lizmat_ [Coke]++ jdv++ 00:03
[Coke] It was just me this time. :) 00:07
lizmat okidoki :-) 00:18
[Coke] jdv did merge a PR for me, that counts, I guess. :) 00:32
.seen jnthn 01:01
tellable6 [Coke], I saw jnthn 2024-12-02T14:00:59Z in #moarvm: <jnthn> :q
patrickb I've just uploaded the precompiled release files to rakudo.org. Thanks for the release [Coke]! 08:25
lizmat nine: looks like we lost all nativecall tests in "make test" under RakuAST 12:00
appears related to the handling of require 12:01
nine lizmat: noticed that, too yesterday evening and working on a fix 12:02
lizmat ack
nine The fix really is making late bound sub calls work in BEGIN blocks 12:03
I.e. my &call-foo := BEGIN { -> { foo } }; sub foo() { }
lizmat nine: running the bootrstapped version: is there something else that I need to do apart from switching to the branch ? 12:16
nine You still need to set RAKUDO_RAKUAST=1 but apart from that no. I actually ran it accidentally a few times 12:17
lizmat also at build time? 12:18
nine yes 12:20
I could probably remove that switch in the branch and force it to RakuAST since it's not buildable with the old frontend anyway 12:21
Geth rakudo/bootstrap-rakuast: 16 commits pushed by (Stefan Seifert)++
review: github.com/rakudo/rakudo/compare/c...b86e17aad6		 13:49
rakudo/main: 37c301eaa0 | (Stefan Seifert)++ | 3 files
RakuAST: fix calling post-declared subs from role methods

When a Role's method tries to call a sub that is declared after the role itself, the lookup for this routine would fail. That's because role bodies are compiled immediately after the role is parsed. At this point the callee does not exist yet, so we won't find it. However at runtime the lookup may very vell succeed. ... (7 more lines)
rakudo/main: 4f5f4cde4f | (Stefan Seifert)++ | src/Raku/ast/variable-declaration.rakumod
RakuAST: fix inlined containers in CStructs
nine NativeCall is back on track and this also gets rid of one more workaround in the bootstrap branch :) 13:50
lizmat nice! 13:51
Geth rakudo/bootstrap-rakuast: 7 commits pushed by (Stefan Seifert)++ 14:16
nine I merged the non-zero-chance-of-failure fixes from bootstrap-rakuast to main 14:17
Now the bootstrap branch contains just 6 workarounds and removes 1 workaround. I'd like to find better solutions for those workarounds but none of them feel terrible. So unless better solutions appear, I intend for them to go in. 14:19
[Coke] releaseable6: status 14:22
releasable6 [Coke], Release date for Rakudo 2025.03 is listed in “Planned future releases”, but it was already released.
[Coke], and I oop! Backtrace: gist.github.com/df5a481ebf861ef185...82616bd278
Geth rakudo/main: 184d325706 | (Will Coleda)++ | docs/release_guide.pod
update next release date
 14:23
[Coke] releaseable6: status
releasable6 [Coke], Release date for Rakudo 2025.03 is listed in “Planned future releases”, but it was already released.
[Coke], and I oop! Backtrace: gist.github.com/0172d26d4caac85b5f...566988eb7a
[Coke] releaseable6: status 14:24
releasable6 [Coke], Release date for Rakudo 2025.03 is listed in “Planned future releases”, but it was already released.
[Coke], and I oop! Backtrace: gist.github.com/96df471088b354c628...46dc275947
Geth rakudo/main: bc5b2673f5 | (Will Coleda)++ | docs/release_guide.pod
actually claim next release
 14:25
[Coke] releaseable6: status
releasable6 [Coke], Release date for Rakudo 2025.03 is listed in “Planned future releases”, but it was already released.
[Coke], and I oop! Backtrace: gist.github.com/7fe3425b8c9ed260c9...9d053a9b10
Geth nqp/main: 06c8273f50 | (Elizabeth Mattijsen)++ | tools/templates/MOAR_REVISION
Bump MoarVM to get 32/64bit writing to file working
 14:26
rakudo/main: 9c9d822e44 | (Will Coleda)++ | docs/release_guide.pod
next release
[Coke] releaseable6: status 14:27
releasable6 [Coke], Release date for Rakudo 2025.03 is listed in “Planned future releases”, but it was already released.
[Coke], and I oop! Backtrace: gist.github.com/cc3122deccbf347605...fe1e4ef45b
[Coke] I think it's running off a cache now; will check in privmsgs going forward 14:33
wonder if it's actually using the month as part of the calculation and I have to wait until Apr-01 14:37
tag is there, release is there.
Geth rakudo/main: 606e4eb084 | (Elizabeth Mattijsen)++ | tools/templates/NQP_REVISION
Bump NQP to get 32/64bit writing to file working

Fixes #5776		 14:38
[Coke] yah, running it against a local copy, it's seeing 2025.04 as the next release. 14:45
.seen alexdaniel 14:47
tellable6 [Coke], I saw alexdaniel 2024-09-01T02:48:19Z in #raku-dev: <AlexDaniel> and I'm happy to see that this tooling is getting some love again <3
lizmat nine: fwiw, I can't get the bootstrapped version to build: 14:48
Cannot find method 'add_attribute' on object of type Perl6::Metamodel::ModuleHOW
at src/Raku/ast/package.rakumod:408 (/Users/liz/Github/rakudo/blib/Perl6/BOOTSTRAP/v6c.moarvm:ATTACH-ATTRIBUTE)
nine Huh...just tried again on a clean directory and after make clean. Just did RAKUDO_RAKUAST=1 make 14:56
lizmat: oh, did you try make install?
lizmat ah, yes... 14:58
nine This was actually my first time ever trying make install with RakuAST :D And yes, see that error as well 14:59
lizmat so where do I find the updated "raku" if I don't do "make installl" ? 15:06
nine ./rakudo-m
in the rakudo directory
lizmat thanks 15:09
so the error that blocked the .assuming, is much more benign in the bootstrapped rakudo 15:10
Could not find a compile-time-value for lexical Nil
nine I'd assume that (pun intended)
lizmat I was hoping It'd just work :-) 15:11
[Coke] jdv: I put myself down for April, unless you want it. 15:44
jdv if its the rakuast one then yes 15:45
iirc the release guide is not updated for the next
[Coke] We don't know when that is yet, I don't think. nine, lizmat, let us know. :)
It is now.
... I am concerned that again, I missed a step here - might need to patch releaseable6 so it's all set before we start going through the changelog next release. I'll take that as a task for this week 15:47
jdv nine: any good guess possible when rakuast will supplant legacy?
[Coke] (the next release wasn't listed in the tag/release in the release_guide but only the announcement). :(
(So I'll do April as penance unless it's the the rakuast one. :) 15:48
lizmat: github.com/Raku/roast/issues/872 - I tried to cherry pick that one commit, but there are conflicts.
jdv i feel like we should know well in advance as well as be very conservative about the schedule so maybe not hard to know whuch release will land it
[Coke] I think getting us to 100% is step one - if that happens before April, then we say May is the one that gets it. 15:50
That gives core devs a bit to do any cleanup before we move it out of experimental. Are we also bumping to 6.e with that release? 15:51
jdv i think that might be a rsc thing. no idea.
[Coke] jdv: can you verify this also happens to you: github.com/rakudo/rakudo/issues/5775 15:54
nine The RSC is the last resort to get some decision when the community fails to achieve consensus
[Coke] and, do you have an opinion on github.com/rakudo/rakudo/issues/5827 ?
jdv [Coke]: can you shoot he a quick email on those? i get a lot of gh noise. 15:56
*me 15:57
does the community agree wut 6.e is? 15:58
[Coke] "should include working rakuast, anything already in 6.e.PREVIEW" - not sure we have a process (or a list that's a work in progress) 15:59
Seems like something we should have in a raku/ repo somewhere for language releases going forward: Language Changelog 16:00
jdv that unfortunately doesnt answer the question if rakuast triggers 6.e though, right?
nine m: my &pos-match-block = { say "here"; True }; my $pos-b where &pos-match-block = 42; use Test; dd $pos-b ~~ Seq:D; is-deeply $pos-b, 42 16:07
camelia Type check failed in assignment to $pos-b; expected <anon> but got Int (42)
in block <unit> at <tmp> line 1
nine But:
m: my &pos-match-block = { say "here"; True }; my $pos-b where &pos-match-block = 42; use Test; is-deeply $pos-b, 42
camelia here
ok 1 -
nine What a bizarre error! The assignment fails if later we would do that type comparison
[Coke] jdv: I would argue that rakuast has to be in 6.e but it doesn't force it to occur. 16:12
I would propose at least (based on previous discussions here) removing experimental macro support (and maybe experimental cache) 16:13
I mean, maybe it makes sense to put this in as a PS ticket, with the result being a doc describing what is in 6.e language. 16:14
github.com/Raku/problem-solving/tr...s/language already has some "what's in language version X" notes 16:15
nine Dubiously legal Raku syntax riddle of the day. What does the following code mean exactly? 17:21
my (2);
m: my (2);
camelia ( no output )
gfldex m: for my (2) { .say }; 17:33
camelia (<anon>)
[Coke] m: dd my ([+]) 18:02
camelia ([],)
jdv nine: stop:) 18:13
[Coke] m: my ('stop') 18:21
camelia ( no output )
nine m: my (2) = 3; 18:26
camelia Type check failed in assignment to anon; expected <anon> but got Int (3)
in block <unit> at <tmp> line 1
nine The error message does give a hint
It also helps to think about what a my ($x); exactly is
[Coke] m: my ($a,$b) =2,3; dd $a; dd $b 18:36
camelia $a = 2
$b = 3
[Coke] shouldn't that still require var names in the parens, though? 18:37
(also is this one of the few places where we use parens, not just a comma?
few/only
m: my (2) = 2; 18:38
camelia ( no output )
[Coke] m: my (2) = 2,3,4; 18:39
camelia ( no output )
jdv how does that even make sense? 18:43
[Coke] ls 19:13
releasable6: status 20:58
releasable6 [Coke], Next release in ≈25 days and ≈22 hours. There are no known blockers. Changelog for this release was not started yet
[Coke], Details: gist.github.com/59125bc6c42b0c20fc...5d5d98c704
[Coke] \o/ it *was* cached, I didn't break it. :)
releasable6: status 21:03
releasable6 [Coke], Next release in ≈25 days and ≈21 hours. There are no known blockers. Changelog for this release was not started yet
[Coke], Details: gist.github.com/c7fb9c4cd6d7fee019...890a7eb569
[Coke] ... ok, it *has* started, stop caching things! :)
releasable6: status 21:04
releasable6 [Coke], Next release in ≈25 days and ≈21 hours. There are no known blockers. 14 out of 366 commits logged
[Coke], Details: gist.github.com/8375974342711cb88c...f93d9830cf
[Coke] .... where is it getting 366...
I did update github.com/rakudo/rakudo/wiki/ChangeLog-Draft for 2025.04 already, trying to at least keep up with the rakuast commits which require processing but no thought. 21:05
nine It makes "sense" once you see it as a signature declaration. 21:29
The part in the parens in my ($a, $b) is a signature, same as on a routine.
So my (2) is the equivalent of a sub (2). 21:30
Geth rakudo/main: 5 commits pushed by (Stefan Seifert)++ 22:08
nine And now we have that madness also with RakuAST ;)
lizmat m: say Q|my (2)|.AST.DEPARSE # wow 22:21
camelia my (2)
lizmat nine: looks like we lost t/12-rakuast/doc-declarator.rakutest after this last batch of commits 22:37
No such method 'meta-object' for invocant of type 'RakuAST::Trait::Of'
afk&
Geth rakudo/main: dd0b647d45 | (Will Coleda)++ | docs/announce/YYYY.MM.md
remove literal YYYY.MM.dd typo
 23:19
coleman This will become an emergency very soon unless addressed github.com/Raku/infra/issues/61 23:38
jdv nine: oh geez 23:55

Powered by Cro and the Raku® Programming Language.

Please report any issues / comments / feature requests as an issue on App::Raku::Log.

Thank you!