Activity
From 01/29/2022 to 02/04/2022
02/04/2022
-
10:41 PM Bug #18501: [BUG] try to mark T_NONE object in RubyVM::InstructionSequence. load_from_binary
- > I'd like to let it run for a few more days to be fully certain though.
We ran it quite enough and saw no more crashes, so we're quite confident the patches work.
I'm marking 3.0 and older as DONTNEED because ne never noticed that... -
10:36 PM Bug #18501 (Closed): [BUG] try to mark T_NONE object in RubyVM::InstructionSequence. load_from_binary
- Applied in changeset commit:git|2a76440fac62bb0f6e53ccada07caf4b47b78cf9.
----------
[Bug #18501] Fire write barrier after hash has been written
Before this change the write barrier was executed before the key and
value were actually r... -
- @tenderlovemaking implemented a patch for this: https://github.com/ruby/ruby/pull/5525, we deployed it on our infra and so far the bug is gone. I'd like to let it run for a few more days to be fully certain though.
@tenderlovemaking, ... -
- Before this change the write barrier was executed before the key and
value were actually reachable via the Hash. This could cause
inconsistencies in object coloration which would lead to accidental
collection of dup'd keys.
Example:
1... - 10:27 PM Revision 06e96b92 (git): * 2022-02-05 [ci skip]
-
10:26 PM Revision 06a28ec4 (git): [DOC] Enhanced RDoc for io.c (#5527)
- Treats:
IO.binread (abbreviated to be like IO.binwrite).
IO.write
IO.binwrite
IO.copystream
IO#external_encoding
IO#internal_encoding
IO#set_encoding -
09:12 PM Bug #18570 (Closed): Please Backport 77fe1fca0abb56f7f07725c0a3803d53a315c853 from the ipaddr gem to Ruby 3.1
-
09:15 AM Bug #18570 (Assigned): Please Backport 77fe1fca0abb56f7f07725c0a3803d53a315c853 from the ipaddr gem to Ruby 3.1
- I've prepared to merge ipaddr-1.2.4 at https://github.com/ruby/ruby/pull/5526
But I have no grant for releasing ipaddr-1.2.4 gem.
@knu Can you release ipaddr-1.2.4 from the v1.2.4 tag of ruby/ipaddr or add me to ipaddr gem owners?... -
08:15 AM Bug #18570 (Closed): Please Backport 77fe1fca0abb56f7f07725c0a3803d53a315c853 from the ipaddr gem to Ruby 3.1
- For some reason the following code causes an exception with Ruby 3.1:
``` ruby
require "ipaddr"
([IPAddr.new("192.168.56.0/21").freeze].freeze).any?{ |range| range.include?(IPAddr.new("192.168.1.0")) }
```
which breaks vagrant with ... -
- I agree, this is expected behavior, since there is no instruction sequence for a composed proc. If you would like an instruction sequence, you need to manually compose the proc in Ruby.
-
-
05:54 PM Bug #18257: rb_mRubyVMFrozenCore is broken by GC run
- So the TracePoint API must use similar paths. Therefore, looking at vm_trace.c, I have noticed this [1] line:
~~~
trace_arg->self != rb_mRubyVMFrozenCore /* skip special methods. TODO: remove it. */) {
~~~
I have traced their ori... -
- It is actually reproducible everywhere. I was able to reproduce it on x86_64 although it fails less frequently.
-
04:10 PM Feature #18568: Explore lazy RubyGems boot to reduce need for --disable-gems
- byroot (Jean Boussier) wrote in #note-12:
> It's not a "one or the other" scenario. We can perfectly explore both.
Yes, but if we manage to slim RubyGems boot like lazy-rubygems does, it loads so little Ruby code that I think there i... -
- > That would only work on CRuby, so as such would be unusable on anything but CRuby.
> ...
It's not a "one or the other" scenario. We can perfectly explore both. -
11:49 AM Feature #18568: Explore lazy RubyGems boot to reduce need for --disable-gems
- > Over 80% of CRuby's base startup is due to eagerly booting RubyGems. We can do better!
It's not the main point here, but perhaps still worth noting: `--disable-gems` implies `--disable=error_highlight,did_you_mean`.
In my own brief l... -
- byroot (Jean Boussier) wrote in #note-9:
> Another thing that could be explored would be to leverage `RubyVM::InstructionSequence.compile_file(path)`. If we indeed have a mapping of the gems content, we could define `load_iseq`, at leas... -
- Another thing that could be explored would be to leverage `RubyVM::InstructionSequence.compile_file(path)`. If we indeed have a mapping of the gems content, we could define `load_iseq`, at least for the duration of rubygems's "boot".
-
- Treats:
IO.foreach
IO.readlines
IO.read
IO.binread -
- * [Feature #18571] Removed the bundled sources from release package after Ruby 3.2
* (dummy comment) -
- We shipped release package with the 3rd party source for mswin environment especially.
The current status is here:
* libyaml: psych gem, ruby repo and package always bundled it.
* libffi: only ruby package bundled it. fiddle ge...
02/03/2022
- 11:26 PM Revision 203b1fc7 (git): Add documentation for regexp emoji named character property
-
08:55 PM Feature #18438: Add `Exception#additional_message` to show additional error information
- > Now Exception#detailed_message has a responsibility to highlight the message, so it need to insert the class name
I see. Thanks for the explanation!
> ...
I don't think adding another option just for that is necessary. I'll first... -
02:06 PM Feature #18438: Add `Exception#additional_message` to show additional error information
- mame (Yusuke Endoh) wrote in #note-25:
> But I agree that this is not a big deal.
Exactly. At this point it's kind of a bikeshed, it's just that according to #note-1 Sentry displays the exception class above the message, so it would be ... -
10:38 AM Feature #18438: Add `Exception#additional_message` to show additional error information
- Thank you for your comment!
st0012 (Stan Lo) wrote in #note-24:
> I think it'll be fine but I'm not sure why the class name needs to be added as well?
The current error printer inserts the exception class name.
```
$ ruby -e '... -
- > My concern was that Sentry will now use #detailed_message instead of #message, but it will now include the exception class, unlike before. But if Sentry is ok with that there's no problem.
I think it'll be fine but I'm not sure why ... -
- Dan0042 (Daniel DeLorme) wrote in #note-22:
> My concern was that Sentry will now use #detailed_message instead of #message, but it will now include the exception class, unlike before. But if Sentry is ok with that there's no problem.
... -
08:07 PM Revision 9fab2c1a (git): Add the size pool slot size to the output of ObjectSpace.dump/dump_all
- 08:07 PM Revision a3caf196 (git): * 2022-02-04 [ci skip]
-
08:06 PM Revision af321ea7 (git): Move total_freed_pages to size pool
-
-
- > https://github.com/rubygems/rubygems/pull/4199
Interesting, that seems a helpful step, although I suspect there would still be a significant overhead over `--disable-gems`, as long as we load so much Ruby code during startup.
I t... -
04:17 PM Feature #18568: Explore lazy RubyGems boot to reduce need for --disable-gems
- > I think a faster boot time while still having RubyGems enabled by default would be good for all Rubies.
Thank you for providing those links... I knew I had seen that code but could not remember where it was located.
Making these ... -
02:21 PM Feature #18568: Explore lazy RubyGems boot to reduce need for --disable-gems
- I worked on https://github.com/rubygems/rubygems/pull/4199 a while ago which avoided all gemspec evaluation during boot if I recall correct, and made things faster. But I never wrapped it up, I'll try to make it happen and re-measure its...
-
- I think a faster boot time while still having RubyGems enabled by default would be good for all Rubies.
https://github.com/oracle/truffleruby/blob/14ec2c2673188d47374a0570cf036864fcafe0b3/src/main/ruby/truffleruby/core/truffle/gem_util.... -
- @deivid That is a good point.
For io-wait/io-nonblock it just seems very unlikely to have security vulnerabilities in them, but for `digest`/`strscan` that is less clear. -
- > Being a default gem is already enough to fix a security issue without a CRuby release.
This is currently enough to be able to _release_ security fixes independently, but only if dependencies are explicitly recorded in lockfiles the... -
- vo.x (Vit Ondruch) wrote in #note-19:
> Unlikely? How about openssl in Ruby 3.1?
By "those parts of stdlib" I meant the 4 gems listed above.
io-wait & io-nonblock have < 5 methods each. Digest and strscan APIs are very stable and ne... -
- Eregon (Benoit Daloze) wrote in #note-14:
> vo.x (Vit Ondruch) wrote in #note-11:
> ...
Unlikely? How about openssl in Ruby 3.1?
> , and even if they do feature checking is always better to support more Ruby versions (e.g., `respond... -
- deivid (David Rodríguez) wrote in #note-15:
> Just to be clear, if in this particular case, for pragmatism, it is considered best that these dependencies should be removed, that's fine with me, but I don't think that warrants making this... -
- Also I recall issues with the `spring` gem that were fixed by correctly specifying their dependency on default gems like `tmpdir`. I could probably find more.
And again, I do understand the argument made that some stuff is not worth g... -
- As an example of 1), I run into trouble with bigdecimal 1.x vs 2.x due to missing proper dependency specifications.
Just to be clear, if in this particular case, for pragmatism, it is considered best that these dependencies should be ... -
- vo.x (Vit Ondruch) wrote in #note-11:
> 1) Not specifying dependencies prevents major changes. If there is default gem A of version 1.5.2 and there should be major update in future version of Ruby, lets say to the version 2.0.0. If some ... -
- I said it in other places too but It's also my opinion that dependencies on default gems should be explicit, for the reasons stated by others.
-
- * [Feature #18568] Explore lazy RubyGems boot to reduce need for --disable-gems
* With more and more default gems, RubyGems has become a major part of Ruby's startup time.
* On CRuby, the base boot time is reduced by 80% using `--dis... -
- gc_marks_continue will start sweeping when it finishes marking. However,
if the heap we are trying to allocate into is full, then the sweeping
may not yield any free slots. If we don't call gc_sweep_continue
immediate after this, then an... -
- Don't take my word as definitive, but I'd say it's expected. `Proc#<<` return has no "source code":
```ruby
>> proc{}.source_location
=> ["(irb)", 4]
> ...
=> nil
```
Think of it as a "delegator", it has no InstructionSequence ... -
10:46 AM Bug #18569 (Rejected): RubyVM::InstructionSequence#disasm returns nil for composed functions
- I was testing function composition Proc#>> (https://ruby-doc.org/core-2.6/Proc.html#method-i-3E-3E) and noticed that when disassembling composed functions it always returns nil.
``` ruby
first_proc = proc { |x| x + 2 }
puts RubyVM::... -
06:54 AM Feature #13110: Byte-based operations for String
- Personally I prefer byterindex because it's a C-like name (I guess String#rindex came from rindex(3)).
However, I follow Matz's decision.
-
02:21 AM Revision 8f3a36fb (git): Fix indents [ci skip]
02/02/2022
-
- Eregon (Benoit Daloze) wrote in #note-21:
> I think it should be `byte_index`, `byte_rindex` and `byte_offset`.
> ...
Yes, byterindex may be hard to read.
If we choose byte_*, it may be good to introduce aliases like byte_slice for ex... -
- I think it should be `byte_index`, `byte_rindex` and `byte_offset`.
I know there is `String#byteslice`/`getbyte`/`setbyte`, but `byterindex` starts to be really hard to read to the point it doesn't look like a proper Ruby method name to... -
- I've created a pull request which adds String#byteindex, String#byterindex, and MatchData#byteoffset:
https://github.com/ruby/ruby/pull/5518 - 11:05 PM Revision 6728f692 (git): * 2022-02-03 [ci skip]
-
11:05 PM Bug #18516: Memory leak on aliasing method to itself
- ruby_3_1 42c9ef769f210d88241a114395dd5ffc27b2fb87 merged revision(s) 7ff1bf317887c0d7b21e91ad548d07b9f05c540c,e89d80702bd98a8276243a7fcaa2a158b3bfb659.
-
- An alias can suppress method redefinition warning
---
test/ruby/test_alias.rb | 11 +++++++++++
1 file changed, 11 insertions(+)
Fix memory leak at the same named alias [Bug #18516]
When aliasi... -
10:07 PM Bug #18567: Depending on default gems in stdlib gems when not needed considered harmful
- I agree that io-wait at a minimum should not be gemified. Perhaps digest. I think this is a different argument than the overall issue you raised.
I disagree that gems made default should not listed. My main scenario is you have a... -
- Eregon (Benoit Daloze) wrote in #note-9:
> These gems are default gems and so available by default, and for the foreseeable future they will always be shipped with Ruby, so there is no risk to not being super explicit about these depend... -
- There are stdlib gems of dubious value, like io-wait or digest, but I'm not sure where to draw the line. My rule of thumb would be that if a stdlib exists as a gem, other gems that need it should explicitly depend upon it, because that's...
-
- These gems are default gems and so available by default, and for the foreseeable future they will always be shipped with Ruby, so there is no risk to not being super explicit about these dependencies (and most code out there using them d...
-
- While I understand the issue, I believe the correct approach is to actually specify the dependencies and specify them always and everywhere. Yes, that will break something somewhere, but once everything is gemifie and all dependencies ar...
-
- Another issue with upstreaming code to default gems is some default gems need many internals and those should not be gems, they should be core or stdlib.
For TruffleRuby for instance, I don't want to expose many internals to support `st... -
- Here is a set of PR to solve the issue for JRuby and TruffleRuby for net-* gems:
https://github.com/ruby/net-protocol/pull/6
https://github.com/ruby/net-imap/pull/56
https://github.com/ruby/net-pop/pull/10
https://github.com/ruby/net... -
- Here is a practical example of needless dependencies on default gems causing JRuby to not be able to install e.g. `net-imap`, even though JRuby has the needed stdlibs:
https://gist.github.com/eregon/5ea86a9625fc32844c81c5f1448d02d0 -
- CRuby over time moves more and more code to default gems, to bundled gems and or even stops shipping some gems which used to be stdlibs (#5481).
I believe the main motivation for that is being to fix security issues without needing to m... -
08:10 PM Revision 7529c538 (git): [ruby/net-http] Do not set SNI hostname if connecting to IP address
- RFC 6066, section 3, explicitly disallows the use of an IP address
as an SNI server name. So check if the connection is being made
to an IP address using the resolv regexps, and do not set an SNI
hostname in that case.
Recent changes t... -
- @eregon I remember seeing code in TruffleRuby that built up an index of default gem files, so it would know which ones might need to trigger a gem load. I cannot find that now.
Could you provide some links to show how this is handled,... -
- Another comparison: in JRuby, RubyGems booting accounts for around 1/3 of JRuby's startup time in our "fastest" mode. Given how poor our startup time is already, we are understandably interested in ways to speed it up.
-
- In https://bugs.ruby-lang.org/issues/17684 there was debate about whether the `--disable-gems` flag should be removed. Several folks were in favor, since Ruby without RubyGems is fairly limited, but others wanted to keep the flag for sma...
-
- And for macOS: try to not mess up your system by installing random stuff in /usr/local, that's the source of many issues reported to `ruby-build`.
Using Homebrew is fine (or MacPorts), but manually installing stuff in /usr/local is aski... -
- After 4 years, We should remove the bundled source code like zlib, libyaml, libffi.
for mswin users: use vcpkg, other platforms: use the official package manager like apt, dnf. - 05:21 PM Revision 3ca244a2 (git): * 2022-02-03 [ci skip]
-
05:20 PM Revision 2913a2f5 (git): Treat TS_ICVARC cache as separate from TS_IVC cache
-
- "Dead-on-arrival" hyberbole aside, it does seem that using `+` semantics would allow Range#step to work "for free" with many existing classes.
More importantly, I realized there is no need to introduce any backward incompatibility: if b... -
- Add a new macro BASE_SLOT_SIZE that determines the slot size.
For Variable Width Allocation (compiled with USE_RVARGC=1), all slot
sizes are powers-of-2 multiples of BASE_SLOT_SIZE.
For USE_RVARGC=0, BASE_SLOT_SIZE is set to sizeof(RVA... -
- My concern was that Sentry will now use #detailed_message instead of #message, but it will now include the exception class, unlike before. But if Sentry is ok with that there's no problem.
-
- @Dan0042 I think your last comment is a different topic. I'd like to keep one proposal as small as I can.
-
01:52 PM Revision af107710 (git): Add `econv_get_encoding` function
-
- I agree this would make a lot of sense, also for `io-nonblock` which is also just a couple IO methods.
-
08:55 AM Feature #18566: Merge `io-wait` and `io-nonblock` gems into core IO
- It makes sense to me.
-
08:52 AM Feature #18566: Merge `io-wait` and `io-nonblock` gems into core IO
- I am happy to do the initial implementation.
-
- I think we should reconsider status of `io-wait`, and consider simply merging it into core's `IO`.
According to @nobu it's only a gem for "historical" reasons.
Any non trivial IO code will likely make use of it, and it's just 400 l... -
12:21 PM Revision d8c54bac (git): [ruby/net-protocol] Fix typo in gem description [ci skip]
- https://github.com/ruby/net-protocol/commit/35d7b08a54
-
- This gem exposes no executables.
https://github.com/ruby/net-protocol/commit/3c4def2a64 - 08:53 AM Revision 7bef8ed1 (git): Update bundled gems list at 2022-02-02
-
- - [Feature #18498] Introduce a public `WeakKeysMap` that compares by equality (byroot)
- Ruby's current weak reference constructs are not very usable.
- The most useful weak collection that is offered by most platform is `WeakKeysMap... -
- * [Feature #18513] Hide patchlevel from `ruby -v`
* There is no reason to display it now.
-
- Merge RubyGems 3.3.x and Bundler 2.3.x
- 04:30 AM Revision 69463805 (git): * 2022-02-02 [ci skip]
-
- It is expected to use on master branch
- 04:26 AM Revision ea6b6f1a (git): Update bundled gems list at d4a8c04dc7bb1c9aa48eaf50dc6b18 [ci skip]
-
04:25 AM Revision d4a8c04d (git): Update RBS (#5515)
- * Bundle RBS 2.1.0
-
- I don't think this is a bug. `strptime` is supposed to be the inverse of `strftime`:
```ruby
DateTime.now.strftime('%I%M%p')
# => "0617PM"
DateTime.now.strftime('%l%M%p')
# => " 617PM"
```
If you pass the same formats to `st... -
02:06 AM Bug #18565 (Rejected): strptime reports invalid date when string is "130AM" and format string is '%I:%M%p'
- `DateTime.strptime("1:30PM", '%I:%M%p')` and `DateTime.strptime("1:30PM", '%l:%M%p')` both return
`#<DateTime: 2022-02-01T13:30:00+00:00 ((2459612j,48600s,0n),+0s,2299161j)>`
`DateTime.strptime("130PM", '%I%M%p')` and `DateTime.strpt... - 12:08 AM Revision bb845fb9 (git): * 2022-02-02 [ci skip]
-
12:08 AM Revision aff4b219 (git): Fix a typo [ci skip]
02/01/2022
-
11:11 PM Bug #18436: Fix Pathname dot directory globbing
- I noticed this change just now. We have some programs which depended on the previous behavior. It loses backward compatibility, so I think it should have been noted in NEWS.md in Ruby 3.1.0. I understand it's too late to say that, though.
-
- FWIW, TruffleRuby has `Primitive.method_unimplement` specifically to achieve to achieve the same as `rb_f_notimplement` but in Ruby code.
Example usage: https://github.com/oracle/truffleruby/blob/4eddae9d18137567beb25e6e4ef2c932fa9de7ab... -
- Thx for the clarification. For now, I have reported the original issue to the PG upstream [1]. Keeping this open should there be a chance to follow the `fork` example.
[1] https://github.com/ged/ruby-pg/issues/423 -
- - [Feature #18249] The ABI version of dev builds of CRuby does not correspond to the ABI
- ABI checking for development Ruby.
- Calculates the MD5 hash of header files during compilation of Ruby, stores this MD5 in native exten... -
- Hello, I’d like to summarize the motivation for this feature and the implementation I proposed.
# Motivation for this feature
At Shopify, we run Ruby head on CI daily against several services. We also run Ruby head on a small porti... -
- (I have added to the description an url to matz's original statement)
-
07:44 PM Feature #18563 (Closed): Add "graphemes" and "each_grapheme" aliases
- https://bugs.ruby-lang.org/issues/13780#note-10
> grapheme sounds like an element in the grapheme cluster. How about each_grapheme_cluster?
> ...
Languages that have added grapheme cluster support seem to be almost exclusively opting... -
- I have created another ticket to summarize the final spec of this proposal: #18564.
-
- (This ticket is for recording the final spec of #18438)
## Proposal
I would introduce a method `Exception#detailed_message`, and let the default error printer use it instead of `Exception#message` to create a error output.
```
... -
- Eregon (Benoit Daloze) wrote in #note-38:
> Note: Whenever a gem is extracted from stdlib, we should be careful the gem can be installed on older Ruby versions, otherwise it's not possible for a library to depend on this newer gem via R... -
- Note: Whenever a gem is extracted from stdlib, we should be careful the gem can be installed on older Ruby versions, otherwise it's not possible for a library to depend on this newer gem via RubyGems (and it has to, otherwise `require` f...
-
- ruby_3_1 cd4aae430da8303dac9f1be2d1e1b7ce1bc8a458 merged revision(s) f27eb8148f5a72bbacfebfecc7de9305471bb5c9,d75f7078c831d45ab5ba2fae4fbb308d01a21e46,d5836db1b398a7936b0461b3011db66f6cc8c490.
-
- Sync io-console gem.
---
ext/io/console/console.c | 8 ++++++--
test/io/console/test_io_console.rb | 8 ++++++++
2 files changed, 14 insertions(+), 2 deletions(-)
[ruby/io-console] bum... -
- This reverts commit aa3b2c81c8de2e70bd100e92436f15c03523eaf7.
-
- ruby_3_1 1f17463dec2ab1535096cb56005a840eae4fea5b merged revision(s) 6a6227e0168b059c3ed34c9f0ace2e5dc2364221.
-
- Shifting zero always results in zero [Bug #18517]
---
numeric.c | 2 ++
1 file changed, 2 insertions(+) -
12:28 PM Misc #16512: Improving `www.ruby-lang.org` reference by merging with `rubyreferences.github.io`
- @hsbt I am up to the task!
There would be a few questions of what's the best way of handling it, I'll prepare the plan/list this week. Basically, there should be made a decision on the process of synchronization of the docs, and updat... -
- @zverok Can you handle this? I'll support to migrate your documentation to www.ruby-lang.org or docs.ruby-lang.org.
- 11:07 AM Revision 0b2f6b94 (git): [rubygems/rubygems] Skip "seller shipped" notification after delivery
- If a Shipment has been delivered, there is no point in notifying the
buyer that the seller shipped. Instead, we should simply notify the
buyer that the shipment was delivered. This is relevant in cases where
the seller is late to mark a ... -
- https://github.com/ruby/io-wait/commit/7be58cd7f4
-
05:38 AM Feature #18462: Proposal to merge WASI based WebAssembly support
- mame (Yusuke Endoh) wrote in #note-4:
> Feel free to ask me if you have a question.
Wow, thanks for putting this together! I did in fact have a question, but I figure we can move the conversation somewhere else to keep this Ruby tick... -
- Hi @jez
Here you are: https://github.com/mame/emruby/
In short, please setup emsdk and then try `./configure --host wasm32-unknown-emscripten --with-static-linked-ext --with-ext=(an extension names that you want to statically link)... -
- Ref: https://github.com/mikel/mail/pull/1439
Some gems depend on io-wait, but still support older rubies,
so they have to chose between droping support or not listing io-wait.
But io-wait could act a a noop on older rubies.
https://gi... - 04:22 AM Revision 804dca7b (git): Update default gems list at fc4fbeef28b1d243e45d7fc8d60778 [ci skip]
-
- https://github.com/ruby/io-wait/commit/e15a2486b2
-
-
- headius (Charles Nutter) wrote in #note-25:
> Is there a way to prohibit the `to_enum` form? I don't think it will be useful to support since it may vary greatly across implementations (depending on what the stack looks like and what ge... -
- Is there a way to prohibit the `to_enum` form? I don't think it will be useful to support since it may vary greatly across implementations (depending on what the stack looks like and what gets included).
It does raise a question for m...
01/31/2022
-
-
- Currently some specs are broken because `rspec-mocks-3.10.3` is used,
which has some breaking changes, apparently.
This change makes ruby-core install the same gems installed upstream for
running bundle specs, so that things never break... -
-
08:45 PM Bug #18562 (Closed): throw_data passed to rescue through require
- When we throw in a require it's incorrectly considered as a raise, making the `T_IMEMO` `throw_data` object accessible via rescue (in a contrived case). In regular usage this is unlikely to cause an issue because `Module#===` immediately...
-
- Related JRuby pull request that makes singleton method definition always public (in response to the ostruct issue linked above):
https://github.com/jruby/jruby/pull/7055 -
- @marcandre provided a possibly more common case, but I still highly doubt that anyone would expect this to behave differently than the unmatched scope version:
```
$ ruby -e 'class Foo; class << Foo; private; Foo.define_singleton_met... -
- Currently nearly all uses of `define_singleton_method` or `def obj.foo` will ignore the caller's frame/cref visibility and use the default visibility of public. I propose this be made explicit in the code, documentation, and ruby specs.
... - 08:20 PM Revision 2b2e3b79 (git): * 2022-02-01 [ci skip]
-
- - it is not used since it is not at the top of the file
- it is not useful anymore
https://github.com/rubygems/rubygems/commit/6aee05d923 -
- katei (Yuta Saito) wrote:
> For example, CRuby already supports WebAssembly target by Emscripten, but Emscripten heavily depends on JavaScript to emulate some missing features in WebAssembly itself.
@katei do you have a link showing ... -
- @duerst Thank you for spelling that out. I figured there are some additional nuances to this, since the unicode tables are also in C code and used internally by many parts of Ruby. In that regard, it is at least much easier for us to imp...
-
- @peterzhu2118 JFYI: If the method body is `rb_f_notimplement`, `Kernel#respond_to?` returns false.
For example, `Kernel#fork` is defined as such in Windows. https://github.com/ruby/ruby/blob/605f226142a47c4f9e4cc88717c5d6170bc3d7f1/pr... -
- I think this is expected. `GC.compact`, `GC.auto_compact=`, `GC.verify_compaction_references` will raise `NotImplementedError` when compaction is not supported on the platform.
`GC.respond_to?(:compact)` is just to check that the meth... -
- Trying to execute test suite of `pg` gem on ppc64le, I observe a following error:
~~~
... snip ...
1) GC.compact should compact PG::TypeMapByClass #328
Failure/Error: GC.verify_compaction_references(toward: :empty, double_... -
- Would it be feasible for #full_message to accept a String/IO argument to which it writes the parts of the message? So the error printer could call `err.full_message(output_to: $stderr)` in order to avoid buffering.
-
- I have given it a try, but I faced two non-trivial test failures.
One is TestException#test_cause_exception_in_cause_message.
https://github.com/ruby/ruby/blob/b01657c4707eadd9de9573ce9818d0b5f0fe3047/test/ruby/test_exception.rb#L9... -
- Thank you, it's much clearer now.
Follow-up question: Should #detailed_message include the class of the error
a) always (unlike #message currently)
b) only if `highlight` keyword is true
c) based on some other keyword (like in #not... -
- I talked with @matz and @nobu about that.
The current implementation of `Exception#full_message` adds `\e[1m` (i.e., bold start) for each beginning of line of `#message`, and `e[m` (i.e., bold end) for each end. (Also, it adds `" (Run... -
02:47 PM Bug #17263: Fiber context switch degrades with number of fibers, limit on number of fibers
- I heard someone ran into this error in a GraphQL-Ruby context, so I thought I'd check out this script on the latest Ruby. It didn't crash as-written, so I added a couple more orders of magnitude. It still finished fine locally, but slowe...
-
- The for loops are not correctly iterating heap pages in
gc_verify_heap_page. -
- Marking this as a feature, because I think it should be improved but can hardly be considered a bug.
### Repro
Consider the following script:
```ruby
# /tmp/allocation-source.rb
require 'objspace'
require 'tmpdir'
source =... -
- ruby_3_1 1c426ddb25e2ccb169720e4f9894605596fe594d merged revision(s) 5c7af72304d0ad33cd3f21b24a4bc44e8acd5b2c,d650b17686d49c2ce8e6a87039861154e93d4621.
-
- Assuming EXIT_SUCCESS equals 0 is not portable
---
test/ruby/test_fiber.rb | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
`rb_fiber_terminate` must not return [Bug #18497]
In a for... -
- ruby_3_1 629b9da7dbdcaff04d6dde70f2431396f66e03e5 merged revision(s) f3c77bd480834f2835fe6fef5c0475336248dbde.
-
- Fix the placeholder subclass entry skipping [Bug #18489]
---
class.c | 3 +--
test/ruby/test_module.rb | 9 +++++++++
2 files changed, 10 insertions(+), 2 deletions(-) -
09:05 AM Bug #18294: error when parsing regexp comment
- I understand it is not easy to fix, and I sure can live with it.
ありがとうMatzさん \^_^ -
02:28 AM Bug #18294: error when parsing regexp comment
- I admit this is a bug and it should be fixed. But implementation-wise, it's difficult to fix. Considering the (small) impact of the bug, its priority is low.
We will fix it but it could take a long time.
Matz.
- 06:35 AM Revision 5b0c7b4a (git): Prefer `wait_readable` rather than `IO.select`.
- 06:35 AM Revision bdd3cf00 (git): Fix incorrect minimum read length.
- 06:35 AM Revision aa3b2c81 (git): Fix console.c usage of `rb_io_wait`.
-
05:02 AM Bug #11064: #singleton_methods for objects with special singleton_class returns an empty array
- matz (Yukihiro Matsumoto) wrote in #note-1:
> `nil` does not have its singleton class, but `NilClass` plays the role of the singleton class
I would appreciate it if @matz could kindly explain what the difference is between "being a s... -
- `nil` does not have its singleton class, but `NilClass` plays the role of the singleton class (since `nil` is the only instance of `NilClass`).
So a singleton method definition defines a method in `NilClass`. It's a kind of special trea... -
- For the record: we confirmed that this issue was fixed by `gem install ffi -- --enable-libffi-alloc`. See #18555 and https://github.com/ffi/ffi/issues/800#issuecomment-1008308614
-
- I believe this is the same case as #18555. We confirmed that that issue was fixed by `gem install ffi -- --enable-libffi-alloc`. See #18555 and https://github.com/ffi/ffi/issues/800#issuecomment-1008308614
-
- I believe this is the same case as #18555. We confirmed that that issue was fixed by `gem install ffi -- --enable-libffi-alloc`. See #18555 and https://github.com/ffi/ffi/issues/800#issuecomment-1008308614
-
- I believe this is the same case as #18555. We confirmed that that issue was fixed by `gem install ffi -- --enable-libffi-alloc`. See #18555 and https://github.com/ffi/ffi/issues/800#issuecomment-1008308614
-
02:23 AM Feature #18515: Add Range#reverse_each implementation
- Dan0042 (Daniel DeLorme) wrote in #note-3:
> I agree with both of those statements, but Integer is the only class that has #pred, so in practice "using pred" and "specialized implementation for Integer" are the same thing.
Yes, that ... -
- ruby_3_1 8e3bce0841497fc00ccdaf45f6583c25720a2d3f merged revision(s) ae5458f228a5477383e9c00425d85d50a3867817.
-
- thread.c: Convert TAG_BREAK to a normal exception at thread top-level
[Bug #18475]
---
test/ruby/test_enum.rb | 11 +++++++++++
thread.c | 3 +++
2 files changed, 14 insertions(+) -
- ruby_3_1 abdd655afd6f3f36553e6bb0d6b3e3eac09ae977 merged revision(s) 92630e2346a221fdfeaa484632938a06765bc6e5.
-
- [ruby/optparse] Fix links to the page directory files [Bug #18468]
https://github.com/ruby/optparse/commit/dab72c543d
---
doc/optparse/creates_option.rdoc | 4 ++--
doc/optparse/option_params.rdoc | 2 +... -
- @anthonyaykut Thank you for your cooperation! Because there are many bug reports about the issue of ffi with M1, this will be a great solution case.
-
-
- * [Feature #13110] Byte-based operations for String
* Are there any objections to introduce byteindex, byterindex and byteoffset? (not bytesplice)
* Should the method names be `byte_index`, `byte_rindex`, and `byte_offset`?
-
- https://github.com/ruby/ruby/pull/5497.diff
01/30/2022
-
- kyanagi (Kouhei Yanagita) wrote in #note-2:
> In general, I think it would be better to improve `#reverse_each` by using `#pred` (predecessor; the inverse of `#succ`) if the element has.
> ...
I agree with both of those statements, but... -
- Very happy about this.
mame (Yusuke Endoh) wrote in #note-16:
> Note: the rdoc of `detailed_message` should explicitly state that the method should accept `highlight` keyword that allows using ANSI terminal escape sequences to highli... - 10:59 PM Revision b01657c4 (git): [DOC] Fix case for new sentence.
-
10:38 PM Revision f924e05b (git): [DOC] Fix a typo in the NEWS.md
- Follow up of https://github.com/ruby/ruby/commit/fbb3cab.
-
- Treats:
Kernel#readline
Kernel#readlines
Kernel#`cmd`
IO#advise
IO.select
IO#ioctl
IO#fcntl
Kernel#syscall
IO.pipe -
- > the implication is that `range.step(1)` (using `+`) would have different semantics than `range.each` (using succ); I have reservations about that.
Well, it is already so to some extent. Say, with numeric ranges `#step` returns `Arit... -
- @zverok if understand correctly, the implication is that `range.step(1)` (using `+`) would have different semantics than `range.each` (using `succ`); I have reservations about that.
Also, due to backward compatibility I don't think it... -
- Some clarifications after rereading the corresponding [dev.meeting log](https://github.com/ruby/dev-meeting-log/blob/master/DevMeeting-2022-01-13.md#feature-18368-rangestep-semantics-for-non-numeric-ranges-zverok):
**My proposal is no... -
- > Here's one idea: how about adding an #increment(n) method. For Numeric, Float, Time, Date, it would be an alias to +. For String it would be equivalent to doing succ n times. So Range#step would have the semantics of using begin.increm...
-
- > matz: I'd like to allow numeric-type '+', but to deny concatination-type '+'
> ...
I fully agree with the above. Here's one idea: how about adding an `#increment(n)` method. For Numeric, Float, Time, Date, it would be an alias to `+`.... -
09:48 PM Feature #16122: Data: simple immutable value object
- I'm quite fond of this proposal--I basically never use Struct unless I specifically need mutability and have been using the values gem for years, which has a simple implementation of about 100 lines:
https://github.com/tcrayford/Value... -
- Eregon (Benoit Daloze) wrote in #note-29:
> ValueStruct or ImmutableStruct or FrozenStruct maybe?
Those are good ideas. Or to highlight the "pared-down" aspect of this strict version of Struct: SimpleStruct / PlainStruct / BasicStruc... -
08:39 PM Feature #16122: Data: simple immutable value object
- Some more alternatives to get the ideas rolling: `Unit` and `Item` (might be paired with Struct)
I also like `Box`. -
- ValueStruct or ImmutableStruct or FrozenStruct maybe?
ImmutableStruct would probably only make sense if values are made immutable too, which doesn't seem proposed here.
I think the nesting of `Struct::Value` feels a bit weird, especi... - 06:47 PM Revision b0bd26ad (git): * 2022-01-31 [ci skip]
-
06:47 PM Revision e75bd8df (git): pend some tests while using to_binary
- ISeqs from ISeq#to_binary/load do not support `keep_script_lines`
now so some tests should be pending tests with
`RUBY_ISEQ_DUMP_DEBUG=to_binary` -
- I think Ruby users already know well what `#<MyObject:0xXXXXXXXXXXXXXXXX>` means, and using another naming is likely more confusing.
Either way it will need some getting used to as it's new output, and then it will just be natural.
The... -
- "undefined method 'time' for object Integer" is grammatically incorrect.
It would need to be "undefined method 'time' for Integer object".
Or "instance of Integer" or "object of type Integer". -
- Great!
@mame Do you plan to make a PR or should I give it a try? -
- ruby_3_1 28ec24db4a8201fa624692cdc1cbef9b00489412 merged revision(s) a79c59472df38297c246b27713c277f2edaefa7a.
-
- Allow include before calling Module#initialize
This is to allow Module subclasses that include modules before
calling super in the subclass's initialize.
Remove rb_module_check_initializable from Module#initiali... -
- ruby_3_1 1a1737530bff683041b8f45e7acac8c5c85c4163 merged revision(s) c764e368bd9c0487e6cc97f8a62e0e0e99a0d7c5.
-
- Fix default --jit-max-cache in `ruby --help`
changed at 028f1887c2bfa50dcfc1b623470d6ab1b880fdde
---
ruby.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-) -
- ruby_3_1 e75246c287436621658bc2b8a5f999d8991e7d1d merged revision(s) 5917f49ad03cd3b2c13d3eea85c969a113b83bf8.
-
- Fix documentation for Dir.glob
Dir.glob("*", File::FNM_DOTMATCH) no longer includes ".." in the
returned array as of Ruby 3.1.
Fixes [Bug #18436]
---
dir.rb | 2 +-
1 file changed, 1 ins... -
- ruby_3_1 2640161df5cf18d08ec86a0c1b913d4ee99e102a merged revision(s) 5414de4b6e4372af832e338f8eb7a9fe8de17c84.
-
- ruby_3_1 20091ccad34904cb5ded13a8787f6662a8e2df68 merged revision(s) 0dc7816c4350683ccd020f5759eee4914de0085d.
-
- ruby_3_1 d3a2a3c5420fecfee54cbf3fbdd9287bc6c58f03 merged revision(s) 9e0a91d0640600f2dfd7fc1d5fae6667019c9ca5.
-
- Don't segfault if Warning.warn is undefined
Check that there is a method entry for the method before passing
it to rb_method_entry_arity.
Fixes [Bug #18458]
---
error.c | 3 +... -
- Yes, see [Feature #16432].
-
- YJIT: Fix SP index with optarg and unordered kwarg
Previously when we were calling a method with an optional argument and
multiple keywords arguments which weren't in the order the receiver
expected we would use ... -
- Make RubyVM::AST.of work with code written in `-e` command-line
option
[Bug #18434]
---
ast.c | 11 +++++++----
test/ruby/test_ast.rb | 5 +++++
2 files changed, 12 inse... -
06:42 AM Bug #18443: IO read/write/wait hook bug fixes.
- I have confirmed the PR https://github.com/ruby/ruby/pull/5473 contains the two whole backports and a partial backport from master branch.
I think e4cd319f3d6560b76dbc1c7e89aa4d790a2d1e43 should be backported. @naruse can I backport i... -
01/29/2022
-
08:34 PM Bug #18555: Running "bundle exec middleman server" on M1 Mac gives [BUG] Bus Error at 0x0000000104b04000
- Right, I have an update -- I worked out that `syntax` was also a gem :) so in the end I added:
```
gem "middleman-syntax"
gem "middleman-sprockets"
gem "middleman-autoprefixer"
gem "nokogiri"
```
to get everything working! It'... -
- mame (Yusuke Endoh) wrote in #note-9:
> anthonyaykut (Anthony Aykut) wrote in #note-8:
> ...
OK, I did as suggested... `bundle install` took a while to complete, but looked good. I tried to run `bundle exec middleman server`, but it co... -
- anthonyaykut (Anthony Aykut) wrote in #note-8:
> and then... `bundle install`:
> ...
...
> Using ffi 1.15.5
...
> ...
I think the command did not recompile ffi gem. Could you delete already-installed gems and retry `bundle install`?... -
07:30 PM Bug #17629: macOS, readline compiled with libedit: adding entries to history doesn't work unless Readline.readline is called before populating it
- Can someone close this ticket? I don't see a possibility to do it myself.
-
- ```ruby
def yield_2_args
yield 1, 2
end
yield_2_args { p [eval("_1"), _2] }
```
Running with ruby 2.7.2, I get output:
```
[1, 2]
[1, 2]
```
With 3.0.2, I get output:
```
[1, 2]
(eval):1:in `block in <main>': unde... -
- Something seems changed on FreeBSD 13.
-
- Define `MAP_ANONYMOUS` to `MAP_ANON` if undefined on old systems.
-
- # The next dev meeting
**Date: 2022/02/17 13:00-17:00** (JST)
Log: https://github.com/ruby/dev-meeting-log/blob/master/DevMeeting-2022-02-17.md
- Dev meeting *IS NOT* a decision-making place. All decisions should be done at the bu... -
-
- Discussed on the dev-meeting.
#18367 was accepted, so @matz said "give it a try". Need to create a PR. -
08:32 AM Revision fbb3cab9 (git): Add a NEWS entry about [Feature #16806]
-
- Discussed on the dev-meeting.
@matz is now negative to allow settings. Having various semantics in one Struct class will bring confusion rather than usability. `keyword_init` settings will be no longer needed after Ruby 3.2. (See #168... -
- @nobu already committed the last step at commit:c956f979e5d05900315d2753d5c3b1389af8dae4. Closing.
-
- @matz accepted the name `detailed_message`. I'll update and improve my PR (or volunteer is welcome)
Note: the rdoc of `detailed_message` should explicitly state that the method should accept `highlight` keyword that allows using ANSI ... -
- Discussed at the dev-meeting but no conclusion.
@akr said the part `0xXXXXXXXXXXXXXXXX` is almost useless, and roughly counter-proposed the following.
```
42.time #=> undefined method `time' for object Integer (NoMethodError)
... -
- https://github.com/ruby/ruby/pull/5506
-
03:08 AM Bug #18556 (Closed): error: use of undeclared identifier 'MAP_ANONYMOUS'
- ruby 3.1.0 fails to build on older versions of macOS such as OS X 10.10.5:
```
gc.c:11141:88: error: use of undeclared identifier 'MAP_ANONYMOUS'
char *ptr = mmap(NULL, alignment + size, PROT_READ | PROT_WRITE, MAP_PRIVATE |... -
- I'm unsure but you failed to recompile ffi gem because the size of ffi_c.bundle did not change.
Before:
```
{
"source" : "P",
"arch" : "arm64",
"base" : 4342497280,
"size" : 98304,
"uuid" : "f086814f-a607-31b9-9fff-7037c97cdf32... -
12:08 AM Bug #18512: MacOS 12.1 Monterey Bug
- IPS Report:
```
oucl5976@ck1 cf-zendesk % cat ~/Library/Logs/DiagnosticReports/ruby-2022-01-29-000341.ips
{"app_name":"ruby","timestamp":"2022-01-29 00:03:41.00 +0000","app_version":"","slice_uuid":"0ef6bea2-99c6-3fd9-a082-80849cd76... -
- ```
oucl5976@ck1 ~ % gem uninstall ffi
You have requested to uninstall the gem:
ffi-1.15.5
rb-inotify-0.9.10 depends on ffi (>= 0.5.0, < 2)
sassc-2.4.0 depends on ffi (~... -
-
-
06:22 AM Revision b90e56e6 (git): mkmf: pkg_config accepts multiple options
-
-
06:08 AM Bug #18514: ruby 2.6.9's Date#jisx0301 does not support Reiwa
- > Can you confirm with date-2.0.3 ?
Thank you so much.
I tried it, then I get good result!
-
- Accepted.
Matz.