Activity
From 09/18/2019 to 09/24/2019
09/24/2019
- 11:24 PM Revision 8e8dd88c (git): * 2019-09-25 [ci skip]
-
11:24 PM Revision 5294ded6 (git): [ruby/io-console] Skip cursor position test on Solaris
- It results in a mysterious failure.
https://github.com/ruby/io-console/commit/e3543c3da4 -
10:39 PM Bug #16178: Numbered parameters: _1 should be the same as |x| and _0 should not exist
- It's all about definitions. How do we explain numbered parameters?
Isn't `_1` the first parameter, as in `x` in `{ |x| }` and `x` in `{ |x,y| }`?
And therefore `_2` the second parameter as in `y` in `{ |x,y| }`?
Of course, `{ |x| ... -
- Here is another inconsistency on current trunk:
```ruby
[1,[2,3]].map { |x| x } # => [1, [2, 3]]
[1,[2,3]].map { |x,| x } # => [1, 2]
[1,[2,3]].map { _1 } # => [1, 2]
[1,[2,3]].map(&-> { _1 }) # => [1, [2, 3]]
[1,[2,3]].map(&... -
- Dan0042 (Daniel DeLorme) wrote:
> `proc{ |x,| }.arity` == 1, so `_1` is consistent with that.
Which sounds like another bug to me, because that block happily accepts more arguments, and should be identical to `|x,*|`, which `proc {|x... -
- zverok (Victor Shepelev) wrote:
> Honestly, we use it pretty frequently in our codebase and find it appealing.
> ...
How frequently compared to just `|x|`? I would guess < 5% on any non-trivial codebase.
But you are right, frequency o... -
09:38 PM Bug #16178: Numbered parameters: _1 should be the same as |x| and _0 should not exist
- `proc{ |x,| }.arity` == 1, so `_1` is consistent with that.
In order to get the tuples' first value you would need to do `array_of_arrays.map{_2;_1}` because otherwise `_1` would mean the entire tuple.
This argument is really weird... -
04:04 PM Bug #16178: Numbered parameters: _1 should be the same as |x| and _0 should not exist
- Hmmm. I find underscores to be harder to read in combination with
other parts (e. g. _abc or _123) whereas I use _ quite a bit in my
own code.
I still like @1 @2 @3 etc... - however had, I have to admit that if
the suggestion is to... -
01:53 PM Bug #16178: Numbered parameters: _1 should be the same as |x| and _0 should not exist
- Some counter-points (that being said, I dislike the "unnamed parameters" idea as a whole, because I am afraid this feature -- being just "shorter" while staying imperative -- will stay in the way of adoption of more promising features li...
-
- nobu (Nobuyoshi Nakada) wrote:
> When `_1` is same as `|x|`, what does `[[1, 2]].map {_1 + _2}` mean?
It means `[[1, 2]].map { |a,b| a + b } # => [3]` of course.
> ...
Yes, just like `Proc#arity` changes.
```ruby
-> { _1 }.arity... -
- When `_1` is same as `|x|`, what does `[[1, 2]].map {_1 + _2}` mean?
The meaning of `_1` changes if `_2` is used or not?
-
- Of course, having `_` as the only unnamed parameter would have `|x|` semantics,
but I guess it's too late for that and now we have `_<n>` parameters. -
- Currently on trunk:
```ruby
array = ["string", 42, [1, 2]]
array.map { |x| x * 2 }
# => ["stringstring", 84, [1, 2, 1, 2]]
array.map { _1 * 2 }
# => ["stringstring", 84, 2]
```
Oops, this trivial code just lost data and com... -
- I ended up rebasing my commits to the latest master and then fiddling with parse.y until I got everything clean and working, including Ripper.
So this is my implementation of omitted parameters based on nobu's original patch:
https://g... -
07:10 PM Bug #16173: ENV.delete returns nil when name does not exist and block given
- I also agree that nobu's patch should be merged.
-
06:31 PM Bug #16173: ENV.delete returns nil when name does not exist and block given
- I, being a member only of The Outer Party, can't merge anything here.
I'm agnostic on whether the change should be to the documentation or to the code. My part was just to note the discrepancy.
Can someone in The Inner Party take ... -
07:08 PM Bug #15244: Method #extname return empty string if filename is dot ('.')
- Example from .NET:
(https://docs.microsoft.com/en-us/dotnet/api/system.io.path.getextension)
```
class Program
{
static void Main(string[] args)
{
string ext = Path.GetExtension("..jpg");
Console.WriteLine(ext);
... -
- As KysyCore.dll is included in the backtrace, that is likely to be the cause of the problem. Is it possible for you to test on a system that doesn't have Kysy loaded when running the tests?
-
06:27 AM Bug #16135: Segmentation fault in Ruby 2.6.1
- Kysy is not installed in the machine .
The above call stack is from my development environment which has Kysy codebase but not Kysy library installed -
-
-
-
12:52 PM Revision a1dcb9da (git): Rename from ruby-x.y.z.ext.draft to ruby-x.y.z-draft.ext [ci skip]
-
- lldb module bundled with Xcode is for Python 3 now.
-
-
- How to reproduce:
1. Create dummy files by `./miniruby -e '%w[a b foo bar test zzz].each{|basename|File.write("#{basename}.rb", "raise %(do not load #{basename}.rb)")}'`
2. Run `ruby/test_autoload.rb` like `make test-all TESTS='ruby... -
- https://github.com/ruby/io-console/commit/9377e37295
-
- To be consistent with `winsize`, changed the cursor position
format from `[x, y]` to `[row, column]`.
https://github.com/ruby/io-console/commit/d1f5ae9286 -
- It seems crashing in ffi-call by rb-readline.
-
02:50 AM Revision 2366c681 (git): suppress meddlesome clang10 warrning
- It says:
vm.c:2519:34: warning: expression does not compute the number of elements in this array; element type is 'const struct __jmp_buf_tag', not 'VALUE' (aka 'unsigned long') [-Wsizeof-array-div]
sizeof(e... -
- cadfaacb2533d47d52dbb5dbefe724d7bf11112e missed it.
09/23/2019
-
- Attached is an alternative approach for implementing this. It uses `VALUE` for the `kwfreeze` variable, so we can use `Qundef` instead of the magic value of `-1` for the default behavior. It also updates the documentation to reflect yo...
-
10:52 PM Bug #16176: Segmentation fault running irb on ruby 2.6.4p104 (2019-08-28 revision 67798) [x64-mingw32]
- Install is from https://github.com/oneclick/rubyinstaller2/releases/download/RubyInstaller-2.6.4-1/rubyinstaller-devkit-2.6.4-1-x64.exe however it seems to work fine under git bash (https://gitforwindows.org/)
me@box MINGW64 ~
$ git ... -
- I'm new to Ruby so bare with me, I have a fresh install of Ruby on a Windows 10 (Build 1803) corporate Dell laptop and getting segmentation fault just trying to launch irb
Dump below
-
10:06 PM Feature #16147: List Comprehensions in Ruby
- Eregon (Benoit Daloze) wrote:
> Just my opinion: I don't think we should have list comprehensions in Ruby.
Thanks I appreciate the feedback!
> ...
You have some very good points! Nested comprehensions can get pretty gnarly. Still... -
- nobu (Nobuyoshi Nakada) wrote:
> sammomichael (Samuel Michael) wrote:
> ...
Hi, I was suggesting a splat operator as one way to create a special syntax, under the hood instead of calling each method it would treat this as a filter map... -
- I think it's important to make a distinction between "immutable" and "frozen".
Some programming languages have immutable data structures, and some programmers find that concept really cool, functional, powerful and whatnot. This is a ... -
- When reading the code it seems like the intent of `foo(test:42)` is for baz to return `[{:test=>42}, {}]` but that's not what I get when running it on 2.7. I get a warning and `[1, {:test=>42}]`. To fix the code on 2.7 I need to change `...
-
- Since I wasn't at the developer meeting I'll post my thoughts/responses to the log here.
> matz: I prefer this style to .map(&:to_s). But I understand it is not flexible enough. Difficult to determine.
This shorthand simplifies one... - 04:28 PM Revision e561e4a8 (git): * 2019-09-24 [ci skip]
-
04:28 PM Revision 74e33662 (git): Make public_send and rb_f_send handle keyword argument separation
- Kernel#send takes a different optimized code path that was already
handled. -
- It looks like Kysy is still installed and active according to the C level backtrace:
```
C:\Windows\SYSTEM32\ntdll.dll(RtlCaptureContext+0x3c3) [0x00007ff9c23c9bd3]
E:\code\KYSY-3153\Source\x64\Release\bin\KysyCore.dll(Kysy::PCIAddr... -
- It works fine in Ruby 2.5.6 version
-
- Please find the error output of Segmentation fault with no Kysy installed.
```
Copying E:/code/KYSY-3153/Source/test/../RegisterDef/data/kysy_platdb.json to E:/code/KYSY-3153/Source/x64/Release/bin directory
Run options: --seed 3459... -
- https://github.com/ruby/io-console/commit/b8017509ef
-
- https://github.com/ruby/io-console/commit/a54b6e4dd1
-
08:33 AM Revision b443bdbd (git): Use short wait for select(2)
- It is one of the reasons why paste to IRB is slow.
-
-
12:25 AM Bug #16151: [PATCH] Fix a class of fstring related use-after-free
- I ran some quick tests on rejecting frozen non-bare strings from the buffer
deduplication code path. On Discourse, about 18% of calls to rb_fstring are
refused with the change in https://bugs.ruby-lang.org/issues/16151#note-5.
So I do...
09/22/2019
-
- At the last developer meeting, matz requested that instead of "k", "e", and "n" modifiers to `rb_scan_args`, a new `rb_scan_args_kw` method be added for that behavior. I have updated the pull request to add that method.
matz also had... -
- IMHO it's better to fix behavior to be consistent with Hash#delete.
And the compatibility risk seems non existent here.
So I think we should change behavior, and not change documentation which would be inconsistent with Hash#delete.
... -
06:29 PM Feature #16119: Optimize Array#flatten and flatten! for already flattened arrays
- nobu (Nobuyoshi Nakada) wrote:
> This patch unrolls the `while`-loop for the already flatten head and seems reasonable.
> ...
Attached is a patch for early return and preventing additional allocs for the empty array case (however not v... -
-
-
- vo.x (Vit Ondruch) wrote:
> However, I wonder if there is some explanation for disabling the test. I see a lot of comments such as `# These clocks in practice on Linux do not seem to match their reported resolution.` or `# These clocks ... -
- Just my opinion: I don't think we should have list comprehensions in Ruby.
They are just a small subset of the far more powerful Enumerable methods, and IMHO are often less readable (e.g., nested list comprehensions in Python are what... -
- sammomichael (Samuel Michael) wrote:
> ```ruby
> ...
It is an `Array` from 0 to 10 now.
> below we propose a syntax in which we splat the for loop to return the stored result not the caller
> ...
Obviously the latter result conflic... -
04:38 PM Bug #14716: SecureRandom throwing an error in Ruby 2.5.1
- I am encountering this in a project as well. In my case, I see this while running rspec regression tests:
```
#<ActionView::Template::Error: failed to get urandom>
./.rvm/rubies/ruby-2.6.3/lib/ruby/2.6.0/securerandom.rb:104:in `uran... -
- 04:04 PM Revision 2afe86e4 (git): * 2019-09-23 [ci skip]
-
- https://sevenzip.osdn.jp/chm/cmdline/switches/method.htm
-
- nobu (Nobuyoshi Nakada) wrote:
> I'm positive about this, except for the performance.
> ...
Using the rdoc gc bench tooling Koichi used in https://bugs.ruby-lang.org/issues/14858, it's difficult to get consistent results between this c... -
- It disables 7z, which seems not to have an option to stop saving
extra file attributes (uid/gid and atime), in order to make zip
packages stable. -
- Extra file attributes (uid/gid and atime) make the packaged zip
file unstable. -
01:12 PM Revision 5f35b8ca (git): st.c: Use rb_st_* prefix instead of st_* (#2479)
- The original st.c was public domain hash table implementation, but
Ruby's st.c is highly modified, and its data structure is not
compatiblie with the original one.
Therefore, when creating an extension library to wrap C code that uses
t... -
- I agree with Jeremy. It is not a bug.
I think that the wording of `freeze: false` is a bit confusing, but the document explains clearly. Do you have a real-world use case for `clone(freeze: true)`?
Anyway, please do not modify th... -
- @jeremyevans0 thanks for your answer.
Let me explain my point a bit.
I come upon this inconsistency (or what **I** see as an inconsistency) when working on 2.4 version of my [RubyChanges](https://rubyreferences.github.io/rubychange... -
- I sort of agree with zverok at the least for the expectation of freeze: true
working. IMO for boolean toggle-values I would think it is simpler to have
both variants work; then again I don't think I need either of the two variants
mys... -
-
- It was originally static inline, but seemed to be accidentally published
at 8f675cdd00e2c5b5a0f143f5e508dbbafdb20ccd. -
- This function was created as a variant of st_copy with firing write
barrier.
It should have more explicit name, such as st_copy_with_write_barrier.
But because it is used only for copying iv_tbl, so I rename it to
rb_iv_tbl_copy now. If... -
- instead of deprecated BigDecimal.new.
-
- nobu (Nobuyoshi Nakada) wrote:
> I'm positive about this, except for the performance.
> ...
Apologies for the delay in replying.
Using `benchmark-driver` script (running set last as it would taint the others by initializing the loca... -
- `check-snapshot-ruby_2_4` uses `make test` instead of `make check`.
https://github.com/ruby/actions/blob/95692e54f4263501db1508d51c1fb67c03b4f575/.github/workflows/snapshot.yml#L448-L449
`draft-release` use `make check` to make it simpl...
09/21/2019
-
11:24 PM Bug #16174: Feature #15958 triggers Rails unit test failure
- Thanks for the fix. https://github.com/rails/rails/pull/37262 has been merged to master, I think this issue can be closed.
-
- https://github.com/rails/rails/pull/37262
-
- This Rails unit test compares the output of Time using pp.
pp Time class object also changed to show subseconds.
* Ruby 2.6.4
```
$ ruby -v
ruby 2.6.4p104 (2019-08-28 revision 67798) [x86_64-linux]
$ irb
irb(main):001:0> requi... -
- Let me inform one of Rails Active Record unit test get failed since https://bugs.ruby-lang.org/issues/15958
I prefer this change but let me inform because it may be one of incompatibility.
* Steps to reproduce
``` ruby
git clon... - 11:10 PM Revision c020fd6a (git): * 2019-09-22 [ci skip]
-
- Applied in changeset commit:git|2e551356a7a6e74ba07283e000ff16f5d1ea6506.
----------
Make Kernel#{Pathname,BigDecimal,Complex} return argument if given correct type
This is how Kernel#{Array,String,Float,Integer,Hash,Rational} work.
Bi... -
- Applied in changeset commit:git|7470f965650bf17875632f0c5f9e5a4d9de9fc3f.
----------
Fix Module#class_variables for singleton classes of classes/modules
Module#class_variables should reflect class variable lookup. For
singleton classes... -
- This is how Kernel#{Array,String,Float,Integer,Hash,Rational} work.
BigDecimal and Complex instances are always frozen, so this should
not cause backwards compatibility issues for those. Pathname
instances are not frozen, so potentially... -
- Module#class_variables should reflect class variable lookup. For
singleton classes of classes/modules, this means the lookup should
be:
* Singleton Class
* Class
* All Ancestors of Class
Note that this doesn't include modules included ... -
- In reiteration of my main points, Ruby is a dynamic multi-paradigm language which should when possible embrace a variety of modes to achieve the same result.
SO what would be the benefits of list comprehension in Ruby and what wou... -
- An interesting bit here is that there is no way to get back the module from which the method was retrieved, even though it's shown in #inspect:
```
> Range.instance_method :map
=> #<UnboundMethod: Range(Enumerable)#map>
```
Maybe ... -
- `nil` does not respond to `to_hash` though, how do you propose to deal with that?
Should `**` call `to_h` rather than `to_hash`, similar to `*` calling `to_a` and not `to_ary`?
Your comment https://bugs.ruby-lang.org/issues/15381#n... -
- I worked on an alternative approach of using `ruby2_keywords` with a hash flag approach instead of a VM frame flag approach: https://github.com/ruby/ruby/pull/2477 . This approach is significantly less invasive to implement, and I think...
-
- The `freeze: false` option was intended to be: "do not freeze clone if receiver is already frozen". The `Object#clone` documentation states: "#clone copies the frozen (unless `:freeze` keyword argument is given with a false value) and ta...
-
- In #12300, the new keyword `freeze:` was introduced, allowing this:
```ruby
h = {}.freeze
h.clone.frozen?
# => true
h.clone(freeze: false).frozen?
# => false
```
Though, it turns to me that behavior is not symmetric:
```ruby
... -
- I'm still against frozen-string-literal by default. It is arguable if the string creation limits performance so much in real-world programs. We need to first measure how much Ruby can be faster by frozen-string-literal. If it is not s...
-
- Thanks, @alanwu. I'm refreshing diff.txt with more fulsome documentation, along with enhanced testing.
-
- Thanks, @nobu, but I'm not going to propose a change to the functionality.
-
- Comparing with `Hash#delete`, it looks that the document is correct and the code is wrong.
```diff
diff --git i/hash.c w/hash.c
index 8b84a14484..7880178dc8 100644
--- i/hash.c
+++ w/hash.c
@@ -4779,7 +4779,7 @@ env_delete_m(VALU... -
- Matz didn't determine this proposal at the meeting. There were four points discussed:
* Before matz accepts this proposal, he must decide the grand design change: Ruby should be immutable by default?
* If we want to make Ruby gradua... -
- Hmm. I can somewhat understand the proposal, but I believe you would also
add a bit more complexity here; you need to remember that ruby users may
have to understand the idea behind this too, and that adds a bit of
complexity. A bit ... -
- to suppress variable unused warning
-
- I published the Log. Closing.
-
- I've added a pull request that adds warnings to setting/access of $SAFE, as well as public C function that deal with $SAFE: https://github.com/ruby/ruby/pull/2476
As the taint tracking/mechanism is being removed, I was not sure if we ... -
07:09 AM Revision b4149993 (git): Add a benchmark for JIT-ed code dispatch
-
04:24 AM Revision e006b992 (git): typedef rb_jmpbuf_t to void *[5] if __builtin_setjmp is used
- The built-in version operates on a buffer of 5 words, much smaller than
the size of jmp_buf defined in libc.
Note, powerpc requires 5 words, while arm and x86_64 just require 3.
09/20/2019
-
- This gives the impression that the return value is always nil, which isn't true when the environment variable exists.
Maybe it's worth clarifying that. -
- Attached diff.txt:
* ENV.delete for nonexistent name and block given:
* Test enhanced to verify return value is nil.
* Documentation corrected to say that return value is nil, not value.
-
- Applied in changeset commit:git|27144de2bd95c5886c3d4400c3b8aea932b856f1.
----------
Fix documentation for ENV.each to return ENV
Also have spec check that it returns ENV.
Mostly from burdettelamar@yahoo.com (Burdette Lamar).
Fixes [... -
- I'm attaching a diff that:
* Enhances the test to verify the returned object, which is ENV.
* Corrects the documentation to say that the returned object is ENV, not a Hash.
-
- Also have spec check that it returns ENV.
Mostly from burdettelamar@yahoo.com (Burdette Lamar).
Fixes [Bug #16164] -
- Inadvertently merged change to suppress warnings by gcc 9.2.
Pointed out by Alan Wu. -
- Can you please post or upload the error output from the failure with Ruby 2.6.3 and no AMD Kysy installed? Also, if possible, please post a self contained minimal example that reproduces the failure.
-
- I have tested in a clean machine with no Kysy ,but it failed there too.
-
- I already tried with Ruby 2.6.3 too and faced same issue .
- 03:48 PM Revision 990e714f (git): * 2019-09-21 [ci skip]
-
-
- Use PRIdPTRDIFF for the platforms where a pointer is larger than a long.
-
- rb_vm_call_kw handles the tmp buffer for you.
Also, change method_missing so it also calls rb_vm_call_kw to
handle the kw_splat flag, instead of requiring callers to handle
kw_splat flag before calling method_missing. This may fix othe... -
- When Object#to_enum is passed a block, the block is called to get
a size with the arguments given to to_enum. This calls the block
with the same keyword flag as to_enum is called with.
This requires adding rb_check_funcall_kw and
rb_ch... -
- If defined in Ruby, dig would be defined as def dig(arg, *rest) end,
it would not use keywords. If the last dig argument was an empty
hash, it could be treated as keyword arguments by the next dig
method. Allow dig to pass along the em... -
-
- `iseq->body->jit_unit->compile_info` should not be referenced before
the null check of `iseq->body->jit_unit`. -
- in vm_push_frame and limit scope of i.
Just a minor maintainability improvement. -
11:32 AM Feature #16153: eventually_frozen flag to gradually phase-in frozen strings
- Dan0042 (Daniel DeLorme) wrote:
> The naming is a bit wonky, I'll freely admit. I'll change the description to your suggestion.
> ...
The details and pros and cons were discussed at yesterday's Ruby committer meeting, with Matz present... -
-
-
- for debug counters
```
../include/ruby/intern.h:1175:137: warning: passing argument 3 of 'rb_define_singleton_method0' from incompatible pointer type [-Wincompatible-pointer-types]
#define rb_define_singleton_method(klass, mid, func, a... -
08:35 AM Bug #16121: Stop making a redundant hash copy in Hash#dup
- I found that
> ` rb_define_method(rb_cArray, "initialize_copy", rb_ary_replace, 1);`
`Array#initialize_copy` == `Array#replace`. Can we use same technique? -
- Thank you. Completely fine.
Do you want to merge via github PR or attached patch by my commit? -
08:12 AM Feature #16170: Remove the unmaintained libraries from Ruby 2.7
- Just FTR, I am using tracer from time to time, typically just from command line as `-rtracer`. I'll be sad to see it go away ...
-
- NMAKE sets MAKE to the full path name, which includes spaces by
the default installation. -
- * Removed excess backslashes
* Fixed the target name to try failure.cpp -
- [Feature #11297] [Feature #16123]
-
- I wanted to briefly add that I think C++ support or have it as a "first class citizen"
could be great - even though competent programmers such as matz or linus are not
necessarily the biggest fan of C++. :-)
I like some parts of C+... -
- We discussed this topic at the developer meeting held yesterday.
The attendees' mood was that we would like to say using C++ for extension libraries is just okay.
However because C and C++ have diverged for ages, it is not trivial to... -
- The idea is very interesting to me, purely from a conceptual point of view
alone. So from this point of view, I like the idea itself, or the thoughts
behind the idea.
HOWEVER had, at the same time, I actually dislike the syntax.
... -
05:22 AM Revision 1de242de (git): Fixed up 37c083942509dd7e0d8b7ccd3575936a2017c400
-
-
-
-
-
-
- Can someone update the "Log" link above?
-
02:10 AM Feature #15381: Let double splat call `to_h` implicitly
- I came here to file this feature request, only to find this had already been proposed. This would be beautiful, indeed. :-)
-
- * array.c (flatten): fix a memory leak in the case of an exception
at conversion of an element to Array. -
-
-
- This patch unrolls the `while`-loop for the already flatten head and seems reasonable.
But I could get only insignificant result, and a worse result for an unflattened array.
```yaml
# array_flatten.yml
prelude: |
ary = (0...100...
09/19/2019
-
- @mame
I think that class is used frequently, though I'm not sure if Rails tries to deduplicate instances with fstrings. Regardless, it seems that removing
buffer deduplication altogether isn't very workable, since we need to backport... -
- The naming is a bit wonky, I'll freely admit. I'll change the description to your suggestion.
Does this mean you agree with the basic idea? -
- `freeze(eventual: false)` gives the impression that it will never be frozen, but the proposal is to have this mean that it is frozen immediately.
So it would be better to say `freeze(immediately: true)` for what happens now, and `froz... -
06:25 PM Feature #11297 (Closed): Allow private method of self to be called
- Applied in changeset commit:git|7fbd2f7cc247ee66e877ab3c88f0274834c6b6c7.
----------
Allow calling a private method with `self.`
This makes it consistent with calling private attribute assignment
methods, which currently is allowed (e.... -
- Applied in changeset commit:git|7fbd2f7cc247ee66e877ab3c88f0274834c6b6c7.
----------
Allow calling a private method with `self.`
This makes it consistent with calling private attribute assignment
methods, which currently is allowed (e.... -
08:17 AM Feature #16123: Allow calling a private method with `self.`
- Accepted. But as I said in #11297, the document should be updated. After the patch, you can say `self.puts("hello")`, which can confuse you.
Matz.
-
- [Feature #11297] [Feature #16123]
-
- [Feature #11297] [Feature #16123]
-
- [Feature #11297] [Feature #16123]
- 05:20 PM Revision 7fbd2f7c (git): Allow calling a private method with `self.`
- This makes it consistent with calling private attribute assignment
methods, which currently is allowed (e.g. `self.value =`).
Calling a private method in this way can be useful when trying to
assign the return value to a local variable ... -
- Define overloading functions of rb_define_virtual_variable and
rb_define_hooked_variable, for combinations with and without
ANYARGS casts. -
- By using template and overloading, instead of transparent union.
-
- At the dev meeting yesterday, matz recommended changing the method name from `pass_keywords` to `ruby2_keywords`. I have updated the pull request to do that: https://github.com/ruby/ruby/pull/2449
akr disliked the approach of using a... - 03:20 PM Revision 5e3771aa (git): * 2019-09-20 [ci skip]
-
- > `Time.now.floor(step: 3600)` doesn't work well with leap seconds.
```ruby
n = (Time.now.to_i / 86400).floor * 86400
ENV["TZ"] = "UTC"; Time.at(n) #=> 2019-09-19 00:00:00 +0000
ENV["TZ"] = "right/UTC"; Time.at(n) #=> 2019-09... -
- As @akr stated, `round` etc.for `Time` class do not work well (rejected).
For float values, we could suffer from errors. Unless there're real-world use-cases, I am not positive.
Matz.
-
09:22 AM Feature #16021: floor/ceil/round/truncate should accept a :step argument
- `Time.now.floor(step: 3600)` doesn't work well with leap seconds.
Also, we want floor/round/ceil for month and year
but one month and one year is not fixed number of seconds.
So, step with number of seconds doesn't for month and yea... -
-
-
-
01:26 PM Feature #16131: Remove $SAFE, taint and trust
- I look forward to removing all tainting logic!
-
- Basically agreed.
My proposal for the schedule:
2.7:
* Remove taint tracking/mechanism.
* Non-verbose warning on setting/access of $SAFE
* taint/trust/untaint/untrust become no-ops, verbose warning when called
3.0:
... -
- To substitute suffixes and VPATH for nmake.
-
-
- This reverts commit 69e209a3450bd6b281dcad1d96a34e9cab184845.
The debug has finishted. -
11:45 AM Revision cbf405fe (git): Specify encoding explicitly for sprintf
-
- Applied in changeset commit:git|5208c431bef3240eb251f5da23723b324431a98e.
----------
Separate Time#inspect from to_s and show subsec [Feature #15958] -
-
-
-
- to fix another keyword argument warning which was added recently.
-
08:35 AM Feature #16163 (Closed): Reduce the output of `RubyVM::InstructionSequence#to_binary`
- Applied in changeset commit:git|20baa08d652b844806fab424a2a590408ab613ef.
----------
Improve the output of `RubyVM::InstructionSequence#to_binary` (#2450)
The output of RubyVM::InstructionSequence#to_binary is extremely large.
We have ... -
- The output of RubyVM::InstructionSequence#to_binary is extremely large.
We have reduced the output of #to_binary by more than 70%.
The execution speed of RubyVM::InstructionSequence.load_from_binary is about 7% slower, but when reading ... -
- Accepted. PR welcome.
Matz. -
- @nobu Could you please review this?
-
-
- I'll review it.
-
-
08:07 AM Feature #15667: Introduce malloc_trim(0) in full gc cycles
- I'd just like to point out that calling malloc_trim(0) according to article explaining the feature - https://www.joyfulbikeshedding.com/blog/2019-03-14-what-causes-ruby-memory-bloat.html - will not fix memory fragmentation - it'll just m...
-
- Accepted.
Matz.
-
- I'm positive about this, except for the performance.
Do you have any numbers? -
- I'm positive about this, except for the performance.
Do you have any numbers? -
- FYI: How many times the gems is required from all latest public gems by using [gem-codesearch](https://github.com/akr/gem-codesearch).
```
$ csearch "require .mutex_m.$" | wc -l
199
$ csearch "require .scanf.$" | wc -l
119
$ csea... -
- >Can we remain it as bundled gem?
sounds good.
We are going to remove them at Ruby 2.7.0-preview2
* cmath
* scanf
* shell
* sync
* thwait
We need to investigate the following stdlibs with irb. After resolve dependencies,... -
- I'll maintain `debug`. Can we remain it as bundled gem?
-
- Can we also remove them?
* debug
* getoptlong
-
- For frozen `Symbol#to_s`, I see a clear benefit. But I worry a little bit for incompatibility.
So how about making an experiment by the next preview(2) to see how big the incompatibility is?
`String#to_s` etc. is a different story. we ... -
- This bug should be fixed. @nobu could you review the patch?
Matz. -
- See also https://travis-ci.org/ruby/ruby/jobs/586452224
-
- I noticed that in case of cache misshit, re-calculated cc->me can
be the same method entry than the pevious one. That is an okay
situation but can't we partially reuse the cache, because cc->call
should still be valid then?
One thing t... -
05:56 AM Misc #16114: Naming of "beginless range"
- I hope this discussion leads to Ruby having Ranges congruent to all the mathematical intervals.
Today we have right-unbounded ranges, and left-unbounded ranges, and right-closed ranges, and left-closed ranges, and right-open ranges --... -
- This `*[]` is not just exception list omitted, but explicitly specifies zero exceptions to catch.
Thus the current behavior is intended.
Matz.
-
- We discussed about this in developer meeting.
This feature has some difficulty in spec for example padding, unknown type (other than AF_INET, AF_INET6, AF_UNIX, and so on), which should it use for the comparison.
Though such difficulty... -
- What's the use case of this comparison?
-
- > Is there a reason why the existing frozen-string-literal: false is not good enough?
My understanding of `frozen-string-literal` is that if true it makes string literals in the same file mutable. If false, it does not do that.
If ... -
- Make draft release packages. (Release packages are official after tests and release announce.)
- Copy from ruby/actions
- Change trigger tags from `draft/v*` to `v*` (I use `draft/v*` on ruby/actions because I want to avoid to cause tro... -
-
-
- rb_vm_call0 allocates its own struct call_info etc. But they are
already there in case of rb_funcallv_with_cc. Let's just pass the
existing ones, instead of re-creation. -
-
-
- [Misc #16172] C++ situation?
- What do we do with C++ extensions? -
- (This issue is MRI specific)
Few weeks ago I pushed a minimalist C++ compatibility layer to the master branch. Then Nobu pushed his MakeMakefile::CXX implementation. It seems our C++ support got suddenly improved. What is our curre... -
-
-
- https://ruby-doc.org/ is not our resource. Please contact admin of ruby-doc.org
09/18/2019
-
-
- @alanwu
Indeed, it makes sense. Thank you.
I am not familiar with Rails at all so I could be wrong, but I hear ActiveSupport::SafeBuffer inherits String and has an instance variable. Is this class not so used frequently?
https... -
- @mame
Thanks for pointing that out. I should clarify what I meant there. In a world where strings don't track taint status, it might make more sense to refuse buffer deduplication for non bare strings.
At that point, my original patc... -
- @alanwu
Just FYI: You said in https://bugs.ruby-lang.org/issues/16152#note-8
> * What do you think about the proposed schedule?
> ...
But, BARE_STRING_P is not only for a tainted string, but also for a string that includes an inst... -
- Replace [arg=1, args] with [arg, args] so we can actually test
the value correctly.
Add some missing tests for **h3 when method accepts (**args).
Add tests for passing positional hashes to (**args) methods and
check for the expected wa... -
- Documentation for PP is in https://ruby-doc.org/stdlib-2.4.6/ but it's no longer in https://ruby-doc.org/stdlib-2.5.6/ or subsequent versions.
Was that intentional? - 03:11 PM Revision 2b708ab7 (git): * 2019-09-19 [ci skip]
-
- Look up language module with `MakeMakefile.[]`, insted of a
accessing constant under that module directly, to get rid of
expose the constant to the toplevel inadvertently. -
-
- * [Feature #16120] Omitted block argument if block starts with dot-method call
* `posts.map{ .author.name }` has *beautiful* readability, no incompatibility, changes a syntax error into an intuitive idiom, already implemented in 6 li... -
- [Feature #16170] Remove the unmaintained libraries from Ruby 2.7
* Is it ok to remove them? -
- It is exposed to the topleven namespace via included MakeMakefile.
-
-
-
- Agreed. Might also be simpler to get full support in gem-spec format
for those libraries that are maintained. :)
Some of them I have not even read before at all, such as mutex_m. -
- I'm working to extract standard libraries to rubygems in recent years. So, the standard libraries will be promoted the default gems or bundled gems.
But I suggest removing some of the unmaintained libraries without default/bundled gem... -
- > And I have a path forward that would easily address any breakage: we add a `mutable-string-literal` pragma.
I don't understand, why a new pragma? Is there a reason why the existing `frozen-string-literal: false` is not good enough?
... -
- > This depends on the ruby code. Some projects will be semi-dormant due to various reasons.
That's for us to address as a community. Are we going to let a single decade-old gem prevent us from moving Ruby forward? What's the threshold... -
- @mame
> Ruby allows to omit parentheses of method calls
Oh, thanks! In fact, that was the piece of understanding I missed trying to catch how space- and spaceless-versions are different from PoW of the interpreter.
The ticket co... -
- Isn't it there? http://rubies.travis-ci.org/
-
-
-
-
-
09:26 AM Revision effacde2 (git): [bundler/bundler] Merge all `bundle check --path` specs
- And skip them all for bundler 3.
https://github.com/bundler/bundler/commit/b88936cdc2 - 09:26 AM Revision 434af730 (git): [bundler/bundler] Merge #7340
- 7340: Fix bundle clean issue r=deivid-rodriguez a=deivid-rodriguez
### What was the end-user problem that led to this PR?
The problem was that `bundle clean` is crashing under some conditions.
### What was your diagnosis of the proble... -
- It was suggested a deprecated command as a fix.
https://github.com/bundler/bundler/commit/e330a9a34f -
- https://github.com/bundler/bundler/commit/6e9774b377
-
09:26 AM Revision 4f2f6aeb (git): [bundler/bundler] Add a spec for installing git deps after packaging w/o git
- https://github.com/bundler/bundler/commit/65351c58b8
-
- This reverts commit 3dc509e645abb497e4dc92a7c42be471ff87db0b.
https://github.com/bundler/bundler/commit/b5766564fb -
- This reverts commit 00b095b98fe4bd44950beaf3bc9f1d91eac7b69e.
https://github.com/bundler/bundler/commit/e93bce3b20 -
- This reverts commit 52c5a0eedec34b5d86464b3cf135dc2002486f1d.
https://github.com/bundler/bundler/commit/b4cc36deb9 -
- This reverts commit 3a2d2f025081755bdb38af660897e7b2f749a33a.
https://github.com/bundler/bundler/commit/13cef81582 - 09:26 AM Revision c27aaf1a (git): [bundler/bundler] Fix comments and messages to refer to https url
- https://github.com/bundler/bundler/commit/a86b49f1b9
-
- https://github.com/bundler/bundler/commit/b7fc6f4187
-
- To not mention that the flag is remembered when it's not.
https://github.com/bundler/bundler/commit/82f0b95854 -
- https://github.com/bundler/bundler/commit/0a0e7cf5ec
-
08:29 AM Revision 9118cb24 (git): Fix typos
-
- Why do you use private attr_accessor? You can access its instance variable directly.
Allowing `self.private_method` has a unpreferable side effect; it allows `self.global_function`, i.e., `self.require "foo"`, `self.lambda { }`, `sel... -