00:26 [Coke]_ joined 00:29 El_Che left, RakuIRCLogger left, [Coke] left, El_Che joined 00:44 [Coke]_ is now known as [Coke] 01:05 hulk joined 01:06 kylese left 01:29 simcop2387 left 01:39 simcop2387 joined
[Coke] anyone know how to do rsync over ssh but specify a specific key like with scp -i ? 01:46
ah, rsync -e 01:47
02:15 hulk left, kylese joined 04:52 wayland76 joined
wayland76 Hi all! I've had an idea. We should look at having a bug verification bot. Its job is to attempt to reproduce the bug that's submitted. My idea was that it could be run across all the old bugs, and try to answer the questions 1. Can we reproduce the bug on the version of Raku that was either mentioned in the ticket, or current at the time of ticket submission, (if not, hit up the user for more info), and 2. Can we reproduce the bug with the current 04:55
raku (if not, then tell the user it'll be resolved in 2 weeks if they don't follow up), and optionally, produce a report on what bugs should be closed our (or just do it itself). That would help keep the queue under control.
Anyway, my instinct was to put this under problem-solving, but I'm proposing a solution, not a problem. 04:56
05:07 leppard left
xinming lizmat: I think with yyyy-mm, We can have mm-dd as short cut too. :-) 05:31
wayland76 Is there a way to override the ordering of multiple dispatch? 05:42
06:26 abraxxa joined
Voldenet m: sub datefmt($d, $_) {S:g/(y|m|d)$0*/{ {d=>$d.day,m=>$d.month,y=>$d.year.substr(*-$/.chars)}.{$0}.fmt(<%0>~$/.chars~ <d>) }/}; class F { has $.d; method FALLBACK($f){datefmt($!d, $f)}}; sub f { F.new: :$^d }; given Date.today.earlier(:10days).&f { .yyy-m.say; .yy-mm-dd.say; .yyyy-m-d.say } 06:29
camelia 026-6
26-06-08
2026-6-8
Voldenet i'm not very happy about .substr(*-$/.chars) because it doesn't handle yyyy
erm, yyyyy
m: sub datefmt($d, $_) {S:g/(y|m|d)$0*/{ {d=>$d.day,m=>$d.month,y=>$d.year.substr(*-$/.chars)}.{$0}.fmt(<%0>~$/.chars~ <d>) }/}; class F { has $.d; method FALLBACK($f){datefmt($!d, $f)}}; sub f { F.new: :$^d }; given Date.today.earlier(:10days).&f { .yyy-m.say; .yy-mm-dd.say; .yyyy-m-d.say, .yyyyy.say }
camelia 026-6
Directive d not applicable for value of type Failure
in code at <tmp> line 1
in sub datefmt at <tmp> line 1
in method FALLBACK at <tmp> line 1
in block <unit> at <tmp> line 1

26-06-08
2026-6-8
Voldenet m: sub datefmt($d, $_) {S:g/(y|m|d)$0*/{ {d=>$d.day,m=>$d.month,y=>$d.year % 10 ** $/.chars}.{$0}.fmt(<%0>~$/.chars~ <d>) }/}; class F { has $.d; method FALLBACK($f){datefmt($!d, $f)}}; sub f { F.new: :$^d }; given Date.today.earlier(:10days).&f { .yyy-m.say; .yy-mm-dd.say; .yyyy-m-d.say, .yyyyy.say } 06:35
camelia 026-6
26-06-08
2026-6-8
02026
Voldenet m: sub datefmt($d, $_) {S:g/(y|m|d)$0*/{ {d=>$d.day,m=>$d.month,y=>$d.year % 10 ** $/.chars}.{$0}.fmt(<%0>~$/.chars~ <d>) }/}; class F { has $.d; method FALLBACK($f){datefmt($!d, $f)}}; sub f { F.new: :$^d }; given Date.today.earlier(:10days).&f { .y-yy-yyy-yyyy-yyyyy-yyyyyy.say }
06:35 abraxxa left
camelia 6-26-026-2026-02026-002026 06:35
06:37 abraxxa joined
Voldenet bulk of the problem was actually implementing datefmt :> 06:38
07:19 Sgeo left 07:37 wayland76 left 07:38 wayland joined 07:59 dakkar joined 08:12 leppard joined
timo wayland: melezhik has something like that 08:16
08:25 jdv left 08:26 jdv joined
timo he harvested code examples and the expected and/or wrong output from issues on github and turned those into runnable tests 08:56
08:57 silug joined
wayland Oh! Lovely! Good work by melezhik! Thanks timo. 10:07
timo can always have more bots though (but no LLM please and thank you)
speaking of bots, does someone want to review my changes to IRC::Client where I add support for a couple of IRCv3 features? 10:13
wayland Hmm. But actually, that's different than what I suggested, though possibly even more useful. I was more coming at it from the POV of getting the 800+-bug queue a bit more under control. 10:14
disbot4 <melezhik.> o/ 10:59
<melezhik.> rockylinux.org/news/2026-06-08-get...-needs-you
<melezhik.> There is sparky mentioned here, if someone wish to contribute , here is the way 11:00
wayland librasteve / lizmat : I think you'll both be interested in looking at the intro to the article melezhik linked (before the first heading), because of what their survey found. HTH, 11:06
ugeze: Should I move github.com/rakudo/rakudo/issues/6288 to the zef repo? 11:07
11:20 apogee_ntv joined
tbrowder speaking of the zef repo, the other day @ugexe mentioned a possible cannonical order for auth, ver, and api. is establishing that a future possibility? 11:27
11:29 lucerne90 joined 11:50 dakkar left 11:51 dakkar joined
lizmat tbrowder: do you know of raku.land/zef:lizmat/Identity::Utils#sanitize ? 11:55
12:40 dakkar left 12:41 dakkar joined
tbrowder ah, i vaguely remember it, thnx! 12:56
and it answers ALL MY 12:58
*my questions 12:59
13:01 vrurg_ joined 13:03 vrurg left
apogee_ntv Released Selkie 0.9.0, has some QOL updates and a Selkie::Trace component that lets you build chrome trace reports. 13:27
disbot4 <librasteve> wayland: that’s an awesome point - I agree - can we condense that to a panel on raku.org or maybe a full new page? 13:36
dakkar uh, looks like the logs are not logging… irclogs.raku.org/raku/live.html is stuck to 16 hours ago 13:40
tbrowder ref canonical order: zef lists —installed in ver, auth, and api. that looks like de facto canonical order to me 14:30
16:29 dakkar left
lizmat PSA: if you're here because you're interested in the Raku Programming Language and its community, please register yourself: raku.foundation 17:34
It will help us reach you!
17:34 abraxxa left 18:07 leppard left 18:20 belluzj joined 18:38 belluzj left 18:40 Sgeo joined 18:41 belluzj joined 19:05 belluzj left 19:49 sgros joined 20:10 leppard joined 20:34 sgros is now known as sgrs
wayland lizmat the Raku IRC log thingy isn't recording anything. Is it me, or is it down again? 20:40
librasteve: If you mean my point about the survey, then I should add that I started developing a whole separate site just dedicated to helping people get involved in the community, but never quite got it up and running. I was using the Elucid8 framework, and it was very very new at the time, and my questions didn't get answered then. I may at some point soon at least be able to get it visible somewhere. 20:45
lizmat: Registered, but a warning that I want to keep myself focussed on my stack of projects (starting with Tree-Oriented Programming and Table-Oriented programming). 20:47
20:48 RakuIRCLogger joined
lizmat wayland: thanks for the nudge, kicked it 20:48
21:11 kaskal left
wayland In case my question got buried in the backlog, let me reiterate it: Is there a way to override the ordering of multiple dispatch? Like, I'd like to provide my own ordering that would then be used by things like `nextsame`? 21:41
tbrowder by my simple tests i could use zef install to load a distro by name, or name and ver, or name and ver and auth, or name and ver and auth and api. requests out of that order failed. 23:01
timo wayland: the logic for multiple dispatch is implemented in nqp or raku code, so yes and no, because i don't think you can just put a new dispatcher of the same name in to replace the existing one 23:14
actually, you can write the body of the proto yourself, so you can just call a different dispatcher instead of the default one 23:31