Geth rakudo: farisbasha++ created pull request #6938:
Fix the dead #raku-dev IRC log link in CONTRIBUTING
02:37
05:52 hurufu joined 05:54 hurufu left 05:55 hurufu joined 07:24 hurufu left 07:26 hurufu joined 08:24 hurufu left 08:25 hurufu joined
Geth rakudo/main: e3613c0d12 | (Faris Basha TM)++ (committed using GitHub Web editor) | CONTRIBUTING.md
Fix the dead #raku-dev IRC log link in CONTRIBUTING (#6938)

The colabti.org irclogger page for #raku-dev now returns 404. Link to the #raku-dev logs on irclogs.raku.org instead.
09:03
lizmat argh... looks like b0be8005a74 broke the JVM build 09:24
linkable6 (2026-11-15) github.com/rakudo/rakudo/commit/b0be8005a7 Simplify some container/attribute logic (#6914)
lizmat which I guess also means that 2026.11 does not build on JVM :-( 09:25
which in turn breaks NQP CI
Geth rakudo/lizmat-136: 34a5e1501c | (Elizabeth Mattijsen)++ | src/Perl6/bootstrap.c/BOOTSTRAP.nqp
Hopefully unbreak the JVM build

Introduced by b0be8005a74
09:26
rakudo: lizmat++ created pull request #6939:
Hopefully unbreak the JVM build
09:27
lizmat I guess this also implies the JVM backend is close to death :-( nobody noticed it not building 09:28
Otoh, by the looks of github.com/Raku/nqp/pull/872 , the JVM build was already broken on Sep 7, possibly going back to before May 11 10:04
10:04 hurufu left
lizmat possibly even before that, as it looks like before Nov 9, 2025 the JVM was not in NQP's CI 10:05
10:13 hurufu joined
refactus on one hand I think about all the extra work we've done to keep things working on jvm and try to maintain some sembalance of feature parity, and it seems like it'd be a shame and a waste to throw it all away. on the other hand I think about all the extra work we've done to continue supporting a backend that almost nobody has used since moar was developed, and it seems like a bit of a shame and a 10:50
waste to keep doing it
lizmat en.wikipedia.org/wiki/Sunk_cost 11:02
:-( 11:03
11:04 hurufu left
apogee_ntv I think the main use I would have for JVM backend (can't speak for anyone else) is being able to target Android. 11:05
11:05 hurufu joined
apogee_ntv Dont know if thats doable or close or far off 11:05
lizmat which raku++ now covers, fwiw
apogee_ntv Interesting
lizmat yeah, the world is changing 11:06
apogee_ntv It's changing fast
refactus I know the last time I tried to fudge a test under jvm for a feature I implemented, someone here was still using it and helped me try to make it work, but I guess that was several years ago now. don't even recall who it was 11:17
it was usev6. github.com/Raku/nqp/issues/727 was what I ran into. half a decade ago. but yes, sunk cost fallacy, I agree. it does kinda look that way if it's been broken for months and nobody reported it 11:39
lizmat oddly enough, the rakudo CI is green on my PR 11:42
which includes JVM builds 11:43
so I wonder what's going on here, maybe there's an issue with the NQP CI creating false negatives 11:44
someone with more Azure CI chops should have a look at github.com/Raku/nqp/blob/main/azur...elines.yml then, japhb ? 11:45
Geth rakudo/main: a364d455f0 | (Elizabeth Mattijsen)++ (committed using GitHub Web editor) | src/Perl6/bootstrap.c/BOOTSTRAP.nqp
Unbreak the JVM build (#6939)

Introduced by b0be8005a74
11:50
13:05 hurufu left, hurufu1 joined 13:07 hurufu1 is now known as hurufu
[Coke] Next scheduled release date is the day after Christmas. 13:29
I am *fine* with no release next month.
14:51 MasterDuke joined 15:08 MasterDuke left 15:35 hurufu left
Geth rakudo/lizmat-137: c8558383f8 | (Elizabeth Mattijsen)++ | 6 files
RakuAST: change ::Origins interface to use positionals

The RakuAST::Origins.new method is basically called for each RakuAST element that is generated from source, so that is a *lot*.
However, the number of places where it is being called, are few and in the core only. So it makes sense to use the slightly cheaper way of using positional argments instead of named arguments.
This also needed changing 1 test that was actually testing the internal interface.
16:59
rakudo: lizmat++ created pull request #6940:
RakuAST: change ::Origins interface to use positionals
17:03 hurufu joined 17:45 hurufu left 17:46 hurufu joined 18:30 hurufu left 18:38 hurufu joined
Geth rakudo/lizmat-138: 6f4fc8674e | (Elizabeth Mattijsen)++ | src/Raku/ast/base.rakumod
RakuAST: introduce RAKUDO_OPTIMIZE_DEBUG env variable

Shows the result of static optimizations (the "optimize" stage) on STDERR. The value specified is the name of the method that should be used to represent the before/after stage of the of the optimization. So with "DEPARSE"
... (17 more lines)
18:58
rakudo: lizmat++ created pull request #6941:
RakuAST: introduce RAKUDO_OPTIMIZE_DEBUG env variable
rakudo/lizmat-139: d26ba2f3fd | (Elizabeth Mattijsen)++ | src/Raku/ast/base.rakumod
RakuAST: reduce conditions for each optimizing recursion

Instead of storing the conditions in variables, and then checking them 4 times, use the conditions to set up an if/elsif/else structure with just the calls needed.
19:06
rakudo: lizmat++ created pull request #6942:
RakuAST: reduce conditions for each optimizing recursion
19:07