Activity
From 11/22/2019 to 11/28/2019
11/28/2019
-
08:40 PM Bug #16380 (Third Party's Issue): OSX Segmentation fault at 0x0000000106e67698 / illegal hardware instruction rake install_latest
- This error occurs in Oracle Instant Client:
```
/usr/lib/system/libsystem_platform.dylib(_sigtramp+0x1d) [0x7fff6968cb1d]
/Users/rok/instantclient_18_1/libclntshcore.dylib.18.1(sltsqKeyDestructorInt+0x68) [0x1065864b8]
/Users/rok/i... -
05:19 PM Bug #16380: OSX Segmentation fault at 0x0000000106e67698 / illegal hardware instruction rake install_latest
- I don't see anything obvious in the crash logs.
As a shot in the dark, could you ask the people that have these troubles to run `gem pristine $gem_name` on all the gems they use? `gem pristine --all` works too, but might take longer.
I... -
05:09 PM Bug #16380 (Third Party's Issue): OSX Segmentation fault at 0x0000000106e67698 / illegal hardware instruction rake install_latest
- Since a while we have across our team 4 or 5 developers who run often into "Segmentation fault" on OSX machines when running a rake script.
Problem:
We start the rake task and sometimes it stops with a **Segmentation fault**.
Envi... -
06:24 PM Revision 6b460a79 (git): export func for MJIT
-
- This reverts commit 2e6f1cf8b264f4c8499c4e5f18bf662fdade04ff.
-
- This reverts commit 27d0d7c0d39076d4bbacd3c3f3864322699db7b4.
- 06:18 PM Revision 27d0d7c0 (git): * remove trailing spaces. [ci skip]
-
- 06:11 PM Revision e5705c35 (git): * 2019-11-29 [ci skip]
- 06:11 PM Revision d104ebf2 (git): * remove trailing spaces. [ci skip]
-
- vm_getivar() provides fastpath for T_OBJECT by caching an index
of ivar. This patch also provides fastpath for FL_EXIVAR objects.
FL_EXIVAR objects have an each ivar array and index can be cached
as T_OBJECT. To access this ivar array, g... -
04:30 PM Feature #16348: Proposal: Symbol#start_with?, Symbol#end_with?, and Symbol#include?
- duerst (Martin Dürst) wrote:
> I personally think that we should be restrictive in adding new methods to `Symbol`. Otherwise, this very quickly becomes a slippery slope, and the distinction between `Symbol` and `String` becomes less and... -
- Applied in changeset commit:git|b5fbefbf2c14742f6d46ecdf3ce712062dfb1d0a.
----------
Added Symbol#start_with? and Symbol#end_with? method. [Feature #16348] -
09:38 AM Feature #16348: Proposal: Symbol#start_with?, Symbol#end_with?, and Symbol#include?
- there is also `symbol[string]` which can be used as replace for Symbol#include?
-
09:08 AM Feature #16348: Proposal: Symbol#start_with?, Symbol#end_with?, and Symbol#include?
- I am OK with adding `start_with?` and `end_with?`.
I am a bit reluctant with adding `include?` since Symbol does not include any character. Persuade me further if you really need `Symbol#include?`.
Matz.
-
-
-
12:45 PM Revision 0b453e2a (git): Get rid of inadvertent label [ci skip]
-
-
-
10:57 AM Bug #13446 (Closed): refinements with prepend for module has strange behavior
- Applied in changeset commit:git|a0579f3606561a74e323f6193b9504c06845236c.
----------
Make prepending a refined module after inclusion not break refinements
After the previous commit, this was still broken. The reason it
was broken is t... -
- Applied in changeset commit:git|5069c5f5214ce68df8b3954321ad9114c5368dc3.
----------
Honor refinements for modules that prepend other modules
This previously did not work, and the reason it did not work is
that:
1) Refining a module o... -
- After the previous commit, this was still broken. The reason it
was broken is that a refined module that hasn't been prepended to
yet keeps the refined methods in the module's method table. When
prepending, the module's method table is m... -
- This previously did not work, and the reason it did not work is
that:
1) Refining a module or class that prepends other modules places
the refinements in the class itself and not the origin iclass.
2) Inclusion of a module that prep... -
10:30 AM Feature #16378: Support leading arguments together with ...
- matz (Yukihiro Matsumoto) wrote:
> I know trailing `...` can be very useful from C experience. But the primary purpose of Ruby `...` is method delegation. We are not going to extend the role of `...` in the language (at least for now).
... -
- I know trailing `...` can be very useful from C experience. But the primary purpose of Ruby `...` is method delegation. We are not going to extend the role of `...` in the language (at least for now).
Matz.
-
- * [Feature #16379] Backporting ... to Ruby 2.4 - 2.6 and pass_keywords (eregon)
* I'd like to have the opinion of the core team on this. -
06:40 AM Misc #16262: DevelopersMeeting20191128Japan
- * [Feature #16377] Regexp literals should be frozen (byroot)
* Previous discussion in [Feature #16345]
* Regexp literals always return the same mutable object, it can cause state to leak. -
- I've heard quite a few people saying that backporting `...` to Ruby 2.4-2.6 would be a good way to do migration for keyword arguments.
In fact, if we have `...` in Ruby 2.4+, I think `...` could be the way to do 90%+ of delegation in a ... -
- I like the idea of helpers in https://bugs.ruby-lang.org/issues/16122#note-18.
We need to discuss further the combination of attributes (immutable, enumerable, etc.)
Matz.
-
- * The name `zip_with` is too confusing with `zip`. We need a new name.
* The behavior can be described by the combination of `zip` and `map`.
* I am not sure how much we need this behavior (yet).
Matz.
-
- As far as I understand, the code with the proposal `[1, 2, 3].zip([4, 5, 6]).each_tuple.map(&:+)` can be written as following with numbered parameters:
```ruby
[1, 2, 3].zip([4, 5, 6]).map{_1 + _2}
```
which is quite plain and sh... -
06:24 AM Feature #16261: Enumerable#each_splat and Enumerator#splat
- I agree this feature would be a nice addition.
Actually I had exactly the same idea, presented at Rails Developer Meetup 2019: https://www.slideshare.net/akinorimushaevolution-of-enumerator (Japanese)
There's a subtle difference be... -
- Note: this is not a final decision:
* Having a method to retrieve the endpoint of a method/proc is OK for me.
* The name of the method should **not** be `code_range`, because the name is confusing.
* Making the return value as a speci... -
04:46 AM Feature #16142: Implement code_range in Proc and Method
- > Maybe we could change #source_location to return a `SourceLocation` and have `SourceLocation#to_a` so `file, line = proc.source_location` would still work?
I've spent some time thinking about this.
I wanted to understand why we h... -
- Accepted. Go ahead.
Matz.
-
- Apparently it's a duplicate of https://bugs.ruby-lang.org/issues/8948
-
06:32 AM Revision 4325f080 (git): Add require "irb" to test/irb/test_completion.rb
-
- `Enumerable#count` should aware of interrupts. Besides this case, `#count` should return `Infinity` (we may need to define `Range#count`).
Matz.
-
- IRB completion logic always needed exponential notation for complex literal
such as 3e6i but it's bug. I fixed to support complex literal without
exponential notation such as 3i. -
- By default, it should emit (deprecation) warnings. It's a good idea to have an environment variable option to suppress deprecation warnings.
Matz.
-
- https://github.com/ruby/ruby/pull/2708
-
- Applied in changeset commit:git|8b4ee5d6ba92a385eedc9235ce0a2d5618deecf0.
----------
Raise `NoMatchingPatternError` when expr `in` pat doesn't match
* `expr in pattern` should raise `NoMatchingError` when unmatched
* `expr in pattern` ... -
- I accept this proposal for two reasons:
* as OP described, returning boolean values from `in` pattern matching masks match failure. This can cause serious problems.
* By this change, we cannot use `in` pattern matching in `if` conditio... -
- * `expr in pattern` should raise `NoMatchingError` when unmatched
* `expr in pattern` should return `nil`. (this is unspecified, but
this feature is experimental, at all)
[Feature #16355] -
04:24 AM Revision 18953416 (git): Fix documentation of `MonitorMixin#new_cond` [ci skip] (#2707)
- Since https://github.com/ruby/ruby/pull/2576,
`new_cond` uses the Monitor object, not the receiver. -
- It seems OK as the original proposal is same as the current behavior.
-
- Accepted.
Matz.
-
-
11/27/2019
-
- The above also shows we should avoid multiple calls to rb_warn* for a single logical warning.
E.g., `warning: The last argument is used as the keyword parameter` and `warning: for method defined here` should be part of a single (multi-l... -
- It's actually only 130 unique warnings:
```
$ wc -l run.txt
2439 run.txt
$ grep -v 'Setting up parallel test mode - starting' run.txt > run2.txt
$ wc -l run2.txt
2408 run2.txt
$ cat run2.txt | sort | uniq | wc -l
130
```
This ... -
11:13 PM Feature #16345: Don't emit deprecation warnings by default.
- I tried Jeremy's patch results are here of running Discourse `bin/turbo_rspec` it runs our test suite using 16 threads on my computer.
2440 warning for every test run, suffice to say at Discourse we will absolutely turn warnings of... -
10:54 PM Feature #16345: Don't emit deprecation warnings by default.
- jeremyevans0 (Jeremy Evans) wrote:
> > However there are several situations that
> ...
Of course, we can enable deprecation warnings for
pry and other test libraries.
-
- sam.saffron (Sam Saffron) wrote:
> Even with @Eregon's proposed patch running our test suite with warnings is unusable cause of pages of:
Do you mean @mame's PR from #16289, that is https://github.com/ruby/ruby/pull/2458 ?
Could ... -
- sam.saffron (Sam Saffron) wrote:
> I agree 100% with akr here.
> ...
Note that the vast majority of the warnings listed are not keyword argument separation warnings, they are safe level and String#match? warnings. I think the patch yo... -
02:01 PM Feature #16345: Don't emit deprecation warnings by default.
- When we consider the default verbosity of warnings, we should care about the "audience" of the warnings.
In the previous developers' meeting, the default warnings should target server operators (such as a SRE team) as a primary audien... -
- I agree 100% with akr here.
Even with @Eregon's proposed patch running our test suite with warnings is unusable cause of pages of:
```
/home/sam/.rbenv/versions/trunk/lib/ruby/gems/2.7.0/gems/ffi-1.10.0/lib/ffi/pointer.rb:97:... -
-
-
-
08:32 PM Revision 416cccc6 (git): Drop windows-2016 on GitHub Actions
- https://help.github.com/en/actions/automating-your-workflow-with-github-actions/software-installed-on-github-hosted-runners#windows-server-2016
> Note: The Windows Server 2016 virtual environment will be removed on December 3, 2019. -
- I think we should just return true for `lambda?`.
Proc has extra confusing behavior, e.g., #16166. - 06:03 PM Revision cdcaf041 (git): * 2019-11-28 [ci skip]
-
- Previously, the rest array was modified, but it turns out that is
not necessary. Not modifying the rest array fixes cases when the
rest array is used more than once. -
- @sawa wrote this in the dev-meeting ticket:
> Unintended arity. This must be fixed in an earlier stage before Ruby 3.
I think matz conclusion is all behavior shown in this bug so far is intended, except for `*foo, **bar`.
@Dan0042... -
- Maybe we could change #source_location to return a `SourceLocation` and have `SourceLocation#to_a` so `file, line = proc.source_location` would still work?
-
- Also I believe `...` could be a good way to do delegation in all versions for lexical cases (the majority):
```ruby
# Could be some constant in a gem
ARGS = RUBY_VERSION < "2.7" ? "*args, &block" : "..."
class_eval <<RUBY
def meth... -
- I think this is very important, otherwise `...` can be used only very rarely.
For instance, `method_missing` typically want to access the method name like:
```ruby
def method_missing(name, ...)
if name.to_s.end_with?('?')
se... -
- I filed #16378 for supporting leading arguments with `...`.
-
- @mame's measurements:
```ruby
def req(x)
x
end
hash = {a: 1}
arr = [hash]
100000000.times { req(*arr) }
```
gives
```
$ time ./miniruby.check bench.rb
real 0m5.726s
user 0m5.689s
sys 0m0.025s
$ time ./miniruby.nocheck b... -
- Eregon (Benoit Daloze) wrote:
> jeremyevans0 (Jeremy Evans) wrote:
> ...
After discussion with you and @mame , it appears my benchmarking was flawed and the performance difference is closer to your measurements.
> Your original diff... -
- jeremyevans0 (Jeremy Evans) wrote:
> Do all non-`ruby2_keywords` tests still pass with that? If so, what are the benchmark results with the patch?
Yes, all specs pass and test-all has just 2 expected failures:
```
1) Failure:
T... -
- jeremyevans0 (Jeremy Evans) wrote:
> There are a ton of other changes in 2.7 that could affect performance besides `ruby2_keywords`.
Any specific guess?
> ...
Your original diff kept quite a lot more complicated code in `setup_par... -
- Eregon (Benoit Daloze) wrote:
> https://github.com/ruby/ruby/compare/master...eregon:no-ruby2_keywords is a more extensive removal of `ruby2_keywords`, although not complete yet.
> ...
With that change applied I could see a slowdown of... -
04:10 PM Feature #16375: Right size regular expression compile buffers for literal regexes and on Regexp#freeze
- shevegen (Robert A. Heiler) wrote:
> > Would any further incremental work towards this be considered worthwhile from a
> ...
Yes this isn't focused on speed, but more so on being careful with auxiliary memory buffers and inferring wher... -
02:43 AM Feature #16375: Right size regular expression compile buffers for literal regexes and on Regexp#freeze
- > Would any further incremental work towards this be considered worthwhile from a
> ...
I can not answer for the ruby-core team, or anyone else, but matz has the motto of
3x3 for ruby 3.0 which is destined for next year.
Some time a... -
- References PR https://github.com/ruby/ruby/pull/2696
As a continuation of [type specific resize on freeze implementations of String and Array](https://bugs.ruby-lang.org/issues/16291) and looking into the `Regexp` type I found these m... -
- Attached is a patch that adds `Zlib::GzipReader.each_file` will which handle multiple concatenated gzip streams in the same file, similar to common tools that operate on `.gz` files. `Zlib::GzipReader.each_file` yields one `Zlib::GzipRe...
-
- I'm in favor of adding top level `ruby2_keywords`.
-
- Makes sense to me, and agreed it's going to be a lot less confusing with it.
I'm not particularly for `ruby2_keywords` in general, but if we have `Module#ruby2_keywords` then I agree `main.ruby2_keywords` makes sense. -
- Extending pattern variables to non-local variables would make pattern-matching more complex and confusing. I reject the idea (at least for now). There might be a chance to make pinned variables (`^a`) to pinned expression in the future.
... -
05:15 AM Feature #16372: Allow pattern matching to bind instance variables
- shevegen (Robert A. Heiler) wrote:
> If I understood you correctly then the suggestion is to allow both local and
> ...
Yes, correct. And people might want to use @@variables and $variables too (I rarely use them, though) . -
- We have discussed about in-line pattern matching and made a decision that it should cause an exception instead of returning false. So this issue will disappear.
We might need to talk about the precedence for e.g. `a = n in 99 rescue n... -
12:49 PM Misc #16262: DevelopersMeeting20191128Japan
- Ugh, forgot one
* [Misc #16260] Symbol#to_proc behaves like lambda, but doesn't aknowledge it *(the title says it all)* -
11:58 AM Misc #16262: DevelopersMeeting20191128Japan
- * [Feature #15605] json library needs more frequent releases
* Should we bump the json version to 2.2.1? -
- * [Misc #16375] Right size regular expression compile buffers for literal regexes and on Regexp#freeze (lourens)
* Builds on Misc #16291 , I think there's potential to apply this pattern to other types at hooks outlined at the end of ... -
12:38 PM Revision 6f27fa4f (git): prefer class_serial over m_tbl
- Decades ago, among all the data that a class has, its method
table was no doubt the most frequently accessed data. Previous
data structures were based on that assumption.
Today that is no longer true. The most frequently accessed fiel... -
- I submit the backport changes from ruby core to flori/json at https://github.com/flori/json/pull/388
We need to bump version to ruby repository like 2.2.1 because the same version of flori/json and ruby core confuse ruby users.
I h... -
- The following script:
```ruby
def mutate
re = /foo/
state = re.instance_variable_get(:@state)
re.instance_variable_set(:@state, state.to_i + 1)
state
end
3.times do
p mutate
end
```
Output this:
```
nil
1... -
05:32 AM Feature #16373: RDoc for some of the Kernel methods cannot be found in Kernel module
- I'd like to share the context. `ri` is not the true problem to me.
I'm working to import the RDoc documentation to RBS files. (https://github.com/ruby/ruby-signature/pull/84) I believe having documentations in RBS makes sense, at leas... -
03:47 AM Bug #16376 (Closed): Stack-buffer-overflow in renumber_by_map in regcomp.c
- I found this bug in ruby regex engine. I also reported this issue to Onigmo(https://github.com/k-takata/Onigmo/issues/144).
I reported this to [email protected], but there has been no reply for more than 2 weeks. So I decide to rep... -
- matz, the following is the current behaviors. Do you mean which should be fixed?
(1) 2.7, via variable
```
b = {"b" => "b"}; p({x: "x", **b}) #=> current behavior: {:x=>"x", "b"=>"b"}
```
(2) 2.6, via variable
```
b = {"b... -
- I think this inconsistency is a bug. Double splat in hash expressions needs to work as interpolation.
Matz.
11/26/2019
-
- Personally I don't quite like method names that end in "fy alone. Crystal has .stringify
such as in **&.name.stringify** , which I always felt was a weird name - does not tell me much at
all. I have no particular opinion on the feature... -
04:12 PM Feature #16374 (Open): Object#nullify to provide scalars with the effect similar to Enumerable#reject
- How about adding a new method to `Object` class?
```ruby
class Object
def nullify &block
block.call(self) ? nil : self
end
end
'asdf'.nullify(&:empty?) #=> "asdf"
''.nullify(&:empty?) #=> nil
```
It can be used toge... - 06:57 PM Revision 177bc652 (git): * 2019-11-27 [ci skip]
-
- rename __builtin_inline!(code) to __builtin_cstmt(code).
Also this commit introduce the following inlining C code features.
* __builtin_cstmt!(STMT)
(renamed from __builtin_inline!)
Define a function which run STMT implicitly and call... -
-
- I still think disabling deprecation warnings is better default.
The warnings are useless for end users of applications written in Ruby.
Consider Ruby 2.7 will be distributed by an OS (such as Debian) and
some gems are not deprecat... -
- @nobu Ugh, sorry, I didn't check the latest one (just judged from the ticket state). All is awesome! Thank you.
-
- @zverok, I think cf934413 made it work as you described, doesn't it?
If something is wrong, let me know.
-
- > Sounds reasonable, and matz says “an error on end-exclusive ranges” at https://bugs.ruby-lang.org/issues/14784#change-81254 and ditto in the log of 20190829 dev meeting, but nothing about beginless/endless ranges.
@nobu Will it be f... -
-
- So are you running just `make test-all` normally? I misguessed that you reconstructed the tarball for rpm package or something. Sorry.
I have never seen the issue in CIs (I may overlook of course), and we have no such a report again... -
08:05 AM Bug #16361: TestEnv#test_fetch failure
- mame (Yusuke Endoh) wrote:
> I cannot reproduce the issue with ruby 2.7.0-preview3.
> ...
That is what I more or less said, quoting myself:
> Trying to execute this test independently, it passes just fine.
You have to run the who... -
02:03 PM Bug #14997: Socket connect timeout exceeds the timeout value for
- tenderlovemaking (Aaron Patterson) wrote:
> This really sounds like a bug to me. Please make a patch and I will apply it.
Do you mind taking a look at https://github.com/ruby/ruby/pull/1806? Based on my testing it's solving the prob... -
- Check keys
* by an internal table, instead of unstable dispatched results
* and by parsed key values, instead of escaped forms in the source -
- > Cons
> ...
I can not answer that - but personally I would prefer to keep it simple and show where it is defined
no matter what other explanations may exist. Personally I like a flat hierarchy so I concur.
I should however had al... -
- > I believe the real question is "why #then defined in Kernel and not in Object"
It makes sense to me. Do we have a reason to keep them in `Kernel` not in `Object`? -
- I believe that there is some complicated discrepancy between where the method "defined" physically and where it "belongs" logically (though, it is not completely clear to me why some of the methods defined where they are).
If we take ... -
- RDoc documentations for some of the `Kernel` methods including `Kernel#then` are registered as `Object` methods. I'd like to propose to move `Kernel` methods listed in `Object` class to `Kernel` class.
The following `ri` results are c... -
-
- Not in `p_kw` itself, which makes key and variable/expr pair only
now. -
-
- Update the target file itself of the dependency on this script.
Fall back to the current working directory if unwritable. -
- If I understood you correctly then the suggestion is to allow both local and
instance variables become assignable, when before pattern matching would only
be possible to local variables, yes?
If that is the case then I think this mi... -
- I tried this code, but got sytax error:
```ruby
case [1, 2]
in @a, @b
end
```
```
syntax error, unexpected instance variable, expecting '}'
in @a, @b
^~
```
I think it would be more useful if we could bind ins... -
-
-
09:20 AM Revision 1fe31737 (git): Slightly fix the warning message [ci skip]
- I'm not sure, but maybe it was gramatically incorrect?
-
-
- `**` operator is for keyword arguments, and in Ruby 2.6, non-symbol key is not allowed in keyword arguments. So in principle, `{x: 'x', **{'b' => 'b'}}` should raise an exception.
Because of the spec change of keyword arguments (#141... -
06:50 AM Bug #16371 (Closed): Inconsistent usage of Double splat operator
- Here is an issue with a weird behavior of ruby double splat operator:
```
a = {a: 'a'}
b = {'b' => 'b'}
{x: 'x', **a}
#=> {:x=>"x", :a=>"a"}
{x: 'x', **b}
#=> TypeError (hash key "b" is not a Symbol)
```
When I do that ... -
- https://github.com/ruby/ruby/commit/7bd7a013b14fb6c4403ff243ec5db13c5c71f552/checks?check_suite_id=328682567#step:10:1634
```
*** Following extensions are not compiled:
openssl:
Could not be configured. It will not be installed.
... -
- Attached is a patch that allows interrupting the iteration. It uses `rb_thread_check_ints` in each loop iteration for Range#count calls (unless a block is passed, as the block case is already interruptable).
-
- I don't think infinite loop is a bug, but having no chance to `^C` is definitely wrong. Must be able to interrupt.
```
zsh % gdb --args ./miniruby -e '(1..).count'
GNU gdb (Ubuntu 8.2-0ubuntu1~18.04) 8.2
Copyright (C) 2018 Free So... - 02:46 AM Revision 307e899d (git): * 2019-11-26 [ci skip]
- 02:44 AM Revision 7bd7a013 (git): Fix `$(RMALL) -r` on Windows
- `set recursive=1 &` sets `1 ` to `recursive`, not `1`.
-
- I wrote this code:
``` ruby
n = 0
ret = n in 99
p ret
```
I expected `ret` was `false` but was `0` (truthy value).
I had to write like this:
```ruby
n = 0
ret = (n in 99)
p ret #=> false
```
However, I would like t... -
01:02 AM Bug #16369: Segfault : crashed on child side of fork pre-exec thread_pool.rb
- After a bit more research, I think this might be the same issue as https://bugs.ruby-lang.org/issues/16288#note-2 so feel free to close if that is the case
-
- It occurs at random in development on my rails 6 app with puma server, so unfortunately I do not have very reproducible steps.
Maybe someone with more knowledge of ruby crash reports could suggest some hints?
11/25/2019
-
- I agree caching `require 'openssl'` should be possible.
I'm OK with changing (or removing) that spec to specify the new behavior, but would like to hear others' opinions as I don't much about the history and rationale behind the current... -
06:44 PM Bug #15856: Performance of redundant `Kernel.require` is slow when many gems are activated
- For what its worth:
We went ahead and incorporated the prior patch into chef's shipping packages starting with Chef 15.5.9, and it will be incorporated into all our shipping products that bake in ruby (inspec, chefdk, chef workstation... -
08:19 PM Bug #16368 (Closed): Error in `YAML.load(symbolize_names: true)` with null column YAML
- Trying to parse YAML which contains null column `symbolize_names: true` would raise `NoMethodError`
```
irb(main):001:0> require 'yaml'
=> true
irb(main):002:0> YAML.load("null: nullnull")
=> {nil=>"nullnull"}
irb(main):003:0> ... -
06:22 PM Bug #16367 (Third Party's Issue): stdin not support ibm866 encoding for windows
- see here https://github.com/piotrmurach/tty-prompt/issues/115
-
- Sounds like a bug - counting to infinity is not easy. :)
-
03:28 PM Bug #16366 (Closed): .count on endless range causes infinite loop
- Out of curiosity, I tried:
``` ruby
(1..).count
```
Which resulted in an unkillable 100% CPU irb session.
I expected either an exception or maybe `Infinity`
Thanks -
- The difference from 0b8c73aa65add5c57b46b0cfdf4e661508802172 is to
add the result of `string_add` event to marking objects.
```C
RNODE($1)->nd_rval = add_mark_object(p, $$);
``` -
- [ci skip]
-
-
06:00 AM Misc #16262: DevelopersMeeting20191128Japan
- * [Feature #16166] Remove exceptional treatment of *foo when it is the sole block parameter (sawa)
* Unintended arity. This must be fixed in an earlier stage before Ruby 3.
* [Feature #16274] Transform hash keys by a hash (sawa)
*... -
- If calling builtin functions with no arguments, we don't need to
calculate argv location. -
-
- Hello. It is definitely a good idea to enhance our CI. But that alone does not improve the code quality. We need someone to fix issues.
Maybe we need a platform maintainer first. -
- Eregon (Benoit Daloze) wrote:
> jeremyevans0 (Jeremy Evans) wrote:
> ...
There are a ton of other changes in 2.7 that could affect performance besides `ruby2_keywords`.
> > If you want to measure the effect of `ruby2_keywords`, you... -
- jeremyevans0 (Jeremy Evans) wrote:
> > * Remove `ruby2_keywords` in Ruby 3.0, just have it in Ruby 2.7 where it's needed. Ruby 3.0 (with the keyword arg separation) doesn't need `ruby2_keywords`.
> ...
I don't think anyone is seriously... -
- https://github.com/ruby/ruby/compare/master...eregon:no-ruby2_keywords is a more extensive removal of `ruby2_keywords`, although not complete yet.
It's actually a lot of extra logic in `setup_parameters_complex`. -
-
01:47 AM
Misc #16365 (Closed): [PATCH] Remove duplicated line in NEWS
- Applied in changeset commit:git|a5f0d1d71598451dd8461ba52a3a143cb33afdd3.
----------
Remove duplicated line in NEWS [ci skip]
[ruby-core:95935] [Misc #16365] -
- I found a duplicated line in NEWS and created a patch to fix it.
- 01:46 AM Revision a5f0d1d7 (git): Remove duplicated line in NEWS [ci skip]
- [ruby-core:95935] [Misc #16365]
-
- [ci skip]
- 12:19 AM Revision ce50af21 (git): Fix the docs for Proc#>>.
- The docs are wrong about the behaviour of `#>>` (looks like it was copied from `#<<`)
In `(prc >> g).call(n)` _prc_ is called first (with _n_), *then* _g_ is called with the result.
Code examples are OK.
11/24/2019
-
- * [Feature #16345] Don't emit deprecation warnings by default. (mame)
* The discussion seems to be agreeing with deduplicated warnings [Feature #16289]. We must decide. -
- * [Misc #16188] The performance overhead of `ruby2_keywords` (eregon)
* It's about 7-8% (@mame measurements) for `foo(*args)` calls where `foo` has little code, both for MRI and TruffleRuby (I measured about 5-11% overhead). That's qu... -
09:19 AM Misc #16262: DevelopersMeeting20191128Japan
- * [Feature #16355]Raise NoMatchingPatternError when `expr in pat` doesn't match(ktsj)
* I'd like to fix this specification before 2.7 release. -
03:14 AM Misc #16262: DevelopersMeeting20191128Japan
- * [Feature #16363] Promote did_you_mean to default gem (yuki24)
* Currently there are two problems with the gem being a bundled gem, one with the availability of the lib and the other about bundler. -
- * [Feature #16364] Top-level ruby2_keywords (mame)
* Currently, there is no top-level `ruby2_keywords`, which would be unuseful for some simple cases. -
- jeremyevans0 (Jeremy Evans) wrote:
> Comparing 2.6 to 2.7 is irrelevant in regards to the discussion of the effect of `ruby2_keywords`. There are many other changes between 2.6 and 2.7 that have a much larger effect than `ruby2_keyword... -
- Eregon (Benoit Daloze) wrote:
> Here is another benchmark, where no keyword arguments are used, yet we see a slowdown of up to 11.5% in MRI 2.7:
> ...
As I explained in my previous comment, the majority of the slowdown is not related t... -
- Eregon (Benoit Daloze) wrote:
> jeremyevans0 (Jeremy Evans) wrote:
> ...
`def a(x) end; a(*arr) end` is the worst case performance decrease, assuming that what we are measuring is the effect of supporting `ruby2_keywords` when not actu... -
- Here is another benchmark, where no keyword arguments are used, yet we see a slowdown of up to 11.5% in MRI 2.7:
https://gist.github.com/eregon/31e155901c995925bd1c661dfa1a71d8
`length([1], [2], [3])` and `length({a: 1}, {b: 2}, {c: ... -
- I also measured on TruffleRuby, and there the diff is minimal, just adding the `ruby2_keywords` check on *splat call sites:
https://github.com/oracle/truffleruby/commit/d143af3626aae009e2414bfe61833565fe3a0476
The results are similar... -
- jeremyevans0 (Jeremy Evans) wrote:
> On CRuby master branch, in the worst possible case I could design, the difference was about 1%.
Why is `def a(x) end; a(*arr)` the worst case? Would it not be more expensive to call a method accep... -
- 09:50 PM Revision 90872520 (git): * 2019-11-25 [ci skip]
-
-
-
- 02:15 PM Revision 41164418 (git): * 2019-11-24 [ci skip]
-
- I cannot reproduce the issue with ruby 2.7.0-preview3.
```
$ make test-all TESTS=test/ruby/test_env.rb
fatal: not a git repository (or any of the parent directories): .git
./revision.h unchanged
config.status: creating ruby-runner... -
-
- I think this basically breaks [Feature #15865].
We should decide if `expr in pattern` can be used as a condition (such as in `if`) or not.
As @palkan said, it's already easy to use `|| raise NoMatchingPatternError` for assignment c... -
- \o/
(^^^ That's an upvote emoji really.) -
- The `did_you_mean` gem has been a bundled gem for over 3 years now and I think it's time to promote it to a standard library. There are two main issues in the current structure:
* A number of implementations, including [`optparse.rb`... -
- How about defining top-level ruby2_keywords?
```
#!/usr/bin/env ruby
def foo(**kw)
kw
end
ruby2_keywords def bar(*a)
foo(*a)
end
bar(k:1) #=> {:k=>1} with no warnings in 2.7
```
Currently, `Module#ruby2_keywords`...
11/23/2019
-
- mame (Yusuke Endoh) wrote:
> How did you run the test?
I am using the snapshot to execute the test suite. Therefore I assume that did_you_mean gem is available and the test suite should pass in this configuration. -
- How did you run the test? Currently, bundled gems should be unavailable in `make test-all`, and some tests don't assume that bundled gems are loaded.
Currently, did_you_mean is a bundled gem, so should not be loaded in `make test-all... -
- When building from tarballs, the source directory is not a git
repository. -
-
- Closing as per OP's request. Thanks.
11/22/2019
-
- and use github.workflow instead. Applied a similar change to MJIT too.
-
- Original Tracer.set_get_line_procs is implemented by
"def set_get_line_procs(p = proc)". It means that original
Tracer.set_get_line_procs supports block and Proc object. -
- Original Tracer.add_filter is implemented by "def add_filter(p = proc)". It
means that original Tracer.add_filter supports block and Proc object. -
08:42 PM Revision 8e743fad (git): Count pinned slots using only bitmap
- This is significantly faster than checking BUILTIN_TYPEs because we
access significantly less memory. We also use popcount to count entire
words at a time.
The only functional difference from the previous implementation is that
T_ZOMBIE... -
- Previously we would count the pinned objects on each comparison. Since
sorting is O(N log N) and we calculated this on both left and right
pages on each comparison this resulted in a extra iterations over the
slots. -
-
08:04 PM Bug #16362: Unable to build Ruby 2.6.5 on Manjaro Linux 18.1.3
- Sorry, I discovered that something was resetting my CFLAGS. Please close this.
-
- Steps to reproduce:
```
$ curl -OL https://cache.ruby-lang.org/pub/ruby/2.6/ruby-2.6.5.tar.gz
$ tar xvf ruby-2.6.5.tar.gz
$ cd ruby-2.6.5
$ ./configure
$ make
```
`./configure` summary:
```
* Installation prefix: /usr/... - 04:58 PM Revision 144bff43 (git): * 2019-11-23 [ci skip]
-
- `fatal: not a git repository (or any of the parent directories): .git`
-
- `git --version` failed as expected when git is not installed,
but unexpectedly pass when git installed and pwd is not in git working directory.
So use `git rev-parse` instead, and it failed when git installed too. -
- emscripten fails to compile ruby due to lack of this.
-
- I am facing following test suite error:
~~~
1) Failure:
TestEnv#test_fetch [/builddir/build/BUILD/ruby-2.7.0-f9d20a1bf1/test/ruby/test_env.rb:123]:
[ruby-core:56062] [Feature #8649].
Expected Exception(KeyError) was raised, but ... -
01:43 PM
Bug #16359 (Closed): Psych cannot build against system libyaml.
- Applied in changeset commit:git|994435d0efa9931d0475de923c4e6c227d03e398.
----------
Revert "Update dependencies"
This reverts commit e1b234148829f65bea9f5ecc7018beb782ea6023.
This allows to build Psych against system libyaml again on... -
- The attached patch which reverts the Psych part of [1] appears to fix the issue for me.
[1]: https://github.com/ruby/ruby/commit/e1b234148829f65bea9f5ecc7018beb782ea6023 -
- Just to be sure that we are building Ruby in Fedora against system libyaml library, the first step is to remove bundled libyaml contained in Psych [1]. Now the build started to fail:
~~~
make[2]: Entering directory '/builddir/build/B... - 01:42 PM Revision 994435d0 (git): Revert "Update dependencies"
- This reverts commit e1b234148829f65bea9f5ecc7018beb782ea6023.
This allows to build Psych against system libyaml again on Fedora.
[Bug #16359] -
12:37 PM Misc #16360 (Closed): Enabling IBM PowerPC/Z cases in Travis CI
- We added `arm64-linux` and `arm32-linux` cases to Travis CI by [the ticket](https://bugs.ruby-lang.org/issues/16234).
The `arm32-linux` case is going to be stable after [this pull-request](https://github.com/ruby/ruby/pull/2686) will be... -
-
- 04:58 AM Revision f9d20a1b (git): * 2019-11-22 [ci skip]
-
- Try to fix download error on Solaris CI
https://rubyci.org/logs/rubyci.s3.amazonaws.com/unstable11x/ruby-master/log/20191121T162422Z.fail.html.gz
```
tool/downloader.rb:243:in `rescue in download': failed to download config.guess (Runti...