🦋 Welcome to the MAIN() IRC channel of the Raku Programming Language (raku.org). Log available at irclogs.raku.org/raku/live.html . If you're a beginner, you can also check out the #raku-beginner channel!
Set by lizmat on 6 September 2022.
00:21 librasteve left 00:33 librasteve joined 00:38 librasteve left 01:00 librasteve joined 01:05 librasteve left 01:32 librasteve joined 01:33 kylese left, kylese joined 01:37 librasteve left 02:05 librasteve joined 02:10 librasteve left 02:15 kylese left, kylese joined 02:24 grondilu left, grondilu joined 02:40 librasteve joined 02:44 librasteve left 02:56 librasteve joined 03:18 stanrifkin_ joined 03:20 stanrifkin left 03:45 kylese left 03:48 kylese joined 05:06 librasteve left, yewscion joined 05:08 yewscion left 05:15 Aedil joined 05:18 librasteve joined 05:23 librasteve left 05:24 librasteve joined 05:33 librasteve left 05:47 librasteve joined 05:48 soverysour joined 06:00 Sgeo left, soverysour left 06:29 dg left 06:38 dg joined 06:48 soverysour joined 06:53 librasteve left, soverysour left 06:57 librasteve joined 07:03 librasteve left 07:16 admin___ joined 07:17 admin___ left 07:22 holyghost joined 07:25 librasteve joined 07:29 librasteve left 07:34 donaldh- left 07:39 soverysour joined 07:50 librasteve joined 08:36 dakkar joined 08:52 soverysour left 08:55 soverysour joined 08:56 librasteve left 08:59 soverysour left 09:06 librasteve joined
lizmat weekly: dev.to/patrickbkr/the-second-raku-...summit-3d2 09:07
notable6 lizmat, Noted! (weekly)
09:11 derpydoo joined 09:15 soverysour joined 09:25 soverysour left 09:26 wayland76 joined 09:27 derpydoo left 09:34 soverysour joined 09:40 donaldh joined 09:53 lichtkind joined 10:05 soverysour left 10:14 librasteve left 10:26 librasteve joined
erry youtu.be/6QSiEuByC1Y?t=140 handwritten japanese is valid raku 10:31
lizmat weekly: youtu.be/6QSiEuByC1Y?t=140 handwritten japanese is valid raku 10:33
notable6 lizmat, Noted! (weekly)
erry what weekly? 10:34
10:34 librasteve left
lizmat erry: rakudoweekly.blog/blog-feed/ 10:34
erry I should note by the waym i used github.com/finanalyst/rakuast-L10N...me-ov-file to run the actual 'raku' code, but python to do ocr and feed it to raku
it's a polyglot endeavour 10:35
lizmat: great,i'm glad to spread this totally normal code to the world
timo yes, it's perfect
lizmat yeah: am overhauling all localizations at the moment, in a new organization:
github.com/Raku-L10N/ 10:36
erry nice :D 10:37
10:38 derpydoo joined
erry lizmat: maybe you want to remove the ?t= 10:40
it starts the video from that timestamp
lizmat erry: I thought that was the point ? but yeah, will incorporate the ?t= URL in the next weekly 10:44
10:46 librasteve joined 10:52 librasteve left
timo well, posting for IRC and posting for the weekly may have different intended audiences right? 10:58
11:03 derpydoo left
lizmat well, yes: posting on IRC is immediate, and posting with weekly: is both immediate and a good chance of being mentioned in the next weekly 11:03
erry oki 11:04
:3
timo right. and you took the message from erry 1:1 and put it into "weekly:", that's why erry pointed out the timecode may not be right to have in the weekly as well
i'm not sure if we're actually in disagreement about anything 11:05
11:05 librasteve joined
lizmat I don't think we are? 11:05
timo cool
11:09 librasteve left 11:25 soverysour joined 11:26 simcop2387 left, librasteve joined 11:27 simcop2387 joined 11:31 librasteve left 11:33 soverysour left 11:35 jjido joined 11:36 simcop2387 left, simcop2387 joined 11:45 librasteve joined 11:49 librasteve left 11:50 librasteve joined, jjido left 11:51 soverysour joined 11:54 librasteve left 12:07 librasteve joined 12:19 tjr joined 12:21 librasteve left 12:27 librasteve joined 12:35 yewscion joined 12:40 yewscion left 12:41 soverysour left 12:46 soverysour joined, soverysour left, soverysour joined 13:13 soverysour left 13:18 soverysour joined 13:29 jjido joined 13:39 librasteve left 13:53 soverysour left 13:55 jjido left 14:02 librasteve joined 14:14 yewscion joined
[Coke] Team - please help me to nominate a list of "essential" modules that a) must succeed on blin runs, and b) might be part of more frequent blin testing. 14:41
jdv we already basically do that 14:42
[Coke] (knowing that some of these have a lot of deps, which are then implicitly covered - ok to list the deps explictly as well)
jdv if its a well known dist it pretty much blocks the release
14:42 soverysour joined, soverysour left, soverysour joined
[Coke] github.com/Raku/Blin/wiki/Essential-Modules 14:44
jdv i'd say all the cro dists, zef, DBIish, DB::Pg, the handful of ssl dists at least
[Coke] 'Cro' by itself would pull in all the cro stuff, yes? 14:46
disbot2 <librasteve> yes
<librasteve> but note its lowercase cro ... also the usual advice is to go --/test since the test time is quite long 14:47
jdv well, the point of testing is to...test:) 14:51
disbot2 <librasteve> amen
14:54 maylay left 15:00 maylay joined 15:09 stanrifkin_ left
disbot2 <librasteve> ===SORRY!=== Error while compiling /Users/stephenroe/Library/CloudStorage/Dropbox/GitWorld/Air-Play/lib/Air/Play/Site14-CounterRed.rakumod (Air::Play::Site14-CounterRed) Incompatible MROs in P6opaque rebless for types Perl6::Metamodel::ClassHOW+{<anon>}+{<anon>} and Perl6::Metamodel::ClassHOW+{<anon>} 15:13
<librasteve> I am use'ing Red:api<2>; red-defaults “SQLite”; and use experimental :rakuast; use RakuAST::Deparse::Highlight; on the other and I wonder if this error is because Red is somehow incompatible with RakuAST (at least maybe I need to call experimental somehow with the Red code...? 15:15
SmokeMachine I wasn't yet able to run Red using RakuAST... That's why I haven"t started using RakeAST on Red... 15:17
disbot2 <librasteve> ah - thanks for the info - looks like I am going to be one of the first "testers" ... meantime I will somehow avoid using both database and code highlighting ;-) 15:20
<librasteve> the joys of life on the (bleeding) edge 15:21
15:29 soverysour left 15:33 soverysour joined, soverysour left, soverysour joined 15:44 soverysour left 15:50 soverysour joined
SmokeMachine librasteve: but what do you want to do with RakuAST::Deparse::Highlight and Red? 15:55
disbot2 <librasteve> I want HARC to work with Red (like in the SearchTable example) and I want HARC to do raku code highlighting 15:56
<librasteve> so I am trying to bring in the Rakudoc Hilite code that finanalyst made that uses Rainbow that uses RakuAST::Deparse::Highlight 15:57
SmokeMachine but do you need 'use experimental :rakuast' to work with code highlighting? I suppose that's only neded to make the compiler to use RakuAST, no? 15:58
lizmat Rainbow uses ::Deparse::Highlight?
SmokeMachine but does it need 'use experimental :rakuast'? I would guess it doesn't... but I might be wrong...
disbot2 <librasteve> it seems that the MRO error is thrown on the use RakuAST::Deparse::Highlight;
<librasteve> hang on a sec 15:59
<librasteve> Hilite / Rainbow seems to work fine without the use experimental :rakuast; 16:00
<librasteve> but the MRO error persists ... so that is definitely caused by combining use Red; and use RakuAST::Deparse::Highlight; 16:01
SmokeMachine Red uses a lot of meta stuff... are you trying to highlight code using Red or are you highlighting some code and using Red only for storing stuff? 16:02
disbot2 <librasteve> i'm going to just reject HARC sites that try to use Red (ie for site data) and Hilite by conditional import ... but that is sub optimal in the medium term 16:03
16:04 soverysour left
SmokeMachine librasteve: usercontent.irccloud-cdn.com/file/.../image.png 16:15
16:19 gabiruh_ left 16:20 gabiruh joined
disbot2 <librasteve> appreciate the help - sadly that's not going to work in this case since the Rakudoc Hilite.rakumod gives errors unless I use RakuAST::Deparse::Highlight; and I do not want to have to rewrite that 16:20
<librasteve> suggest I punt on this issue until Hilite is released as a stand alone module (I have made some changes to support this) --- then you are welcome to go in and fix that ... but my preference would be to get Red to work wirh RakuAST ;-) 16:22
Geth ¦ problem-solving: lizmat self-assigned Towards a Raku Foundation github.com/Raku/problem-solving/issues/477 16:26
SmokeMachine I really think that once Red works without RakuAST, the correct would be to make RakeAST work with Red...
16:26 soverysour joined
SmokeMachine make sense? 16:27
I mean Red works... if RakuAST intends to work transparently, that should work with Red... 16:28
disbot2 <librasteve> errr nope ... I think the current status is "the core team is working on RakuAST and that has increasing improved coverage on the raky test suite - therefore when RakuAST is passing all core tests, the next challenge for the team will be to make RakuAST work with important modules such as Red"
<librasteve> yeah - I agree that Red should just work with RakuAST when that works is at a sufficiently high quality 16:29
16:34 wayland joined, wayland76 left
SmokeMachine interesting: usercontent.irccloud-cdn.com/file/.../image.png 16:35
disbot2 <librasteve> oh - I just remembered that Hilite is going to be a combination of something like Air::Plugin::Hilite [that uses a standalone raku module called Hilite] when this stabilizes ... so you will have to go use Air::Plugin::Hilite in the preamble - so in the short term the guidance will be to not do that if you want to use Red
lizmat weekly: dev.to/lizmat/towards-a-raku-foundation-3ne2 16:36
notable6 lizmat, Noted! (weekly)
disbot2 <librasteve> appreciate your help
SmokeMachine it sees it as a package...
16:39 soverysour left 16:41 dakkar left 17:04 librasteve left 17:06 soverysour joined 17:13 soverysour left 17:20 librasteve joined 17:28 soverysour joined, soverysour left, soverysour joined 17:35 soverysour left
El_Che hallo 17:39
tellable6 2025-05-17T16:41:44Z #raku-dev <jdv> El_Che 2025.05 happened
El_Che damn, I haven't been on irc much
17:49 holyghost left 18:47 librasteve left 18:51 soverysour joined 18:57 librasteve joined 19:12 stanrifkin joined 19:16 librasteve left 19:32 nine left, nine joined 19:33 librasteve joined 19:38 librasteve left 19:50 soverysour left 19:51 apac joined, librasteve joined 19:57 ACfromTX left 20:01 jjido joined 20:08 apac left 20:31 stanrifkin left 20:46 librasteve left 20:55 apac joined 20:59 librasteve joined 21:01 yewscion_ joined 21:02 yewscion left 21:04 librasteve left 21:05 stanrifkin joined 21:07 yewscion_ left 21:16 librasteve joined
timo weekly: wakelift.de/2025/06/18/re-emergenc...-the-blog/ 21:17
notable6 timo, Noted! (weekly)
21:21 librasteve left
[Coke] timo++ 21:22
disbot2 <DataKinds> Hey all, is there a page where performance regression testing for Rakudo is posted? Currently trying to compare a program's runtime across compiler versions to some baseline 21:27
21:32 Aedil left
timo no, there's a package that i've used a rather long time ago that automatically built different rakudo versions and ran a little suite of microbenchmarks and one or two not-so-micro benchmarks and generated html pages with javascript graphs, but i haven't used it in a long long time 21:32
disbot2 <DataKinds> Ahh that's okay. I can compare a few versions manually, thanks 21:33
timo we do have the benchable6 bot here on irc which can look for performance regressions by bisecting rakudo versions against code you give it
disbot2 <DataKinds> how does the bot work?
timo and sourceable6 has download links for every rakudo commit
benchable6: help
benchable6 timo, Like this: benchable6: f583f22,HEAD my $a = ‘a’ x 2¹⁶; for ^1000 {my $b = $a.chop($_)} # See wiki for more examples: github.com/Raku/whateverable/wiki/Benchable
timo are you looking for anything in particular? did you find something that seems to have become slower? 21:35
disbot2 <DataKinds> I found a microbenchmark someone else wrote that is slower than I expected it to be & am wondering if it was a regression. Currently rewriting it to one line to run through the bot 21:37
timo i think the bot can take gist links, too 21:44
i've got a lot of experience looking at performance stuff so maybe i can figure out what's wrong 21:45
21:51 librasteve joined 21:57 librasteve left, jjido left, apac left 22:02 yewscion joined
timo but first, some amount of sleep 22:05
Voldenet I read about missing docs about PromiseStatus and I think that this is not really too comfortable to use in a lot of cases 22:06
m: say Promise.kept(42).status ~~ PromiseStatus::Kept
camelia True
[Coke] ?
Voldenet this is quite lengthy 22:07
m: say Promise.kept(2).kept # this could exist
camelia Cannot resolve caller kept(Promise:D: ); none of these signatures matches:
(Promise:U $:: *%_)
(Promise:U $:: Mu \result, *%_)
in block <unit> at <tmp> line 1
[Coke] m: say Promise.kept(42).status ~~ Kept
camelia True
Voldenet sure, but it's still not remotely as nice as single method for this 22:08
> .kept .broken .completed
22:09 lichtkind left
[Coke] .kept is already in use for saying it's kept. 22:09
Voldenet but only for Promise:U 22:10
[Coke] are you saying you want a single method that does kept, broken or complete?
Voldenet well, three helper methods
.then is rarely used, but not-so-nice syntax makes it even more obscure 22:11
m: say Promise.kept(2).then( -> $p { $p.status ~~ Kept ?? "Yes" !! "No" }).&await 22:13
camelia Yes
Voldenet m: use MONKEY-TYPING; augment class Promise { multi method kept(Promise:D:) { self.status ~~ Kept }}; say Promise.kept(2).then( -> $p { $p.kept ?? "Yes" !! "No" }).&await 22:15
camelia Yes
Voldenet I know that improvements I promise are not major and add some maintenance cost 22:16
s/promise/propose/ ┐(´~`;)┌
that's why I'm not writing github issues with these, I only feel that promises lack the brevity I'm used to in raku 22:19
22:19 librasteve joined, yewscion_ joined
Voldenet and I certainly don't want to turn it into project reactor, which has obnoxious number of methods for Mono<T> 22:21
22:22 yewscion left
Voldenet for reference: projectreactor.io/docs/core/releas.../Mono.html 22:22
22:23 Sgeo joined 22:26 librasteve left
Voldenet I stand corrected, `.completed` already exists as `.Bool` 22:26
lizmat m: say Promise.kept(42).status ~~ Kept # shorter 22:27
camelia True
lizmat ah, already said
Voldenet though maybe the way forward is actually spectesting andthen/orelse pair, not adding more code 22:29
m: say Promise.kept(2).andthen({"Yes"}).orelse({"No"}).&await # this already works 22:30
camelia Yes
Voldenet but first I'd have to work on MR for github.com/rakudo/rakudo/issues/5892 22:31
22:41 librasteve joined 22:42 yewscion_ left
lizmat weekly: wakelift.de/2025/06/18/re-emergenc...-the-blog/ 22:48
notable6 lizmat, Noted! (weekly)
23:12 wayland left