Misc #18039: DevelopersMeeting20210819Japan

• Status: Closed
• Assignee: -

[mame (Yusuke Endoh)]

The next dev meeting

Date: 2021/08/19 13:00-17:00
Place/Sign-up/Agenda/Log: https://github.com/ruby/dev-meeting-log/blob/master/DevelopersMeeting20210819Japan.md

• Dev meeting IS NOT a decision-making place. All decisions should be done at the bug tracker.
• Dev meeting is a place we can ask Matz, nobu, nurse and other developers directly.
• Matz is a very busy person. Take this opportunity to ask him. If you can not attend, other attendees can ask instead of you (if attendees can understand your issue).
• We will write a log about the discussion to a file or to each ticket in English.
• All activities are best-effort (keep in mind that most of us are volunteer developers).
• The date, time and place are scheduled according to when/where we can reserve Matz's time.
• DO NOT discuss then on this ticket, please.

Call for agenda items

If you have a ticket that you want matz and committers to discuss, please post it into this ticket in the following format:

* [Ticket ref] Ticket title (your name)
  * Comment (A summary of the ticket, why you put this ticket here, what point should be discussed, etc.)

Example:

* [Feature #14609] `Kernel#p` without args shows the receiver (ko1)
  * I feel this feature is very useful and some people say :+1: so let discuss this feature.

• It is recommended to add a comment by 2021/08/16. We hold a preparatory meeting to create an agenda a few days before the dev-meeting.
• The format is strict. We'll use this script to automatically create an markdown-style agenda. We may ignore a comment that does not follow the format.
• Your comment is mandatory. We cannot read all discussion of the ticket in a limited time. We appreciate it if you could write a short summary and update from a previous discussion.

[mame (Yusuke Endoh)]

• Related to Misc #14770: [META] DevelopersMeeting added

[nobu (Nobuyoshi Nakada)]

• [Misc #18059] Which FL_USERx are open to extension libraries? (nobu)
  • (need comment)

[nobu (Nobuyoshi Nakada)]

• [Feature #18051] Move symbols exported under internal (nobu)

  • Some symbols exported in headers under internal are declared and used in some extension libraries.
    These seem OK to be moved to public headers.

    • ruby_scan_digits, ruby_hexdigits, ruby_digit36_to_number_table
    • rb_int_positive_pow
    • rb_deprecate_constant

[alanwu (Alan Wu)]

• [Bug #13392] TracePoint return event location is incorrect for methods defined with define_method (alanwu)
  • There are two PRs available, one more complex but opens the possibility for future optimizations. Which do you prefer?
• [Bug #18060] Fix infinite loop when b_return TracePoint throws (alanwu)
  • Does the fix make sense?
• [Bug #18031] Nested TracePoint#enable with define_method target crashes (alanwu)
  • Does my fix look good? It assumes that multiple TracePoints targeting the same bmethod should work similar to other targets.

[duerst (Martin Dürst)]

• [Feature #18045] Variable Width Allocation Phase II (peterzhu2118)
  • Variable Width Allocation is a major change to Ruby internals and therefore should be discussed seriously.

[byroot (Jean Boussier)]

• [Feature #10917] Add GC.stat[:total_time] (byroot)
  • Time spent in GC is a very important metric for application performance monitoring.
  • Right now only GC::Profiler can be used, but it is ill suited when there are multiple callers because it is stateful. It also need to be gated by a Mutex.
  • A single monotonically increasing time metric would be much more usable.
  • If the getrusage_time overhead is dimmed to big, it could be a an opt-in metrics through a flag on ::GC.
  • I don't ask to merge that particular patch, but wether the general idea is acceptable.
  • @Eregon (Benoit Daloze) suggested GC.stat[:time] in milliseconds, as that's what JRuby and TruffleRuby already expose.

[jeremyevans0 (Jeremy Evans)]

• [Bug #18036] Pthread fibers become invalid on fork - different from normal fibers. (jeremyevans0)
  • fork terminates other threads.
  • When using pthread coroutine, all fibers are threads, therefore fork terminates all other fibers.
  • Are we OK with this runtime difference with pthread coroutine? (I am.)
  • Do we want to terminate other fibers when forking always, or just when using the pthread coroutine?
  • Do we want to raise an exception if calling fork from a non-main fiber?
• [Feature #16889] TracePoint.enable { ... } also activates the TracePoint for other threads, even outside the block (jeremyevans0)
  • I think this behavior is currently expected. At least the tests expect it, and the behavior seems reasonable.
  • Ruby could benefit from a nicer way to pass a block and only allow tracing for the block.
  • I implemented this nicer way by supporting a target: :block option.
  • Is the current behavior of TracePoint.enable with a block expected?
  • Is it OK to add support for a target: :block option using the patch?
• [Bug #15404] Endless range has inconsistent chaining behaviour (jeremyevans0)
  • Do we want to make ranges of ranges a parse error (e.g. (1..2)..(1..3))?
  • I think we shouldn't. It's possible to define Range#<=> and Range#succ such that a range of ranges could be useful.
• [Bug #15428] Refactor Proc#>> and #<< (jeremyevans0)
  • Do we want Proc#<< and Proc#>> to call to_proc on the argument given?
  • Doing so would allow usage of symbols and other things that implement to_proc but not call.
  • Doing so always would break usage with arguments that respond to call but do not respond to to_proc.
  • Doing so only if the argument responds to to_proc and not call may result in inconsistent behavior.
  • I see no problem with the current situation, but I'm not opposed to calling to_proc if the object does not respond to call.
• [Bug #14744] Refinements modules have a superclass (jeremyevans0)
  • Should we hide the superclass of a refinement module from reflection APIs such as ancestors?

[ktsj (Kazuki Tsujimoto)]

• [Feature #16182] Should expr in a, b, c be allowed or not? (ktsj)
  • How about allowing brackets/braces to be omitted in one-line pattern matching?

[tenderlovemaking (Aaron Patterson)]

• [Feature #18007] C extension "allocation function" check
  • Should we add a config option and put in 3.1, or just wait for 3.2?

[ioquatix (Samuel Williams)]

• [Feature #18083] Capture error in ensure block.
  • Is it acceptable?

[nobu (Nobuyoshi Nakada)]

• [Bug #18077] Marshal.dump(closed_io) raises IOError instead of TypeError (nobu)
  • Marshal.dump is expected to raise a TypeError for unmarshallable objects. But closed streams raise an IOError

[mame (Yusuke Endoh)]

• Description updated (diff)

[hsbt (Hiroshi SHIBATA)]

• Status changed from Open to Closed