finanalyst coleman: ping 09:15
.tell coleman can we have a convo about best practice for adding to new-website? 09:16
tellable6 finanalyst, I'll pass your message to coleman
coleman . 14:46
tellable6 2025-03-23T09:16:30Z #raku-dev <finanalyst> coleman can we have a convo about best practice for adding to new-website?
coleman .tell finanalyst if you have any questions or links to failing tests add them to this ticket github.com/Raku/doc-website/issues/434 14:49
tellable6 coleman, I'll pass your message to finanalyst
finanalyst coleman: ping 16:06
tellable6 2025-03-23T14:49:39Z #raku-dev <coleman> finanalyst if you have any questions or links to failing tests add them to this ticket github.com/Raku/doc-website/issues/434
coleman finanalyst: I'll be around for a couple of hours. I can't do a call today, though. sorry for being out of pocket lately, but I just finished my first week at a new job 16:45
[Coke] I'm here, apologies, took longer to get everything back to normal. 17:35
nine: What's the current status on these numbers:
x/153 (make test), and 1295/1355 (make spectest)
releasable6: status 17:36
releasable6 [Coke], Next release will happen when it's ready. There are no known blockers. 277 out of 348 commits logged
[Coke], Details: gist.github.com/f06df1a2d9b8602bde...d5b9925df7
[Coke] Found 1298 spectest in backscroll - anyone see the latest number for 'make test'? 17:46
m: say 40*7+6 17:48
camelia 286
jdv its usually in the last weekly possibly outdated slightly 18:03
[Coke] I'll take any numner at this point. :) 18:06
releasable6: status 18:09
releasable6 [Coke], Next release will happen when it's ready. There are no known blockers. 347 out of 348 commits logged
[Coke], Details: gist.github.com/64a38ede0b4822a0b8...ed317e90d3
[Coke] releasable6: status 18:16
releasable6 [Coke], Next release will happen when it's ready. There are no known blockers. 348 out of 348 commits logged
[Coke] ok, definitely want to do the changelog weekly next time I have a release, wowza.
nine 1298 is latest 18:18
jdv the 2 big pre-release tasks are blin and changelogs 18:20
[Coke] Yup, shouldn't have saved CL for last day 18:27
nine- did find that, just did a make test run and got 143?
which *is* 1 higher than the number in last announce. 18:30
nine That's likely. lizmat++ did some work there 18:33
[Coke] Undefined template variable 'contributors' 18:38
[Coke] want to update the contributors script so you can have nqp be *next* to the rakudo folder instead of *in* it. 18:43
(or rather, in either location)
Geth rakudo/release-2025.03: 63c133a185 | (Will Coleda)++ | 3 files
Update changelog + announcement

Deliberately not logged:
  [3a336975][ac556502][2d12d3ed][775e8ff9][992f7417][60309b6a]
  [9233a89f][02482743][e633a64c][2e82abc3][fa7697e2][40576e0e]		 18:53
[Coke] ... note to self, run the release commands in a tmux window on the remote box so when my (*@#&$#@$ POS mactop reboots because I plugged in a USB hub, I don't lose my place. 19:06
was it bsd tar that we want? 19:16
coleman I forget what caused problems last time. Which file attribute? 19:18
[Coke] the date/time 19:21
I think installing bsdtar on this box and doing an alias tar=bsdtar will work
(also, writing this down)
coleman that seems safe for the actual output. bsd will have a subset of what gnu provides, I'd wager 19:22
if anything fails it might be command line syntax related but you'll see
[Coke] The moarvm instructions need to be sanitized. there's an extra 0 in the version info that is getting copied through a bunch of places. 19:24
need to make sure it doesn't show up in the final upload.
jdv [Coke]: wut? iirc the extra 0 in the version is there for a reason 19:27
[Coke] Really? if you can tell me what it is.. 19:38
It appears to be "to confuse me". :)
It might be "to avoid confusing between tag and branch", but there's probabably a better way to do that. 19:39
Looking at moarvm.org/releases.html I see 2025.02 19:44
I don't see it in my clone. wtf. 19:45
[Coke] git pull was doing nothing, had to reclone. weird. 19:48
[Coke] HEY. Can someone give me rights to moarvm.org please. 19:51
(that's why I couldn't see anything, it was pointed at my fork)
github.com/MoarVM/moarvm.org/pull/22 19:53
coleman I don't have admin on that 19:54
[Coke] I would argue you should have admin.
Wonder if we have to ping jnthn. 19:55
github.com/MoarVM/MoarVM/releases/tag/2025.03 - if someone can verify that the .targz for this release looks less squirrely 19:59
And also please apply my PR for moarvm.org 20:02
I'll go grab a hot bevvie before starting the next big step
going to put 4/26 as the next release, not 4/19 20:16
... ahhhh, no I'm not. 20:17
Geth rakudo/release-2025.03: 7fca580b83 | (Will Coleda)++ | docs/announce/2025.03.md
Update announcement
[Coke] ah, thank goodness there was no moarvm 2023.03 20:22
(or this would have been a very long detour. :)
jdv: running 'ake human-1' step on the azure box and getting an error that it is unable to sign the tag. 20:25
I cannot remember if, two months ago, I was running ake on my mac or on the ubuntu cloud box. 20:26
export GPG_TTY=$(tty) 20:28
t/12-rakuast/xx-fixed-in-rakuast.rakutest (Wstat: 0 Tests: 105 Failed: 0) TODO passed: 82-83 20:43
In general, this release is going smoother than 2025.01 - next one I do will go even smoother if I do the prep work *in advance* 20:47
jdv probably need your php key loaded in the agent 20:52
[Coke] it was the TTY 20:53
jdv *pgp
oh
[Coke] setting the TTY allowed it to prompt me so I could enter the passphrase.
Don't remember that step from last time, but writing it down in a different rakudo checkout for later. :)
jdv weird default but sure
[Coke] on `ake rakudo-test`
Don't want to hardcode it anywhere on a cloud box. 20:54
jdv: next time you do a release, please verify for me if "rakudo-stress" starts out saying Inline::Perl5 isn't installed 21:02
I think it was not the bsdtar perse, but the extra 0 on the branch name that did it. one instruction says do 2025.03 (but is really using 2025.030); and a future instruction tells you to use 2025.03 (but throws an error if you don't use the 2025.030) in moarvm. but that's also the thing that generates the tar.gz file, so you have to rename the file before you upload it. 21:10
`ake rakudo-stress-errata` 21:11
(bsdtar) using the wrong release # there caused a git command to not return a date, so then tar tried to use something NaN as a number and that's what did it. 21:13
I'll do a test release in a week or so to see if I can narrow down the bits (and wonder why this hasn't ever bitten jdv) 21:14
github.com/Raku/roast/issues/872 21:27
Geth: you here/
geth: status
oops: github.com/rakudo/rakudo/issues/5829 21:58
So I think our last spec test run of the installed rakudo has been testing on 6.d errata this whole time, not on the main branch. 21:59
jdv [Coke]: i've never had that extra zero issue 22:15
yes, i did notice that roast branch issue but we test them all on the day of so...meh 22:16
[Coke] m: my $x will init { say 3 } 22:21
camelia ===SORRY!=== Error while compiling <tmp>
Variable trait 'will init {...}' not yet implemented. Sorry.
at <tmp>:1
------> my $x will init { say 3 }<HERE><EOL>
expecting any of:
constraint
[Coke] ... dammit.
I think the errata tests leave around fudged files that preferentially get run. 22:27
which means that even though I tried to switch it back, this test still failed, because it was getting the fudged one from the branch, not the one that lizmat fixed. 22:28
... or something else weird is going on. now getting failures on S04-declarations/will.t even though I'm on master? trying again...
jdv isn't rakudo-post-test on a git-less archive expansion? 22:30
probably need to fix that up before the archive command 22:31
[Coke] Maybe I was looking in the wrong dir 22:32
redoing rakudo-post-build and will then do rakudo-post-test 22:33
... oh. I wonder if the archive has been archiving the errata dir ??
since it never got switched back? I switch it back, then reran the *post-test*, but if it was already archived... 22:35
ok, hopeful this gets us there. never would have noticed if not for this spec file failing in the errata. 22:36
updated github.com/rakudo/rakudo/issues/5829 with my thoughts on the t/spec dir. 22:38
nine, lizmat, jdv: any thoughts on ^^ appreciated - I assume we want to be shipping roast's master in the rakudo tar.gz 22:39
jdv idk if shipping roast was ever a huge consideration 23:04
sure, flip it to master. i'd bet almost nobody uses that copy. 23:06
Geth rakudo: coke++ created pull request #5830:
Release 2025.03		 23:07
rakudo/release-2025.03: 4315de3f06 | (Will Coleda)++ | tools/templates/NQP_REVISION
[release] Bump NQP revision to 2025.03
 23:09
rakudo/release-2025.03: 316ddd21b2 | (Will Coleda)++ | VERSION
[release] Bump VERSION to 2025.03
nqp/main: 55825b7cea | (Will Coleda)++ | tools/templates/MOAR_REVISION
[release] Bump MoarVM revision to 2025.03
nqp/main: bc9658359b | (Will Coleda)++ | VERSION
[release] Bump VERSION to 2025.03
Geth rakudo: coke++ created pull request #5831:
Release 2025.03		 23:17
rakudo/main: 5 commits pushed by (Will Coleda)++ 23:30
[Coke] github.com/rakudo/rakudo/releases/tag/2025.03 23:32
release guide complete; only thing left is binary packages. 23:38
(pinged them in #raku)
jdv as far as i'm concerned "the release" is done:)
[Coke] results this time: much smoother, opened a roast ticket that I did not fix directly for the errata, opened one or two RELENG tickets on rakudo, and have some minor edits pending. 23:39
jdv - if I make any changes to the Ake stuff, I'll mark you as a reviewer on a PR.
want at least one more set of eyes before submitting (same for release guide updates) 23:40
now off to have a sparkling beverage and chill out
jdv i can look but idk if i'd be helpful
nite

Powered by Cro and the Raku® Programming Language.

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

Thank you!