Activity
From 05/25/2021 to 05/31/2021
05/31/2021
-
11:48 PM Bug #17933 (Closed): `Net::HTTP#write_timeout` doesn't work with `body_stream`
- While testing `Net::HTTP#write_timeout` using the server example from the [Feature Issue](https://bugs.ruby-lang.org/issues/13396), I've noticed that [Faraday](https://github.com/lostisland/faraday) multipart requests (with a file parame...
-
10:43 PM Misc #17932 (Rejected): 90s design (please lets move to 21st century)
- I always welcome to your improvement at https://github.com/ruby/b.r-l.o .
-
08:45 PM Misc #17932: 90s design (please lets move to 21st century)
- I do not understand what "search is broken" means. Could you explain what you tried to do? Searching in Ruby Issue Tracking System seems to just work for me.
-
07:07 PM Misc #17932: 90s design (please lets move to 21st century)
- Typo
-
- Hello there,
I want to modernizer the "[Ruby Issue Tracker](bugs.ruby-lang.org) because every time I access that site, I want to throw up 🤮¹
Here my different proposals to solve this MAJOR issue:
a) Update the design of [redmi... -
04:09 PM Bug #17887: Missed constant lookup after prepend
- Eregon (Benoit Daloze) wrote in #note-4:
> jeremyevans0 (Jeremy Evans) wrote in #note-3:
> ...
That I do not know. Most of the talk when prepend was added was about method lookup, but the same issues apply to constant lookup.
> > T... -
11:21 AM Bug #17887: Missed constant lookup after prepend
- jeremyevans0 (Jeremy Evans) wrote in #note-3:
> This is because unlike method tables, constant tables are not moved to origin classes.
Do we know if this was somehow intentional or more like an oversight/incorrect optimization? I w... - 02:47 PM Revision 9e0e9902 (git): bump the verion of RDoc for previous merge
-
02:46 PM Bug #17877: Backport a fix of RDoc for CVE-2021-31799
- ruby_2_7 483f303d02e768b69e476e0b9be4ab2f26389522.
-
- seems that the test never passes on Windows because cannot create such file.
@aycabta, how do you think? - 02:44 PM Revision 483f303d (git): merge revision(s) a7f5d6ab88 c9ab8fe2 [Backport#17877]
- a fix of RDoc for CVE-2021-31799
-
02:34 PM Revision c9ab8fe2 (git): The test for command injection on Unix platforms should be omitted on Windows
- 02:24 PM Revision ff6f4e63 (git): tool/ci_functions.sh: Fix typos and improve the comment. [ci skip]
- 02:10 PM Revision d8bbbc30 (git): bump patchlevel for previous merge commit
-
- merged into ruby_2_7
https://github.com/ruby/ruby/commit/29bbad939939c6dceb804aac667ba372fdee4ef5 -
02:04 PM Revision 29bbad93 (git): Fix 2.7 build (#4359)
- * merge revision(s) fcc88da5eb162043adcba552646677d2ab5adf55:
configure.ac: fix for upcoming autoconf-2.70
The failure initially noticed on `autoconf-2.69d` (soon to become 2.70):
```
$ ./configure
... - 02:01 PM Revision fd95a180 (git): merge revision(s) d8a13e504992a45d52063f7c925408d7aad3595a: [Backport #17780]
- [Bug #17780] Fix Method#super_method for module alias
Method#super_method crashes for aliased module methods because they are
not defined on a class. This bug was introduced in
c60aaed1856b2b6f90de0992c3477183001... -
12:31 PM Feature #17930: Add column information into error backtrace
- Hi, I like this idea!
This feature would be quite useful when you get `NoMethodError` in the production environment and error tracking tools like [Sentry](https://sentry.io/welcome/) inform you of that error.
Consider the example I've at... -
09:30 AM Feature #17930 (Closed): Add column information into error backtrace
- Consider the following code and error.
```
data["data"].first["field"] #=> undefined method `[]` for nil:NilClass
```
There are two possibilities; the variable `data` is nil, or the return value of `first` is nil. Unfortunately, ... -
12:26 PM Revision 39547990 (git): Tweak skipped files in bundler gemspec
- We won't be using the `extra_rdoc_files` field, because it's very slow
for markdown files. -
11:30 AM Bug #17931: Compile fails setup option nodynamic
- my setup file
``` shell
option nodynamic
#bigdecimal
#cgi/escape
#continuation
#coverage
#date
#dbm
#digest/bubblebabble
#digest
#digest/md5
#digest/rmd160
#digest/sha1
#digest/sha2
#etc
#fcntl
#ffi_c
#fiber
fiddle... -
- I can compile the ruby ​​normally without deselecting the #option nodinamyc and some other extension, but when deselecting I get the following error
``` shell
rbconfig.rb updated
generating enc.mk
making srcs under enc
generating ... -
- I think `binding.debug` is best. It's not that long to type, and people are already used to e.g. `binding.irb/pry`.
Too short feels cryptic and easier to get accidentally committed. -
02:43 AM Feature #17863: rewrite lib/debug.rb with latest API
- ko1 (Koichi Sasada) wrote in #note-4:
> I want to avoid `Kernel#...` for compatibility.
> ...
or both? (`#bp` and `#debug`). mame-san said `binding` is too long and he doesn't want to write it :p -
- I want to avoid `Kernel#...` for compatibility.
If `binding.debug` makes sense, I prefer it.
-
- * [Bug #17887] Missed constant lookup after prepend (eregon)
* Is it a bug? (I think so, feels inconsistent and makes `prepend` useless for constants) Should it be fixed? How about Jeremy's PR?
-
- I can compile Ruby normally without --with-static-linked-ext, but when I try to use this command the following problem occurs:
``` shell
win32ole_typelib.c
compiling C:/ruby-3.0.1/ext/win32ole/win32ole_variable.c
win32ole_variable.... -
- https://github.com/ruby/net-protocol/commit/97c4b68528
-
05:38 AM Bug #17928 (Closed): Backport 1aac0e88193a82ed36b43e852c46414181b66455 to Ruby 3.0
-
04:46 AM Bug #17928 (Closed): Backport 1aac0e88193a82ed36b43e852c46414181b66455 to Ruby 3.0
- Please consider baqckporting commit:1aac0e88193a82ed36b43e852c46414181b66455 to ruby_3_0. This is a fix for a bug related to an optimization introduced in Ruby 3.0.
-
- Applied in changeset commit:git|1aac0e88193a82ed36b43e852c46414181b66455.
----------
Mark inlined ISeqs during MJIT compilation (#4539)
[Bug #17584] -
- [Bug #17584]
- 12:26 AM Revision 2cc3639e (git): * 2021-05-31 [ci skip]
-
12:25 AM Revision 43ac55ad (git): Update bundled_gems
05/30/2021
-
11:19 PM Bug #17927: INSTALL CLTypingLabel 0.4.0 ON MacBook Pro M1
- I tried but it doesn't work at all.
I tried to reinstall cocoapods with an old version but I always get this error -
-
- I believe this is an [Ethon](https://github.com/typhoeus/ethon) bug, similar to #17862 or #17922.
-
- I HAVE A BUG ABOUT CLTypingLabel 0.4.0 WHEN I WANT TO INSTALL ON TERMINAL
Analyzing dependencies
/Library/Ruby/Gems/2.6.0/gems/ethon-0.14.0/lib/ethon/curls/classes.rb:36: [BUG] Illegal instruction at 0x0000000102da0000
ruby 2.6.3p6... -
09:37 AM Bug #17926 (Closed): spec/ruby/core/file/atime_spec.rb: a random failing test on Travis ppc64le
- I just observed the following test failing randomly on Travis ppc64le at the commit:fd65ef2a5aa8629676d1edb6410e4d4cf60b8045. When running this commit on my forked repo, the test passed.
https://travis-ci.com/github/ruby/ruby/jobs/509... -
08:41 AM Bug #17925: Pattern matching syntax using semicolon one-line
- > NOTE 1: I understand that only syntax that doesn't use case and end is experimental one-line pattern matching syntax.
A little supplement. The following is also an experimental one-line pattern matching syntax since Ruby 3.0, but th... -
- This is because unlike method tables, constant tables are not moved to origin classes. This behavior has been present since origin classes were added in Ruby 2.0. I submitted a pull request to work around this: https://github.com/ruby/...
05/29/2021
-
09:08 PM Revision fd65ef2a (git): Attempt to fix floating point test failure
- The previous behavior depending on exact float values, it seemed
to work OK on amd64 and i386, but other CI platforms are
experiencing non-deterministic test failures with it. Relax test
slightly to hopefully pass on such platforms. -
- The same failure observed in WSL2 (Ubuntu) CI:
http://rubyci.s3.amazonaws.com/wsl2/ruby-master/log/20210529T190003Z.log.html.gz -
- Applied in changeset commit:git|f516379853f36d143d820c55d5eeaa9fc410ef52.
----------
Fix Enumerator::ArithmeticSequence handling of float ranges
Depending on the float range, there could be an off-by-one error,
where the last result th... - 03:56 PM Revision 9bb76201 (git): * 2021-05-30 [ci skip]
-
- Depending on the float range, there could be an off-by-one error,
where the last result that should be in the range was missed. Fix
this by checking if the computed value for the expected value
outside the range is still inside the range... -
- ## Summary
There are the following differences between `case ... when` and` case ... in`. Is this an expected behavior?
```console
% ruby -v
ruby 3.1.0dev (2021-05-28T16:34:27Z master e56ba6231f) [x86_64-darwin19]
% ruby -ce ... -
08:51 AM Bug #17864: [BUG] try to mark T_NONE object (in Ractors)
- I didn't know those existed, thanks a lot.
I could test with: ruby 3.1.0dev (2021-04-28 master 8e2ac2140d) [i386-mingw32]
I ran my 200 times command twice and both had a success rate of 100%. I guess ruby 3.1.0dev solves the issue :)... -
- ruby_3_0 a21ec8d1ecc6e978ea6b18a27046c424e2849752 merged revision(s) 636d4f7eb9f3fcb088e1a44af4181c4aa36789b4.
-
- ruby_3_0 d47df50678b00bd622e6be474031204ed2e52b31 merged revision(s) 39a2ba5cc559900c30c3143da32446c2f20a7484.
-
- ruby_3_0 7b6a2ad04a3272a31323493133498dfc60d77d76 merged revision(s) 4b36a597f48c857aa5eb9ed80fec0d02f6284646.
-
- Avoid setting the visibility of refinement method entries
Since refinement search is always performed, these entries should always
be public. The method entry that the refinement search returns decides
the visibi... -
- Method cache: fix refinement entry handling
To invalidate some callable method entries, we replace the entry in the
class. Most types of method entries are on the method table of the
origin class, but refinement ... -
- Fix setting method visibility for a refinement without an origin
class
If a class has been refined but does not have an origin class,
there is a single method entry marked with VM_METHOD_TYPE_REFINED,
bu... -
- withdrawn backport request for now.
I think this is a spec change in practice even if the prior behavior violates any standards.
If you have any objections, please let me know. -
- ruby_3_0 20bf397f7cebd8156d76941dd54659cf9b8c8fe7 merged revision(s) fd8991f797c145615820b4c76e6b7ac5a71339ea.
-
- ruby_3_0 9d63a8dd0beefa49796eb86ab688c2eb0cd89a20 merged revision(s) 5cdf99f64e344b8e4638824d55f5caf33be682ca.
-
- Fix lazy enumerator with index size
Fixes [Bug #17889]
---
enumerator.c | 7 ++++++-
test/ruby/test_lazy_enumerator.rb | 4 ++++
2 files changed, 10 insertions(+), 1 deletion(-) -
- Honestly, I cannot judge did this change keep backward compatibility.
I mean, is the current master works fine with -gdwarf-4 option or old GCC? -
- mkmf: fixed install directories of header files in extension
libraries [Bug #17761]
When installing an extension library which provides a header, that
header should be installed under site_ruby (or vendor_ruby w...
05/28/2021
-
- It compiles & run completely with those changes.
Thanks a lot. -
- user:byroot #note-1 looks like another bug. Please create a separate ticket if need be.
-
- This issue reminded me of #17606, which seems to actually affect also i386-mswin32 according to:
https://docs.microsoft.com/en-us/cpp/build/reference/zp-struct-member-alignment
If so, the following patch is worth applying:
```di... -
01:10 PM Bug #17923: try to mark T_NONE object with miniruby x86 Win32 buildt with MSVC
- I've just hit the same bug on linux:
```
<OBJ_INFO:gc_mark_ptr@gc.c:6605> 0x00007f1d22e9fc00 [2 M ] T_NONE
doesnt-matter.rb:224: [BUG] try to mark T_NONE object
ruby 3.1.0dev (2021-05-27T21:17:54Z master 2ad7a550f3) [x86_64-linu... -
- Hello, I was asked to open a new bug ticket for this specific issue in this thread: https://bugs.ruby-lang.org/issues/17864#note-6
When I try to build ruby using "Visual Studio 2019 Developer Command Prompt v16.9.4".
```<OBJ_INFO:g... -
- Now #note-5 has a separate ticket: #17923
Back to the main issue, there are pre-built binaries of Ruby master (head) in the wild. You may want to use them if you fail to build yourself, although they might not be the latest.
https:... -
06:25 PM Feature #17873: Update of default gems in Ruby 3.1
- byroot (Jean Boussier) wrote in #note-16:
> Just for the record:
> ...
Thanks.
I'll repeat: could we consider to still bundle these gems in next major version but with some warning? If you "update" the gem then the conversion warnin... -
- The `net-protocol` one seems similar to: https://github.com/ruby/net-imap/issues/16.
-
- > I'll have a look at this.
Thank you!
I used digest as an example, but there is also `net/protocol`: https://buildkite.com/rails/rails/builds/77709#8f4e417a-153d-403c-9ad0-6579b3c04adf/999-1010 -
- This sometimes happens when `rubygems` loads a default gem too early before `bundler` has been setup. Then when user code loads the actual gem it uses the `$LOAD_PATH` setup by bundler instead of the default one, causing the double load....
-
- Actually, there's a bit of a problem. If you add a default gem to the gemspec it cause double loading issues, e.g. with `mail`, `net-smtp` pulls `digest`:
```
/opt/rubies/2.7.2/lib/ruby/2.7.0/digest.rb:6: warning: previous definition... -
- Just for the record:
- The matrix removal broke the `capybara` gem: https://github.com/teamcapybara/capybara/pull/2468
- The net-smtp remove broke the `mail` gem: https://github.com/mikel/mail/pull/1439
Both are Rails dependen... - 04:34 PM Revision e56ba623 (git): * 2021-05-29 [ci skip]
-
04:34 PM Revision 5ada23ac (git): compile.c: Emit send for === calls in when statements
- The checkmatch instruction with VM_CHECKMATCH_TYPE_CASE calls
=== without a call cache. Emit a send instruction to make the call
instead. It includes a call cache.
The call cache improves throughput of using when statements to check the... -
- It looks for "checkmatch", when it could be applied to anything that has
"newrange".
Making the optimization target more ranges might only be fair play when
all ranges are frozen. So I'm putting a reference to the ticket that
froze all ... -
02:46 PM Feature #17924 (Open): Range#infinite?
- A boolean method on Range returning true if the range is infinite and false if it is not.
-
- > Are you talking about the code duplication of rubygems and bundler? This ticket is about default gems, isn't it?
Yes, I'm making a general case against maintaining the same code at two different places, but as I worded this issue, i... -
- @deivid @k0kubun Are you talking about the code duplication of rubygems and bundler? This ticket is about default gems, isn't it?
Because @hsbt is promoting default gems to bundled gems step by step (for example, #17873), I think this... -
-
-
-
- This variable had a typo (it's `@gemhome`), but the test is still
passing, so I assume it's not needed.
https://github.com/rubygems/rubygems/commit/3b88642bdb -
- https://github.com/rubygems/rubygems/commit/a678959eda
-
- https://github.com/rubygems/rubygems/commit/0efb894c3b
-
- https://github.com/rubygems/rubygems/commit/8dfe1e30b5
-
- https://github.com/rubygems/rubygems/commit/c77868a555
-
- https://github.com/rubygems/rubygems/commit/aa390a3500
-
- * Port
https://github.com/ruby/ruby/commit/8e91b969df08b7a2eb27a5d6d38733eea42dc7ad
from ruby-core, and make it compatible with psych 3 & 4. -
- There is no reason to care for minitest gem.
05/27/2021
-
- I think a middle ground would be that on one hand some ruby-core contributors get direct write access to upstream repos to commit their test patches directly, but on the other hand this gets acknowledged as an issue and future improvemen...
-
- I don't have any issues with that since as of now, commits from ruby-core contributors to keep ruby-core CI green don't go through any review process and are just ported unconditionally. But I would prefer if you created a PRs upstream w...
-
- ruby/ruby's CI maintainers often modify a test code of rubygems and bundler from ruby/ruby because we don't have write access to these repositories. Do you think their owners can share write access to these repositories to ruby/ruby main...
-
- It's been 6 months, I guess it could be worth asking whether anything has changed here.
I'm happy to be proved wrong but ruby-core run of bundler & rubygems tests has never detected any realworld problem with either rubygems or bundle... -
10:07 PM Feature #8460: PATCH: optparse: add `keep_unknown` option
- Nobody had requested any tests. It would have been much easier if 7 years ago you had written the two lines of code required to test this.
Anyway, I've rebased my changes on top of the latest master, and added some tests. -
- Applied in changeset commit:git|fd8991f797c145615820b4c76e6b7ac5a71339ea.
----------
Fix lazy enumerator with index size
Fixes [Bug #17889] -
- This turned out to be fairly simple to fix: https://github.com/ruby/ruby/pull/4534
-
- I didn't bisect, but this is probably related to commit:85e43e1dfecef69b935c48c235cc20f21bd4f0d4, as it affects 2.7 and 3.0 but not 2.6. I'll see if there is a way to fix this.
- 09:17 PM Revision 2ad7a550 (git): * 2021-05-28 [ci skip]
-
- Fixes [Bug #17889]
-
- From the crash dump, this appears to be an issue in ethon (a third-party gem), not Ruby. Possibly related to https://github.com/typhoeus/ethon/issues/180, but maybe Apple Silicon specific?
-
- The situation looks similar to that of #17862. You might want to open an issue on https://github.com/typhoeus/ethon/issues/ if one does not yet exist.
-
07:57 PM Bug #17922 (Third Party's Issue): pod install 'Firebase/Auth' results in [BUG] Bus Error
- **Steps:**
Attempted to install Cocoapods for Google Firebase after installing Cocoapods on new Mac mini
Please see details below in the following sections (bolded): Environment, Pod File, Mac Terminal Output, Crash Report
... -
- What's wrong with this, which works fine?
```ruby
def transaction
begin
puts "Begin Transaction"
result = yield
success = true
result
ensure
if success
puts "Commit Transaction"
else
pu... -
- ioquatix (Samuel Williams) wrote in #note-14:
> Here is working example:
> ...
You can simplify this slightly using a local variable:
```ruby
def transaction
failed = false
begin
puts "Begin Transaction"
... -
11:13 AM Feature #15567: Allow ensure to match specific situations
- Here is working example:
~~~ ruby
def transaction
failed = false
begin
puts "Begin Transaction"
yield
rescue Exception
puts "Abort Transaction"
failed = true
raise
ensure
puts "Commit Transaction" unless ... -
- I revisited this situation again.
This behaviour is very confusing:
~~~ ruby
def transaction
begin
puts "Begin Transaction"
yield
rescue
puts "Abort Transaction"
else
puts "Commit Transaction"
end
end
catch... -
11:03 AM Feature #17853: Add Thread#thread_id
- Thanks a lot for building this, @naruse!
If I can ask for a tiny addition, it would be really cool to get the `native_thread_id`, if available, in `Thread#inspect` :) -
- I just would update the current status.
Travis was revived with only free non-x86 cases. (#17818)
Here is a list of the possible servers that I plan to install Cirrus CI persistent workers.
* arm64 Ubuntu focal server, if we can... -
-
-
- I applied this proposal at https://github.com/ruby/ruby/pull/4530
But I keep to open this until releasing preview or rc version of Ruby 3.1. -
-
-
-
-
-
-
-
-
-
-
05/26/2021
-
-
- So after digging more I noticed the String was coming from the `memcached` client, and created by this dirty hack: https://github.com/arthurnn/memcached/blob/e5390741c008c4c7cbb91634009bfd57e99ec851/ext/rlibmemcached/rlibmemcached.i#L159...
- 09:22 PM Revision eae7fd0e (git): * 2021-05-27 [ci skip]
-
09:21 PM Revision 8fdb15fd (git): Fill out switch statement in push_mark_stack
- When objects are popped from the mark stack, we check that the object is
the right type (otherwise an rb_bug happens). The problem is that when
we pop a bad object from the stack, we have no idea what pushed the bad
object on the stack.... -
05:21 PM Bug #17889 (Closed): Enumerator::Lazy#with_index should return size
- The following example shows `#size` returns `nil` when the `size` was provided to `new` as `3`.
**Example**
``` ruby
p Enumerator::Lazy.new([1, 2, 3], 3){|y, v| y << v}.with_index.size
```
**Expected Result**
``` ruby
3
```
*... -
- I also think that the same `.ancestors` should yield the same constant lookup result.
It would be harder for users to predict constant lookup otherwise.
By the way, the repro script has behaved the same since [Ruby 2.2.6](https://wa... -
01:06 PM Misc #17888: "What's Here" section for class and module documentation
- It's always great when someone volunteers to work on documentation!
But doesn't the documentation already have a table-of-contents for each class? I see it down the left-hand-side here.
https://docs.ruby-lang.org/en/master/Array.ht... -
- https://github.com/nobu/ruby/runs/2672425246?check_suite_focus=true#step:15:239
A test in rbs passes keyword arguments to `pp`.
Also @ko1 says he uses keywords arguments with `pp` very often. -
- @tenderlovemaking thanks for fixing it! I removed the skipped gc tests on Travis ppc64le at the commit:af43198738bf45d55d91d7f48b197f94dc526967 .
-
- Applied in changeset commit:git|fc832ffbfaf581ff63ef40dc3f4ec5c8ff39aae6.
----------
Disable compaction on platforms that can't support it
Manual compaction also requires a read barrier, so we need to disable
even manual compaction on ... - 09:33 AM Revision af431987 (git): .travis.yml: Remove skipped tests on ppc64le.
- It was fixed at fc832ffbfaf581ff63ef40dc3f4ec5c8ff39aae6 .
-
- http://rubyci.s3.amazonaws.com/solaris10-gcc/ruby-master/log/20210526T070003Z.fail.html.gz
```
1) Error:
TestThread#test_thread_native_thread_id:
NotImplementedError: native_thread_id() function is unimplemented on this machine
/ex... -
- Ok, so I was able to reproduce with this small script:
```ruby
require "set"
iseq = RubyVM::InstructionSequence.compile <<~RUBY
class Id
attr_reader :db_id
def initialize(db_id)
@db_id = db_id
end
def ==(other)... -
- > I assumed you already knew it because of the title
Totally understandable.
Thanks for the explanation, however I'm afraid my understanding of the problem is still too limited to figure out a repro script. -
-
06:18 AM Revision 9ddc7674 (git): Add NEWS about 46655156dcc37509dcb69fcd0717c110eb1c624a
- * Add Thread#native_thread_id [Feature #17853]
-
- Applied in changeset commit:git|46655156dcc37509dcb69fcd0717c110eb1c624a.
----------
Add Thread#native_thread_id [Feature #17853] -
-
-
- 12:37 AM Revision 59928a75 (git): * 2021-05-26 [ci skip]
-
- Manual compaction also requires a read barrier, so we need to disable
even manual compaction on platforms that don't support mprotect.
[Bug #17871]
05/25/2021
-
- xtkoba (Tee KOBAYASHI) wrote in #note-6:
> user:mame wrote in #note-4:
> ...
Ya, that's correct. Even manual compaction requires the mprotect read barrier. Basically we need to disable compaction on platforms that don't support it. ... -
- Just to clarify my previous comment.
Unconditionally enabling bundler everytime `ruby` is used (i.e., from `rubygems.rb`) is not even on the table. I only made some progress on making the `RUBYGEMS_GEMDEPS` environment variable option... -
04:45 PM Feature #17883: Load bundler/setup earlier to make `bundle exec ruby -r` respect Gemfile
- To reiterate what Jeremy said, I have several projects where I have a
Gemfile, but never run the scripts with bundle exec. The Gemfile is
just to make sure that other developers have the necessary gems
installed.
-
-
- jeremyevans0 (Jeremy Evans) wrote in #note-16:
> I'm against this in the strongest possible terms. There are more ruby scripts not using bundler than there are ruby applications using bundler. We should not force the usage of bundler ... -
- nobu (Nobuyoshi Nakada) wrote in #note-15:
> Not true for `-I` at least.
> ...
Mmh, so it's more complex due to unshift'ing into `$:`, but still:
```
$ RUBYOPT=-Ia ruby -Ib -e 'puts $:'
/home/eregon/b
/home/eregon/a
/home/eregon/.rubies/... -
- mame (Yusuke Endoh) wrote in #note-10:
> BTW, I noticed that my patch could be even much simpler. If you like `BUNDLER_SETUP`:
> ...
While I still prefer a more generic approach, I'm OK with this idea as it allows the user to opt-in to... -
- Eregon (Benoit Daloze) wrote in #note-13:
> In other words, `RUBYOPT` should behave like `RUBYOPT=$RUBYOPT ruby $CLI_ARGS` -> `ruby $RUBYOPT $CLI_ARGS`.
> ...
Not true for `-I` at least.
Paths by `-I` in `RUBYOPT` (and `RUBYLIB`) are ... -
- I think the workaround to solve it for `bundle exec ruby -r` can be valuable (https://github.com/rubygems/rubygems/pull/4616).
It seems a much smaller issue that a script doing https://bugs.ruby-lang.org/issues/17883#note-5 does not w... -
- mame (Yusuke Endoh) wrote in #note-3:
> In principle, explicit command-line arguments are "stronger" than RUBYOPT. For `RUBYOPT=-W2 ruby -W0`, `-W0` wins; `-W2` is just ignored. For `RUBYOPT=-ra ruby -rb`, `-rb` wins; `b` is loaded earl... -
01:16 PM Feature #17883: Load bundler/setup earlier to make `bundle exec ruby -r` respect Gemfile
- mame (Yusuke Endoh) wrote in #note-10:
> (This is off-topic: Nowadays, many people are using Bundler, so it might be a good time to enable Bundler by default? We can start it with `require "bundler/setup" unless ENV["NO_BUNDLER"]` :-)... -
- > Command-line arguments are what a user passed explicitly, so it is natural to me to prioritize them over RUBYOPT.
Yes, that's where we disagree. In my opinion, prioritizing `-r` scripts means letting them run last, so that they can c... -
- Sorry for repeating myself again, but I don't think the current behavior of RUBYOPT is wrong.
Command-line arguments are what a user passed explicitly, so it is natural to me to prioritize them over RUBYOPT.
The issue is that RUBYOPT i... -
- I closed the bundler PR in favour of the approach being discussed here. Even if this was fixed in ruby, it would still make sense to add this logic to rubygems require so that this works on all supported rubies.
-
- In my opinion, all approaches not involving fixing `-r` priority feel hacky.
The `BUNDLE_BIN_PATH` environment variable is something I had considered to remove in the future since it's only used for monkeypatching rubygems helpers to se... -
- deivid (David RodrĂguez) wrote in #note-6:
> Would it be enough to check whether `-rbundler/setup` is included in `RUBYOPT`?
I think it is a possible design choice, though it is very hacky.
But in the first place, I think Bundler ... -
- Indeed @mame, good catch. It sounds like this can't really be fixed in general inside bundler.
In my opinion, making handling of RUBYOPT consistent as suggested by @Eregon would be the best way to fix this. I can see it could cause is... -
- deivid (David RodrĂguez) wrote in #note-4:
> special casing `bundle exec ruby` and adding `-rbundler/setup` before any other CLI args: https://github.com/rubygems/rubygems/pull/4616.
@deivid Thanks for your work, but it looks too ad-... -
06:55 PM Misc #17886: DevelopersMeeting20210617Japan
- * [Misc #17888] "What's Here" section for class and module documentation (BurdetteLamar).
* I feel this section is very useful; some have reservations; let's discuss whether and, if so, how. -
- Proposed: That a "What's Here" section is a useful enhancement to the documentation for a class or module.
Actually, this work is already begun; these are already merged into master:
- [Array](https://docs.ruby-lang.org/en/master/Ar... -
- Ah okay, I assumed you already knew it because of the title, but it was just what the error message says.
* opt_setinlinecache: YARV insn to set a constant value to inline cache. Once the cache is filled, it won't be called again unti... -
- > We probably want a unit test that reproduces this when the canary is enabled before doing that though.
I'd be happy to try to reproduce this in isolation, but I don't quite understand what the bug is about, so I have no idea where t... -
- > I guess that it is possible to write such library in C extension by using finalizers
Ah I see. Yeah I didn't think about C-exts. That could work, would probably need to pin all references to avoid compaction problems etc.
But it ... -
- byroot (Jean Boussier) wrote in #note-13:
> I'm not sure how one could write such library today without a WeakMap (or other Weak* datastructure) like I'm requesting.
I guess that it is possible to write such library in C extension by... -
- byroot (Jean Boussier) wrote in #note-12:
> > My understanding is that ObjectSpace::WeakMap is weak values
> ...
I actually asked this question on the ruby-core Slack a while ago and IIRC I could not get a clear answer.
Maybe @nobu knows... -
- > Instead of exploiting WeakMap, a dedicated cache library makes more sense.
I'm not sure how one could write such library today without a WeakMap (or other Weak* datastructure) like I'm requesting.
-
- > My understanding is that ObjectSpace::WeakMap is weak values
I believe it's both weak keys and weak values:
```ruby
> ...
=> #<ObjectSpace::WeakMap:0x00007fa56bf75ea0>
>> Object.new.tap { |o| m[o] = o }
=> #<Object:0x00007fa57... -
- mame (Yusuke Endoh) wrote in #note-10:
> This is just my opinion. I want this kind of feature (non-memory-leakable hash for object dedup), but I'm unsure whether WeakMap is actually usable for that because it is difficult to predict its... -
- In the ancestors, `M` is before `A`, so this seems like a bug:
```
B.ancestors # => [B, M, A, Object, Kernel, BasicObject]
```
Also note that with `B.include M`, which reports the same ancestors, it works as expected:
```ruby
mod... -
- I would close this ticket, as the gcc latest version is tested on the GitHub Actions compilers.yml now.
https://github.com/ruby/ruby/blob/master/.github/workflows/compilers.yml
-
- I would close this ticket, as I think Travis is stable by the commit https://github.com/ruby/ruby/commit/9d4266fd5555c9c4388e2e8592008d0e0d8ccf21 now.
-
- I would close this ticket as both Travis ppc64le and s390x were enabled at https://github.com/ruby/ruby/commit/9d4266fd5555c9c4388e2e8592008d0e0d8ccf21 . We do not see the issue https://bugs.ruby-lang.org/issues/16360#note-7 on Travis pp...
-
- The PR was merged to the master branch. Here is the main commit of the commits.
https://github.com/ruby/ruby/commit/9d4266fd5555c9c4388e2e8592008d0e0d8ccf21
- 12:42 PM Revision 2a29a531 (git): Add Travis badge image to `README.md`. [ci skip] (#4527)
- Align the order of the badges with ones on the wiki page.
https://bugs.ruby-lang.org/projects/ruby/wiki/CIServers -
-
- 10:38 AM Revision 9e56a22b (git): bootstraptest/test_ractor.rb: Skip an assertion on Travis arm64.
- Skip the assertion to test the `Ractor.select` from multiple ractors that rarely
fails on Travis arm64.
See <https://bugs.ruby-lang.org/issues/17878>. - 10:38 AM Revision 9d4266fd (git): .travis.yml: Update to use arm64/ppc64le/s390x.
- * Use only the free pipelines arm64/ppc64le/s390x without credits on Travis.
See <https://docs.travis-ci.com/user/billing-overview/#partner-queue-solution>.
* Add ppc64le case.
* Update the gcc to gcc-11.
* Remove unused logic.
* Add t... -
10:38 AM Revision 83efb2e4 (git): .travis.yml: Ubuntu 20.04 Focal Fossa
- This is a LTS. Must be better than Xenial.
- 10:38 AM Revision 4f735903 (git): Revert "Remove .travis.yml" to revive Travis.
- This reverts commit 6b978d542704a5614af5e9375c4b31b8d2618652.
-
-
- akr (Akira Tanaka) wrote in #note-8:
> nobu (Nobuyoshi Nakada) wrote:
> ...
I prefer @nobu's patch to mine. Since it was fine to @akr, how about committing it? -