Misc #19525: DevMeeting-2023-04-13

• Status: Closed
• Assignee: -

[mame (Yusuke Endoh)]

The next dev meeting

Date: 2023/04/13 13:00-17:00 (JST)
Log: https://github.com/ruby/dev-meeting-log/blob/master/2023/DevMeeting-2023-04-11.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 record of the discussion in the 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 of the meeting 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 2023/04/10. 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

[hsbt (Hiroshi SHIBATA)]

• [Bug #19551] Backport commits for CI failures
  • Can I merge PR like "Fix CI failure" for stable branches myself from April 2023?

[byroot (Jean Boussier)]

• [Feature #18885] End of boot advisory API (byroot)
  • I'm about to merge the first optimizations, are we still ok with Process.warmup as a name?
• [Feature #19561] ObjectSpace::WeakMap#delete and ObjectSpace::WeakKeyMap#delete (byroot)
  • When used as a weak set, it can be convenient to eagerly delete some keys.
• [Feature #19538] Performance warnings (byroot and eregon)
  • Disabled by default, enabled with -W:performance.
  • Would allow the VM or YJIT to indicate when it has to deoptimize because of some construct.
  • For example: transitioning to SHAPE_TOO_COMPLEX, mega-morphic call sites, etc.
• [Feature #19528] JSON.load enabling create_additions: true by default is surprising and lead to security vulnerabilities (byroot)
  • JSON.load is the natural method to reach to as "load/dump" is the standard Ruby interface (Marshal.load, YAML.load, etc).
  • Various linters have to ben load and recommend to use parse instead.
  • I think json 2.x should warn when this feature is implicitly enabled, and json 3.x should disable it by default.
  • This has led to multiple CVE in the past.
• [Feature #19560] IO#close_on_fork= and IO#close_on_fork? (byroot)
  • O_CLOFORK was added to the POSIX spec in 2020, and some system (not Linux) supports it.
  • Ruby could emulate it.
  • Such feature would make it much easier to write fork-safe network clients.

[peterzhu2118 (Peter Zhu)]

• [Feature #19571] Add REMEMBERED_WB_UNPROTECTED_OBJECTS_LIMIT_RATIO to the GC (peterzhu2118)
  • We tested on production traffic on Shopify's Storefront Renderer.
  • We saw a significant improvement in GC time (0.67x on average and 0.49x for p99) and response time (0.96x on average and 0.86x for p99).

[jeremyevans0 (Jeremy Evans)]

• [Bug #4040] SystemStackError with Hash[*a] for Large a (jeremyevans0)
  • @ko1 (Koichi Sasada) fixed this issue for iseq methods accepting splats in Ruby 2.2.
  • We fixed this issue for cfunc methods in January.
  • I found this issue applies to all other method calls.
  • I developed a fix for all cases I could identity (yjit will need updates for it).
  • The fix results in a minor performance decrease in microbenchmarks.
  • I have attempted to offset this performance decrease with major optimizations to bmethod, send, symproc, and method_missing calls, using knowledge gained from fixing this bug.
  • Do we want to fix this issue, and if so, how much of a performance loss are we willing to accept?
  • I propose we should accept this fix if it doesn't result in more than 0.2% performance decrease in production benchmarks such as railsbench.
• [Bug #19533] Behavior of ===/include? on a beginless/endless range (nil..nil) changed in ruby 3.2 (jeremyevans0)
  • Do we want to make beginless+endless Range#include? return true for all objects?
  • A decision was made in August 2022 to have include? raise exception for beginless/endless ranges.
  • Looking back, it's possible that it was desired for the beginless^endless case and not the beginless+endless case.
• [Bug #19564] Range.cover? fails for Range wrapped in SimpleDelegator (jeremyevans0)
  • Do we want to handle delegate objects in Range#cover??
  • If so, do we want to add Range#to_range and have Range#cover? check for that method and call it if it exists?
• [Feature #19573] Add Class#singleton_inherited (jeremyevans0)
  • This would be similar to Class#inherited, but called for singleton class creation instead of normal subclass creation.
  • This could be used to warn or raise on singleton class creation, or modify the object to change behavior.

[matz (Yukihiro Matsumoto)]

• [Bug #19246] Rebuilding the loaded feature index much slower in Ruby 3.1
  • Personally asked for the situation.

[palkan (Vladimir Dementyev)]

• [Fearure #19591] Add symbolize_keys to MatchData#named_captures
  • Looking for Matz's feedback

[mame (Yusuke Endoh)]

• Description updated (diff)
• Status changed from Open to Closed