00:10 finanalyst left
Geth rakudo/main: 69f2b375f4 | (Nick Logan)++ (committed using GitHub Web editor) | 2 files
Keep EVAL of a RakuAST node from running its MAIN (#6755)

Previously EVAL compiled a RakuAST node without setting $*INSIDE-EVAL. The node's mainline was then wrapped in a call to RUN-MAIN, which ran MAIN against the program's @*ARGS. Without arguments,
  `Q|sub MAIN($x) { }|.AST.EVAL` printed a usage message and exited. With
matching arguments MAIN ran and EVAL returned Nil. This sets
  $*INSIDE-EVAL while the node compiles, as the string form already does.		 00:14
japhb ugexe: Thanks for the Test::Async fix. Testing now while I eat dinner. :-) 01:57
ugexe [Coke]: i was thinking i would finish reviewing github.com/rakudo/rakudo/pull/6754 tonight but i dont think i will, so im done with commits tonight 02:01
japhb ugexe: Test::Async got farther but still failed, this time on the async stress test, where it tries to run 4001 times, but only managed 3999. 02:45
I'm assuming that's flaky, so trying again.
... aaaaand JSON::Class pins the old version of Test::Async. 03:05
03:30 coleman left, coleman joined 05:00 Pixi` is now known as Pixi 09:35 finanalyst joined
patrickb Binaries for 2026.09 are out. 10:25
tellable6 2026-09-28T17:48:05Z #raku-dev <[Coke]> patrickb - source release is done for 2026.09
lizmat whee! 10:34
apogee_ntv patrickb++ 10:43
11:40 finanalyst left
lizmat notable6: weekly 11:52
notable6 lizmat, 1 note: 2026-09-28T08:27:30Z <lizmat>: hankache.github.io/raku-koans/
lizmat notable6: weekly reset
notable6 lizmat, Moved existing notes to “weekly_2026-09-29T11:52:36Z”
[Coke] patrickb: I am very happy you didn't report any issues. :) 13:02
I get concerned when I have to pause half way and then go back. :)
13:03 [Tux] left
[Coke] Thanks to everyone who contributed to the release (lots of commits) but esp. those who jumped in to resolve issues found during the release process. 13:06
(pinned versions) btw, this is another thing that blin doesn't do. you get the latest version, IIRC, not a pinned version. 13:07
13:09 [Tux] joined
[Coke] patrickb: does your process also build rakudo.star or did that one kick off automatically at some point? (got the email on it) 13:20
patrickb I don't do star. That's a GH actions pipeline afaik. 13:21
lizmat And yet another Rakudo Weekly News hits the Net: rakudoweekly.blog/2026/09/29/2026-...nstreamed/ 13:22
[Coke] so, automatic. kewl. 13:26
lizmat wish making the Weekly was automatic :-) 13:31
anyways, sorry for it being one day late: a bad bug caught me on Sunday, and I slept most of Monday. Feeling much better now :-) 13:32
[Coke] I hope some parts of it are automatic! 13:38
someone could script something to get new modules, read the bot etc.
thanks for taking it back on.
lizmat yeah, librasteve did that 13:45
[Coke] oh, good. 13:52
still a bunch of manual stuff, no doubt.
[Tux] lizmat, CSV has compile issues now 14:04
Type check failed in binding to parameter '$identifier'; expected Str but got Array (["callframe"])
at test-t.pl:4 14:05
------> <BOL>⏏use Text::CSV;
[Coke] ugexe: new blin run has at least one find so far. Will report out 14:29
Wonder if that's a) not tested in Blin for some reason, b) Slang related. 14:50
yah, it's AlwaysFail at the moment. :| 14:51
Geth marketing/master: 5 commits pushed by (Matt Doughty)++, (Will Coleda)++ 14:54
[Coke] [Tux]: I just (on mac) did a zef install --deps-only Text::CSV - that was fine. did a 'zef look Text::CSV' and then 'zef test .' - tons of deprecation warnings about perl, but it passed the tests. 15:00
I was able to install it. 15:02
ugexe probably needs to update their dependencies 15:05
[Coke] (git head does fail one test, but not with that failure) 15:06
[Tux]: I assume this is on 2026.09 raku, yes? 15:07
ugexe github.com/raku-community-modules/...16fL14-R17 15:08
15:20 kjp left
lizmat yeah, I think [Tux] needs to do an update on Slang::Tuxic 15:48
16:54 kjp joined
Geth setup-raku/dependabot/npm_and_yarn/undici-6.29.0: b063c25299 | dependabot[bot]++ (committed using GitHub Web editor) | package-lock.json
Bump undici from 6.28.0 to 6.29.0

Bumps [undici](github.com/nodejs/undici) from 6.28.0 to 6.29.0.
  - [Release notes](github.com/nodejs/undici/releases)
  - [Commits](github.com/nodejs/undici/compare/v.....v6.29.0)
  ---
... (7 more lines)		 17:52
setup-raku: dependabot[bot]++ created pull request #64:
Bump undici from 6.28.0 to 6.29.0
18:23 notable6 left, benchable6 left 18:24 notable6 joined, benchable6 joined
[Coke] github.com/coke/raku-blin-release-results 19:00
Mux might be real. redoing them both (one was OOM) 19:01
through 69f2b375f4 19:03
linkable6 (2026-09-28) github.com/rakudo/rakudo/commit/69f2b375f4 Keep EVAL of a RakuAST node from running its MAIN (#6755)
[Coke] lizmat: also, glad you're feeling better!! 19:05
ugexe Mux has failed in previous runs
19:06 timo left
ugexe Memory peak: 1G (swap: 418.9M) 19:14
i wonder if it should automatically run things getting killed using that little (relatively, for raku) amount of memory
rerun rather
[Coke] ugexe: nevermind, all good on rerun. Mux must have a bad flapper (blin does flap protection) 19:16
ah, or a resource issue
interesting that it uses a gig, didn't die from OOM, but failed a test 19:17
ugexe i thought it just detects that overall its going to run out of memory, and then systemd or whatever starts killing processes based on some heuristic 19:18
the OOM i was mentioning earlier was for crust though, which didnt fail any test 19:19
[Coke] going forward, will skip the rakuast section of the changelog 21:19
If we want to have those sections of fix/add/improvement, would be nice if there was a hint in the commits. 21:20
(and internal)
at work we have a bunch of markdown in each commit message which is probably overkill 21:22
ugexe yeah im not sure if its worth mentioning the RakuAST anymore or not. ive still been including it since we still havent technically removed the legacy backend and i want to make it clear certain changes may not be getting backported 21:26
Geth rakudo/main: c6fff750c2 | (Will Coleda)++ | docs/release_guide.pod
claim 2026.10
 21:29
[Coke] ok - I'll prbably sort them into the remaining buckets for changelog, but feel free to use it in commits.
ugexe i think its safe to assume anything im doing is fixes 21:36
unless stated otherwise
Geth ¦ rakudo: lizmat self-assigned RakuAST: adverb elems in sequence endpoint github.com/rakudo/rakudo/issues/6046 22:28
[Coke] changelog caught up 23:07
Geth setup-raku: 99d3a48056 | dependabot[bot]++ (committed using GitHub Web editor) | package-lock.json
Bump undici from 6.28.0 to 6.29.0 (#64)

Bumps [undici](github.com/nodejs/undici) from 6.28.0 to 6.29.0.
  - [Release notes](github.com/nodejs/undici/releases)
  - [Commits](github.com/nodejs/undici/compare/v.....v6.29.0)
  ---
... (8 more lines)		 23:53

Powered by Cro and the Raku® Programming Language.

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

Thank you!