01:13 MasterDuke joined
[Coke] Feels like it slowed down this run is at 42% completion. 01:52
jdv - when you do the blin run pre-release - how many commits get in between your run and the version that's released? Is there a gap where some commits aren't blin tested? 02:37
(and if there are, do you then include the commit after the one tested for the next run so there's a chance to catch issues, even if it's a release late?) 02:39
07:18 lizmat_ joined 07:19 [Coke]_ joined 07:21 [Coke] left, lizmat left 07:50 sena_kun joined
timo i've already got the jitted frames to show up "half"-way 08:07
08:23 sena_kun left 08:39 lizmat_ left, lizmat joined
[Tux] Rakudo v2024.08-33-ga8111db83 (v6.d) on MoarVM 2024.08-8-gf14712b6b
csv-ip5xs0.260 - 0.266
csv-ip5xs-201.106 - 1.114
csv-parser1.481 - 1.526
csv-test-xs-200.142 - 0.143
test1.832 - 1.871
test-t0.401 - 0.407
test-t --race0.263 - 0.270
test-t-204.778 - 4.803
test-t-20 --race1.173 - 1.178
 09:49
tux.nl/Talks/CSV6/speed4-20.html / tux.nl/Talks/CSV6/speed4.html tux.nl/Talks/CSV6/speed.log
10:16 [Coke]_ is now known as [Coke]
[Coke] github.com/Raku/doc/pull/4513#issu...2328410128 - is the Test module considered part of the language? (My take was yes, JJ seems to be saying no here) 10:17
lizmat I'd say, since roast depends on Test, it is ? 10:21
[Coke] OK. 10:24
timo it looks like IO::Handle.seek is not spec-tested to have a default value for $whence, so the docs are correct in not showing that it has a default value ... but i kind of like having SeekFromBeginning as a default value for seek :| 10:27
bisectable6: say IO::Handle.^lookup("seek").candidates.map(*.signature.params[2].optional) 10:30
bisectable6 timo, Will bisect the whole range automagically because no endpoints were provided, hang tight
timo, ¦6c (83 commits): «(True)␤» 10:31
timo, Nothing to bisect!
Geth ¦ roast: coke self-assigned subtest variants github.com/Raku/roast/issues/858 10:50
¦ roast: coke self-unassigned subtest variants github.com/Raku/roast/issues/858
¦ roast: JJ self-assigned subtest variants github.com/Raku/roast/issues/858 10:53
rakudo/main: a260251b4d | (Elizabeth Mattijsen)++ | lib/RakuAST/Deparse/Highlight.rakumod
RakuAST: prevent unnecessary comment highlighting
 11:08
[Coke] Raku was removed from debian testing, the last version there is 2022.12 which has a problem compiling with gcc14 (per perl6-users email). Need a volunteer to take over packaging. Dominique Dumont notes that if it's not fixed, it will eventually be removed from unstable, at which point getting it back into Debian will be more difficult. 11:09
packages.debian.org/experimental/rakudo - this says 2024.02 is available, so I'm confused. 11:15
11:19 [Tux] left
timo qa.debian.org/excuses.php?package=moarvm 11:20
looks like being in experimental is not good enough if a package can't make it to testing
11:24 [Tux] joined
[Coke] how did you find that? 11:29
timo tracker.debian.org/pkg/moarvm via this page 11:30
on the right hand sidebar there's a link labelled "developer information"
packages.debian.org/experimental/moarvm-dev from here
[Coke] wow it's so old. 11:32
the "stable" version. 11:33
timo: I tried to login to salsa.debian.org using my gitlab account, login to gitlab, salsa tells me I can't sign in using a gitlab account without a pre-existing gitlab account. :? 11:36
timo i don't know anything about salsa. what path did you take to get to a gitlab login? 11:38
[Coke] tried to sign in one of the options was google, bitbucket, gitlab.com 11:39
(for SSO)
I'm guessing they mean you need a *local* gitlab account (because it's an external gitlab instance) but the language is confusing. :) 11:40
(but then why even have the link option...)
timo "Waiting for GitLab to boot" coke what did you do 11:41
coke you can't just nuke the debian infrastructure whenever their rakudo packages are too old, please
i understand your frustration, and i applaud the commitment, but i'm afraid it's a bit much ... 11:42
Geth roast/✅-subtest-pair: f730aad289 | (JJ Merelo)++ | S24-testing/12-subtest-todo.t
closes #858 :white_check_mark:
[Coke] If I can take down your VCS by logging into it..
timo in any case, github.com/MoarVM/MoarVM/commit/4c...5513.patch is a patch that applies directly on top of moarvm 2024.02, which is the package that's currently in experimental. do you want to try to figure out how to add it to the debian patch series/quilt? 11:43
Geth roast: JJ++ created pull request #859:
Adds tests for both versions of Pair		 11:44
timo oh hey it's back up
do i have to wait for the app to warm up or something, or is my tethered internet connection too slow to go to the ... never mind 11:46
ok yeah maybe it's actually clicking the "sign in with gitlab.com" that causes it to go down??? because i tried to click that and immediately got "waitin for gitlab to boot" again 11:47
[Coke] timo: it's back down again. 11:48
wow.
timo i hope it wasn't me and there's just something else going on that's more serious
i guess we'll know if they remove that "sign in with gitlab.com" button from the login page soon
[Coke] *us*. :)
timo there's no like status.debian.org that shows information about outages of their web services or anything 11:50
(i think there's also nothing like that for raku?)
there are pseudo-packages for a bunch of things in debian that are meant to receive reports for problems with debian infrastructure, but there doesn't seem to be one for salsa yet, at least not on this list i'm looking at right now 11:52
[Coke] I just clicked "sign in" just got the 502 again 11:59
Didn't eve make it to the login page
timo "[Yes, salsa is experiencing performance issues - and yes we know it sucks]"
the topic of #salsa on OFTC
[Coke] did you say anything in the channel yet? 12:04
ah, there you are
Bisecting Audio::PortMIDI - that's new 12:17
... which then emitted a git error and an "uh oh" 12:18
Geth rakudo/main: 07c8b1193b | (Elizabeth Mattijsen)++ | lib/RakuAST/Deparse/Highlight.rakumod
RakuAST: Remove support for markup in syntax highlighting

It's impossible to implement in a generic fashion. See
   github.com/Raku/RakuDoc-GAMMA/issues/47		 12:21
rakudo/main: d0a702818c | (Elizabeth Mattijsen)++ | t/12-rakuast/highlight.rakutest
RakuAST: add another "use" test in highlighting
lizmat </headbanging> 12:23
[Coke] ab5tract: github.com/Raku/Blin/issues/35 12:25
to answer your question, lizmat, I don't think there's a general way to do it given that it's not valid code - I don't think it's worth doing partial support - extra work for something that will still not work in some cases? meh 12:27
lizmat yeah, took me over a week to reach that coclusion :-( 12:28
*conclusion :-) 12:32
ab5tract [Coke]: I don’t recognize that code. M I in the blame history or something? 12:48
[Coke] no, you just have pending patches in Blin so I thought maybe you might want that one. :) 12:59
I hestitate to muck with anything while I'm 56% of the way through a run. :)
looks like that's from "initial commit". :) 13:06
ab5tract Ah, gotcha :) 13:21
timo coke, did you get some good advice on the debian-mentors channel yet? i only now joined so there's no backlog form e 13:58
[Coke]: do you get highlighted on just "coke" as well? 14:02
jdv [Coke]: there are smaller commits that never get blin'ed. i typically do a blin run beginning of the release week so its only a few days worth and big stuff is held back during that period. 14:04
[Coke] I only joined the channel, didn't get further. 14:05
timo: try to highlight me again with just coke? 14:06
jdv a clean blin run takes ~8h for me. if anyone else wants to run it more have at it:)
timo does the coke see this? 14:07
[Coke] Just saw it as a message, not to me.
timo coke, and when put at the start like this?
[Coke] jdv: I'm at about 24 hours so far, and am shy of 60%
timo: Sadly I was already looking at it when it showed up, so I didn't get the irrsi title bar anyway. 14:08
here, I'll close my eyes and you can try again. :)
timo coke, okay 14:09
coleman id like to buy the world a coke 14:35
[Coke] yes, that worked. 14:41
timo cool cool. 14:43
[Coke] Looks like the main throttle here is CPU. I have a 4-cpu box pegged, using <8GB of a 16G box. Can probably find a better azure VM that has more cores and not much more memory.
timo i also signed up for salsa now so i can see what you see and such
[Coke]: did you get an email confirmation of some kind right after you signed up, or only one after the verification has gone through? 14:52
ok i got verified 14:59
but there seems to not be any projects with raku or rakudo or moarvm or perl6 in their name that the search finds. are we not at the right address with salsa? 15:01
do we want to create something like "raku-team"? 15:02
oh, when you said the moarvm version was ancient, did you maybe see the only version that's available for the x32 arch? that's 2017.06, but for the other archs there's 2022.12 15:04
[Coke] 2022.12 is still old. 15:05
timo: there's a perl6 team or some such, yes? Let's join that and rename it. 15:06
15:08 MasterDuke left
ugexe www.mail-archive.com/perl6-all@per...13816.html - maybe not useful or is already known, but here is the message they left when they stepped down from packaging rakudo for debian 15:12
timo ah, perl6-team, right: salsa.debian.org/perl6-team/ 15:13
[Coke] Thanks, ugexe. 15:21
timo [Coke]: do you want to contact Dominique about giving your salsa user and mine ("timotimo") access to the perl6-team and whatever we need to have access to in order to move forward? 15:28
[Coke] I 15:29
am dayjobbing at the moment, will send an email later this afternoon if you don't beat me to it. 15:30
Geth roast/✅-subtest-pair: e519d2ca1d | (JJ Merelo)++ | S24-testing/12-subtest-todo.t
Fix number of tests

A bit of DRY, too		 15:34
timo just to make sure, your user on salsa is "coke"? it does have your name, but what if someone swooped in and stole it
... how do i even know it's actually you on irc here? 15:35
[Coke] AHAHAHAHA 15:36
my TWENTY YEAR PLAN HA.... ahem. yes it's me, of course it is.
Yes, I just registered coke on salsa today 15:37
timo OK I sent a request, publically on the mailing list on the thread. I hope I didn't overreach by speaking for you like that 15:50
maybe there's time for a nap ... 16:07
[Coke] timo: that was fine. 16:41
we need to be "-guest" in the gitlab repo?
(getting 502s again) 16:46
jdv- when you run blin do you use nproc-multiplier? 18:44
I've been running with a multiplier of 1.5 (wonder if that's too high) 18:45
jdv iirc i run nproc=12 (got 16 cores) 18:48
[Coke] ok, so mine sounds too high. 18:49
jdv it could be;)
timo i'm very slowly moving towards the account rename feature 18:55
19:24 coverable6 left, coverable6 joined 19:30 rypervenche left 19:32 rypervenche joined
timo i looked up why they call it salsa: gobby.debian.org/export/Sprints/Al...17/Minutes "(salsa--where we keep our secret sauce?)" 20:02
21:45 finanalyst joined 22:07 finanalyst left 22:18 sena_kun joined 22:34 Altai-man joined 22:38 sena_kun left 23:41 Altai-man left

Powered by Cro and the Raku® Programming Language.

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

Thank you!