre
Matz. -
- Accepted. It should raise an error on end-exclusive ranges (`...`).
Matz. -
- I understand there's theoretically exist a situation this feature is useful.
But I think it doesn't exist in practice.
I object to provide an additional utility to support legacy encoding. -
- Depending on usage, distinction of UTF-8 (with/without BOM), UTF-16LE without BOM, UTF-16BE with or without BOM, and so on may also be necessary. Also, for Japanese, traditionally distinction between EUC-JP, Shift_JIS, and ISO-2022-JP ca...
-
- This reverts commits:
* d365fd5a024254d7c105a62a015a7ea29ccf3e5d
* d780c3662484d6072b3a6945b840049de72c2096
* aa7211836b769231a2a8ef6b6ec2fd0ec882ef29
* 043f010c28e82ea38978bf8ed885416f133b5b75
* bb4dd7c6af05c7821d572e2592ea3d0cc748d81f
... -
- During today's meeting, it was mentioned that production deployments may not want to include this functionality.
To do this, one solution would be to make this functionality available as a bundled gem. -
- 06:12 AM Revision 07f9cd96 (git): * 2019-08-29 [ci skip]
-
- Agreed.
Matz. -
- After experiments, `|>` have caused more confusion and controversy far more than I expected. I still value the chaining operator, but drawbacks are bigger than the benefit. So I just give up the idea now. Maybe we would revisit the idea...
-
-
-
- After discussion and consideration, we picked `_1`, `_2`... as numbered parameters. I investigated replaced examples, the current `@1` etc. are too confusing with instance variables. So I picked underscore for numbered parameters (a la S...
-
05:05 AM Feature #15588: String#each_chunk and #chunks
- Just an idea, this method may be useful to treat data of fixed-length record format if it accepts multi column lengths, such as
```ruby
records = []
fixed_length_records_data.each_slice(7, 10, 20) do |zip, tel, name|
records.push... -
- As @shyouhei mentioned, we'd like to hear the real-world use-case. Extracting fixed-width records may be the purpose. I'm curious about the OP's opinion.
Matz. -
- Applied in changeset commit:git|761346a9604ca2c79777d1d67fb5dcc3c30dbf69.
----------
Show the previous definition location,
when reopened class/module redefinition mismatched the previous
definition. [Feature #11460] -
- Accepted.
Matz.
-
- usa said it might not work on windows when the install path includes non-ASCII characters. Please check it out, @nobu .
-
- when reopened class/module redefinition mismatched the previous
definition. [Feature #11460] -
- Dan0042 (Daniel DeLorme) wrote:
> ```
> ...
Correct. That is expected, as the method takes no arguments, only keyword parameters.
> ```
> ...
Correct. This is expected, as the method does not support a keyword parameter for `9`.
... -
- I tried testing to see what behavior changes are introduced via this. This is what I found so far.
```
def foo(*c, **f); p [c,f]; end; foo(9=>9, f:12)
before: [[{9=>9}], {:f=>12}]
after: [[], {9=>9, :f=>12}]
```
Good bugfix I th... -
- A range literal in conditional expression is turned into a
flip-flop, as a Range object is never falsy and does not make a
sense. -
-
- This reverts commit bf7a32d22079cc44eb19794e41d82b886d5d17b3.
flip-flop is no longer deprecated.
[Feature #5400] -
- I cannot reproduce this on 2.5.6, 2.6.4, 2.7.0-preview1, or the master branch. It was probably fixed by commit:f91879a7b548284c93743168acfd11e3d2aeefac. If you are still having problems with it, please reopen.
08/28/2019
-
- Attached is a patch that fixes this issue by switching from `rb_block_call` to `rb_funcall_with_block`.
-
- Applied in changeset commit:git|cd0e208963bdf9ee2fec30e83bdc8f6bc77a7b8d.
----------
Document {n}? regexp pattern is optional and not non-greedy [ci skip]
While not consistent with {n,}?, {,m}?, and {n,m}?, it is arguably
more useful, ... -
- While not consistent with {n,}?, {,m}?, and {n,m}?, it is arguably
more useful, as otherwise the ? would have no effect.
Fixes [Bug #15583] -
- This problem appears fixed in 2.6.4 and 2.7.0-preview1 in my environment (OpenBSD-amd64). It still occurs in 2.5.6. If it still occurs for you in 2.6.4, please reopen.
-
- @zverok Thank you!
-
- @mame OK, sorry. I was not sure how the process looks internally (e.g. should I save the space and attention of the reader in agenda by just listing things shortly, or provide the full context in-place). Here are updated descriptions, ju...
-
- zverok (Victor Shepelev) wrote:
> * [Feature #14781] `Enumerator.generate`. Patch provided; Matz seem to have liked the idea and name proposed, can we have it in 2.7?
> ...
The two tickets were discussed at the past meetings. Next tim... -
- Looks like we have too many topics at the next meeting. Please do not expect all tickets to be discussed. Thank you for your understanding.
-
10:58 AM Bug #15900: Resolv::DNS#each_name doesn't accept Resolv::IPv{4,6} against documents
- Could you please backport to 2.5 and 2.6?
-
-
- This reverts commit 6454808c52fff445ff09fefb0fb96988f82aaa3c.
It is no longer needed, as `VCS::SVN#get_revisions` now returns
`Integer` as revision numbers, and `short_revision` should deal
with it. -
- `_FILE_OFFSET_BITS` should be defined before system headers.
-
- Remove jquery.js
Remove debug print [ci skip]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_6@67797 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
- Fixed unmatched indent of the closing bracket for
`:irb_current_working_workspace`, and adjusted following elements. - 06:57 AM Revision 2b55df73 (git): `revision` might be an Integer
- 06:45 AM Revision aca4e654 (git): merge revision(s): f308ab2131ee675000926540cbb8c13c91dc3be5
- Remove jquery.js
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_4@67794 b2dd03c8-39d4-4d8f-98ff-823fe69b080e - 06:43 AM Revision 32aca3dd (git): merger revision(s): f308ab2131ee675000926540cbb8c13c91dc3be5
- Remove jquery.js
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_5@67793 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
-
- I backported 4e038a7e into ruby_2_6 at r67792.
-
- Revert "parse.y: Deprecate flip-flops"
This reverts commit bae638ad5b782c44c80efe33834cb9039279af46.
[Feature #5400]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_6@67792 b2dd03c8-39d4-4d8f-98ff-823fe69b080e - 05:00 AM Revision 47403b71 (git): merge revision(s) 989e8ad322afdfa3aa06e74b89fc42aef42895d0:
- Remove debug print [ci skip]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_4@67791 b2dd03c8-39d4-4d8f-98ff-823fe69b080e - 04:57 AM Revision 73eb2984 (git): merge revision(s) 989e8ad322afdfa3aa06e74b89fc42aef42895d0:
- Remove debug print [ci skip]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_2_5@67790 b2dd03c8-39d4-4d8f-98ff-823fe69b080e -
-
- This isn't a bug. It is not in Ruby's control to release memory to the operating system. Ruby just calls `free` on the underlying memory pointer. If the libc `free` function does not end up releasing the memory to the operating system...
-
03:08 AM Feature #15631 (Closed): Let round_capa for ID table not allocate excess capacity for power of 2 ints >= 4 - Applied in changeset commit:git|8e13da1ee83028000e5d7f9f9526379e32765a81.
----------
optimize get_power2 [Feature #15631]
Merged: https://github.com/ruby/ruby/pull/2292 - 02:29 AM Revision 8e13da1e (git): optimize get_power2 [Feature #15631]
- Merged: https://github.com/ruby/ruby/pull/2292
-
-
-
- I don't think the current behavior is a bug, as the parameters for a non-lambda proc are optional. Also, changing the behavior could be backwards incompatible.
I propose instead a `lambda` keyword to `Proc#parameters`, which would re... -
-
-
- Does this issue still occur with 2.5.5 or 2.6.3?
-
- Does this issue still occur with 2.5.5 or 2.6.3?
-
- Testing Ruby versions from 1.9 to 2.7, only Ruby 2.3 does not raise a `NoMethodError` for the example code. I think that can be considered a bug in Ruby 2.3. As Ruby 2.3 is no longer supported, the bug will not be fixed.
-
-
- I don't think that this should be considered a bug. Going from immutable to mutable should not break things, while going from mutable to immutable can. It is true that the using `String#+` can decrease performance slightly, but that sh...
-
-
-
-
- because default value overrides it.
https://help.github.com/en/articles/virtual-environments-for-github-actions#environment-variables -
-