DevMeeting-2018-05-17

Date: 2018/05/17 (Thu) Time: 14:00-18:00 (JST) Place: Cookpad Inc. (Tokyo, Japan) Sign-up: https://ruby.connpass.com/event/85917/

Please comment your favorite ticket numbers you want to ask to discuss with your SHORT comment or summary. (your summary/comment will help us because we don’t need to read all of ticket comments)

DO NOT discuss then on this ticket, please.

From this time, we use a ticket to make dev-meeting agenda page instead of a wiki page https://bugs.ruby-lang.org/projects/ruby/wiki.

NOTE

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 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 is scheduled according to when/where we can reserve Matz’s time.

Agenda

Next dev-meeting

https://bugs.ruby-lang.org/issues/14769

About 2.6 timeframe

preview2 will be released RubyKaigi 2018 day 1 or 2

From Attendees

• [Bug #14699] Subtle behaviors with endless range (mame)
• [Feature #14724] chains of inequalities (Martin)
  • Proposal by gotoken (Kentaro Goto) to allow 0 <= a < 10 as a shortcut of 0<=a && a<10, and so on; patch by Nobu avaliable

(will be edited later) (if you have a write access permission, please list directly)

From non-attendees

• [Feature #14697] Introducing Range#% as an alias to Range#step (mrkn)
  • Matz already commented LGTM. Please judge whether this is acceptable.
• [Feature #14724] chains of inequations (mrkn)
  • This new syntax can also reduce the duplicated evaluations of common terms, such as timeval.tv_sec in the descriptiohn of the proposal.

Functional programming: (zverok)

• [Feature #6284] Add composition for procs
  • 6-year-old proposal. Matz: “Positive about adding function composition. But we need method name consensus before adding it? Is #* OK for everyone?”
• [Feature #13581] Syntax sugar for method reference
  • 1-year-old proposal. Matz: “I am for adding syntax sugar for method reference. But I don’t like proposed syntax (e.g. ->). Any other idea?”
• [Feature #11161] Proc/Method#rcurry working like curry but in reverse order
  • 3-year-old proposal with absolutely no reaction. I’ve added real-life examples there several months ago.
• [Feature #14390] UnboundMethod#to_proc
• [Feature #14423] Enumerator from single object (Object#enumerate)

Misc:

• [Bug #14575] Switch Range#=== to use cover? instead of include? (zverok)
• [Feature #14473] Add Range#subrange? ( greggzst )
• [Feature #14097] Add union and difference to Array ( ana06)
• [Feature #14105] Introduce xor as alias for Set#^ ( ana06)
• [Misc #14760] cross-thread IO#close semantics (normal)

Docs (probably not to discuss on development meeting, but I am not sure where I can post it to have it more visible) (zverok):

• Method: https://bugs.ruby-lang.org/issues/14483
• Proc: https://bugs.ruby-lang.org/issues/14610
• MatchData: https://bugs.ruby-lang.org/issues/14450
• yield_self: https://bugs.ruby-lang.org/issues/14436

(will be edited later) (if you have a write access, please list directly)

Log

Date: 2018/05/17 (Thu)

Time: 14:00-18:00 (JST)

Place: Cookpad Inc. (Tokyo, Japan)

logs

Next dev-meeting

• 5/31 11:30- at RubyKaigi
• 6/21 14:00- at Cookpad

About 2.6 timeframe

• Usa: preview2…
• Hsbt: next week?
• Naruse: not that immediate.
• Naruse: hopefully the day before the Kaigi.

About Neon EOL

• Neon (one of our oldest host) is running debian wheezly, which will be EOLed this month.

From Attendees

• [Bug #14699] Subtle behaviors with endless range (mame)

• Mame: what to do with #max for endless ones?

• Usa:

• Exception?

• Retuen nil?

• Akr: is it possible to detect if a range is endless or not?
• Mame: #end can be used that case.
• Mrkn: I like nil. (by text chat)

• Ko1: let’s discuss this issue with mrkn.

• [Feature #14724] chains of inequalities (Martin)

• Proposal by gotoken (Kentaro Goto) to allow 0 <= a < 10 as a shortcut of 0<=a && a<10, and so on; patch by Nobu avaliable
• Ko1: what do you think matz?
• Matz: I understand people want to write this.
• Matz: what’s this nobu’s patch?
• Nobu: that patch breaks existing code.
• Matz: what code?
• Nobu: shell.rb
• Ko1: so this patch breaks DSLs that use < characters.
• Usa: what about #< to return self instead?
• Matz: I think I tested that before, but did not work.
• Akr: ruby up to 0.51 behave like that.

From non-attendees

• [Feature #14697] Introducing Range#% as an alias to Range#step (mrkn)

• Matz already commented LGTM. Please judge whether this is acceptable.
• Akr: I think this should go.

• (nobody at the meeting had any objections)

• [Feature #14724] chains of inequations (mrkn)

• This new syntax can also reduce the duplicated evaluations of common terms, such as timeval.tv_sec in the descriptiohn of the proposal.

• Same issue as above. See above.

• Object#then (Usa)

• Usa: no one can commit this feature but matz.

• Matz: I already did it for mruby.

• Jemalloc (shyouhei)

• Shyouhei: We are going to change configure default, OK?

• [Feature #5400] Remove flip-flops in 2.0 (mame)

• Mame: now is the time?
• Matz: now?
• Usa: everyone who use this feature are in this meeting, I think
• Akr: is this worth deleting?  Nobody uses this means it has no benefit.

Functional programming: (zverok)

• [Feature #6284] Add composition for procs

• 6-year-old proposal. Matz: “Positive about adding function composition. But we need method name consensus before adding it? Is #* OK for everyone?”
• (ko1 is describing the Ruby Hack Challenge results)
• Usa: is it OK to add #compose at the first place? Apart from the operators?
• Shyouhei: I think it’s OK to have the feature itself but for the operators?
• Mame: do we need composition?
• Matz: it seems comment #60 introduces Symbol#».
• Shyouhei: that sounds radical.
• Matz: shouldn’t we start with Proc…

• Matz: and I think we need both OOP-ish / FP-ish way.

• [Feature #13581] Syntax sugar for method reference

• 1-year-old proposal. Matz: “I am for adding syntax sugar for method reference. But I don’t like proposed syntax (e.g. ->). Any other idea?”
• (reviewed comment #21)
• Preference by the attendees is .:
• Matz: .: is better than ->

• Mame: there is a patch cf #12125 https://bugs.ruby-lang.org/issues/12125

• [Feature #11161] Proc/Method#rcurry working like curry but in reverse order

• 3-year-old proposal with absolutely no reaction. I’ve added real-life examples there several months ago.
• Mame: I don’t think reverse currying works for variadic methods

• Akr: the “real-life” example is adding keyword arguments, which is not what currying is talking about

• [Feature #14390] UnboundMethod#to_proc

• What is the expected behaviour here?

• [Feature #14423] Enumerator from single object (Object#enumerate)

• Naruse: interesting proposal
• Akr: adding method to Object sounds too radical to me.
• Usa: I think there is a chance if this is a method of Enumerable.
• Shyouhei: my feeling is this should start as a gem.

Misc:

• [Bug #14575] Switch Range#=== to use cover? instead of include? (zverok)

• Seems there is a compatibility issue.
• Usa: does it matter? No one actually expects Range#=== to behave as #cover?
• Matz: the problem, if any, is String.  What about keep string’s behaviour as is?

Docs (probably not to discuss on development meeting, but I am not sure where I can post it to have it more visible) (zverok)

• Method: https://bugs.ruby-lang.org/issues/14483
• Proc: https://bugs.ruby-lang.org/issues/14610
• MatchData: https://bugs.ruby-lang.org/issues/14450

• yield_self: https://bugs.ruby-lang.org/issues/14436

• Naruse: This meeting is not for documents. Just assign to docs member group.

• [Feature #14473] Add Range#subrange? ( greggzst )

• Shyouhei: what to do with String this case?
• Akr: the simplest is to only concern the both ends.
• Usa: Hash has > and <

• Mame: use case not clear.

• [Feature #14097] Add union and difference to Array ( ana06)

• Matz: create #union and #union!, or create mutating #union ?
• Matz: I want to add mutating method but #union does not sound mutating.
• Knu: Set has #union, which is not destructive.

• Akr: to me it seems mutating or not is not the key point of this proposal. The key point is provide a (readable) name addition to an operator, I guess.

• [Feature #14105] Introduce xor as alias for Set#^ ( ana06)

• Assign to knu

• [Misc #14760] cross-thread IO#close semantics (normal)

• Ko1: I made it theoretically possible to unblock IO#select. However there is a difficult problem for  IO#copy_stream.