Misc #19240: DevMeeting-2023-01-19

• Status: Closed
• Assignee: -

[mame (Yusuke Endoh)]

The next dev meeting

Date: 2023/01/19 13:00-17:00 (JST)
Log: https://github.com/ruby/dev-meeting-log/blob/master/2023/DevMeeting-2023-01-19.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 2023/01/16. 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

[duerst (Martin Dürst)]

• [Feature #13890] Allow a regexp as an argument to 'count'
  • Recent interest, should be easy to implement

[byroot (Jean Boussier)]

• [Feature #19245] Array#pack should have a strict mode (byroot)
  • Currently it silently modulo the arguments if they are too big for the type.
  • This can can lead to silent data loss.
  • I believe pack need a "strict mode" that would instead raise an error, e.g. [4096].pack("C", strict: true) # => 4096 out of char range (RangeError) (like Integer.chr)
  • It may make sense to gradually make the strict mode the default in future versions.
• [Feature #19236] Allow to create hashes with a specific capacity from Ruby (byroot)
  • This is significantly more efficient when you need to create large hashes of known size.
  • rb_hash_new_capa(long) was added as a C-API in 3.2
  • We're still missing a similar API in Ruby.
  • What about Hash.create(capacity: 4096) ?

[kjtsanaktsidis (KJ Tsanaktsidis)]

• [Feature #19179] Support parsing SCM_CRED(ENTIALS) messages from ancillary messages
  • Various UNIX-y OS's support getting the (e)uid, (e)gid, and pid from unix sockets - some with a socket option, some with an ancillary message
  • Ruby's socket module already has code to parse these, but it's only used for Socket::Options#inspect and Socket::AncillaryData#inspect
  • I propose enhancing Socket::Options and Socket::AncillaryData to conveniently support this functionality where it's available
• [Feature #19144] Ruby should set AI_V4MAPPED | AI_ADDRCONFIG getaddrinfo flags by default
  • Many older and widely-deployed glibc versions have a bug which can cause a 5-second delay in DNS resolutions when IPv4/IPv6 are both enabled
  • When IPv6 is not in use, the AI_ADDRINFO flag to getaddrinfo works around this bug by not making AAAA requests if the host does not have a useable IPv6 address.
  • This bug often shows up in Ruby programs in particular, because Ruby accidentally overwrites one of the overlapping levels of glibc defaults, and causes AI_ADDRCONFIG to be unset where it would normally be set by default.

[hsbt (Hiroshi SHIBATA)]

• [Bug #19260] ruby/spec is failed with Ruby 3.3
  • How handles these issues?
  • https://bugs.ruby-lang.org/issues/13671 is still open status
  • https://bugs.ruby-lang.org/issues/18518#note-8 @nobu (Nobuyoshi Nakada) ?

[mame (Yusuke Endoh)]

• [Feature #18285] NoMethodError#message uses a lot of CPU/is really expensive to call (mame)
  • I have created a PR to change the message of NoMethodError based on @akr (Akira Tanaka) 's proposal. https://github.com/ruby/ruby/pull/6950
  • Why don't you give it a try on master?

[hsbt (Hiroshi SHIBATA)]

• [Misc #16671] BASERUBY version policy
  • How about my proposal?

[shugo (Shugo Maeda)]

• [Feature #19314] String#bytesplice should support partial copy
  • Not only I but also kazuho-san wants this feature: https://twitter.com/kazuho/status/1611279616098070532
  • Can the length be omitted in the destination?

[Eregon (Benoit Daloze)]

EDIT: I remove #19315 for this meeting's agenda, let's discuss it at a later meeting.

[Eregon (Benoit Daloze)]

• [Bug #18518] NoMemoryError + [FATAL] failed to allocate memory for twice 1 << large (eregon)
  • I would like a decision on this ticket
  • I propose to raise RangeError (shift width too big) if shift width is >= 2**31 (32-bit signed integer) for all platforms, like it already behaves on 32-bit platforms.
  • Current behavior is useless on 64-bit platforms, it's slow and then eventually NoMemoryError.
  • This will restore consistency for this between Ruby, JRuby and TruffleRuby.

[peterzhu2118 (Peter Zhu)]

• [Bug #19289] RbConfig::CONFIG["STRIP"] should keep rb_abi_version and rb_abi_version should always be part of Ruby
  • In dev meeting #18557 it was decided that we would keep rb_abi_version on both development and release (so there would be no difference in behavior between development and release).
  • In commit cd1a0b3 it was changed so that rb_abi_version would not be defined in release.
  • This has caused issues in gems such as gRPC and rb-sys (which compiles Rust based gems) because they expect rb_abi_version to be present.
  • This will continue to cause problems in the future due to this difference in behavior.

[hsbt (Hiroshi SHIBATA)]

• [Bug #18658] Need openssl 3 support for Ubuntu 22.04 (Ruby 2.7.x and 3.0.x)
  • Should we merge openssl-3.0.x into ruby_3_0 before moving security only maintenance phase?

[ioquatix (Samuel Williams)]

• [Feature #19333] Setting (Fiber Local|Thread Local|Fiber Storage) to nil should delete value in order to avoid memory leaks.

[jeremyevans0 (Jeremy Evans)]

• [Feature #19347] Add Dir.fchdir (jeremyevans0)
  • This method is useful when passing directory file descriptors through UNIX sockets or to child processes to avoid TOCTOU vulnerabilities.
  • We already have Dir#fileno, but nothing that can use the returned directory file descriptor.
  • Is it OK to add this method?
  • If so, is the implementation in the pull request acceptable?
• [Bug #19237] Hash default_proc is not thread-safe to lazy-initialize value for a given key (jeremyevans0)
  • Similar to hash[key] ||= value in multiple concurrent threads.
  • One workaround is using a Mutex and mutex.synchronize{hash[key] = value unless hash.key?(key)} inside the default proc.
  • Another workaround is to avoid using a default proc to set values in a hash accessed by multiple concurrent threads.
  • Is this a bug? If not, should we update the documentation of Hash.new and/or Hash#default_proc= regarding use with concurrent threads?
• [Bug #19286] What should kwargs' arity be? (jeremyevans0)
  • Do we want to change how arity is calculated for methods/procs accepting/requiring keywords?
  • I don't think it is worth breaking backwards compatibility here (#parameters can be used for keyword argument details).
• [Bug #19293] The new Time.new(String) API is nice... but we still need a stricter version of this (jeremyevans0)
  • Do we want to add a keyword argument to Time.new for stricter parsing, or another Time method that is stricter?

[mame (Yusuke Endoh)]

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