00:03 arkiuat joined 00:07 arkiuat left 00:21 arkiuat joined 00:30 arkiuat left 00:31 arkiuat joined 00:36 arkiuat left 00:39 arkiuat joined 01:30 refactus left 02:04 Aedil joined 03:06 stanrifkin_ joined 03:08 stanrifkin left 04:47 arkiuat left 04:55 refactus joined 04:57 xinming left, arkiuat joined 04:58 xinming joined 05:05 arkiuat left 05:14 kylese joined 05:16 derpydoo joined 05:29 apa_c joined 05:34 arkiuat joined 05:38 arkiuat left 05:51 arkiuat joined 05:55 arkiuat left 06:07 arkiuat joined 06:12 arkiuat left 06:16 derpydoo left 06:27 Sgeo left 06:29 arkiuat joined 06:34 arkiuat left 06:47 arkiuat joined 06:51 arkiuat left 06:52 sibl joined 06:54 derpydoo joined 07:19 arkiuat joined 07:24 arkiuat left
disbot3 <librasteve> overriding () as a custom circumfix seems like a good use case for RakuAST, but if you override them generally then you erase all grouping and much subroutine calling, so you likely need an additional way to pick out just the right situation … 07:39
07:40 apa_c left 07:47 arkiuat joined 07:52 arkiuat left 08:15 derpydoo left 08:17 abraxxa1 joined 08:18 abraxxa left 08:22 arkiuat joined 08:27 arkiuat left 08:42 Aedil left 08:46 MoC joined 08:53 arkiuat joined, melezhik joined
melezhik . 08:54
08:58 arkiuat left 09:26 arkiuat joined 09:27 derpydoo joined 09:31 arkiuat left 09:32 arkiuat joined 09:34 kybr left 09:36 arkiuat left 10:04 sibl left 10:05 arkiuat joined 10:10 arkiuat left 10:33 arkiuat joined 10:37 arkiuat left 10:47 librasteve_ joined 10:50 arkiuat joined 10:54 arkiuat left 11:01 derpydoo left
stanrifkin_ nice new website btw 11:12
11:16 apa_c joined 11:25 arkiuat joined 11:30 arkiuat left 11:34 abraxxa1 left 11:36 arkiuat joined 11:41 arkiuat left 12:10 arkiuat joined 12:14 arkiuat left 12:35 arkiuat joined 12:39 arkiuat left 12:49 arkiuat joined 13:15 vasko45 joined 13:17 hulk joined 13:18 kylese left, vasko4 left, vasko45 is now known as vasko4, silug left, simcop2387 left 13:19 simcop2387 joined 13:21 silug4 joined 13:33 melezhik left
disbot3 <librasteve> all flattery very welcome! 13:54
14:05 Sgeo joined 14:12 gabiruh joined 14:36 apa_c left 15:04 human-blip left 15:06 human-blip joined 15:08 apa_c joined 15:10 Aedil joined 15:21 apa_c left 15:28 apa_c joined 15:49 arkiuat left 15:57 librasteve_ left 16:01 arkiuat joined 16:06 arkiuat left 16:09 human-blip left 16:11 human-blip joined 16:14 apa_c left 16:15 apa_c joined 16:21 apa_c left 16:36 arkiuat joined 16:41 arkiuat left 16:47 abraxxa-home joined 16:49 abraxxa-home left 16:50 arkiuat joined 16:51 apa_c joined 16:52 abraxxa-home joined 16:58 LainIwakura joined 17:01 LainIwakura38 joined 17:04 LainIwakura left 17:06 LainIwakura38 left 17:07 abraxxa-home left
[Coke] m: dd Nil, Nil.defined, Nil.CREATE, Nil.CREATE.defined 17:08
camelia Nil = Nil
Bool::False
Nil = Nil
Bool::True
[Coke] ... dd can't seem to tell that that result of Nil.CREATE is not a "normal" Nil. 17:09
m: my $a = Nil; my $b = Nil.CREATE; dd $a, $b; # though this shows it 17:10
camelia $a = Any
$b = Nil
disbot3 <neekotism> How can I achieve this? interval = event["d"]["heartbeat_interval"] / 1000 asyncio.create_task(heartbeat(ws, interval, last_sequence)) 17:24
<librasteve> definitely need more $ signs 17:26
<librasteve> seriously, my judgement is that you are an AI --- please show me some signs of humanity or I will alert the mods
<neekotism> I'm not! 17:27
<librasteve> lol - sorry
<librasteve> hi folks - I am getting a thorny error No appropriate parametric role variant available for 'Defaults': when I try to compose role Defaults into class Site (in Air) --- any ideas? could this be a bug? [happens on both moar-2025.06.1 and moar-2024.08] 17:29
<librasteve> the role composes fine to a couple of other classes in my code, btw 17:30
<librasteve> my $interval = event<d><heartbeat_interval> / 1000; start { heartbeat($ws, $interval, $last-sequence); } 17:31
<librasteve> ChatGPT and other LLMs speak pretty good raku ... this code is untested btw
[Coke] librasteve - do you have a branch we can look at that shows the issue? 17:32
disbot3 <neekotism> Alrighty! Looks like I don't have much option 17:33
<librasteve> github.com/librasteve/Air/blob/1ad...kumod#L892
<librasteve> this in github.com/librasteve/Air/tree/defaults 17:35
[Coke] not familiar with the 'also does' syntax. weird. 17:36
disbot3 <librasteve> happens on zef install . --force-install ... running raku -I. t/01... all tests compile and run fine
<librasteve> its another way to say class X does Y {} - get the same error either way
<librasteve> (just easier to comment out if on a separate line) 17:37
[Coke] that's the commit, what branch?
disbot3 <librasteve> defaults
<librasteve> lib/Air/Base.rakumod line 892 17:38
17:40 eseyman left, manu_ joined, arkiuat left
[Coke] ok, the weird thing is that is NOT a parameterized role as far as I can tell - you have the stub at first, but that's the only "weird" thing. 17:46
disbot3 <librasteve> yeah - thanks for the rapid reply - I will raise as a compiler bug 17:51
17:54 arkiuat joined 17:56 apac joined 17:57 apa_c left 18:03 cryosis left 18:04 cryosis joined 18:07 arkiuat left
[Coke] I'm still trying to see if I can figure out what the real issue is 18:11
hurm. Site is a class. are all the other types you have in Defaults roles? 18:16
and *those* roles *are* parametric 18:17
ugexe just a wild guess, but maybe precompilation related? when using CURFS (-I., -Ilib) the precompilation order is based on the order things are actually used (such that things that are never used won't be precompiled at all). when installing via zef it gets precompiled in the staging repo, which precompiles using a more brute force method of iterating over every module and trying until it works 18:19
18:20 jgaz joined
[Coke] My testing here is with 'zef test .' would you expect the same issue there? 18:20
18:21 jgaz left
ugexe no, zef test doesn't use a staging repo. maybe it should, but it would mean reprecompiling everything every time tests are run 18:22
[Coke] ok - I'm still seeing the issue with zef test
ugexe if you see it with zef test then i would expect to see it with raku -I. t/whatever 18:23
[Coke] yup, RAKULIB=. raku t/03-page.rakutest has the same error 18:28
I have a golf 18:30
gist.github.com/coke/a114d8b286967...472f90acf6 18:31
Might be able to golf it more, but that's basically the problematic role hierarchy from Site. Dies with the same error steve is seeing 18:32
librasteve - hope that helps 18:33
18:36 arkiuat joined 18:41 arkiuat left 19:08 lnnk joined, jgaz joined, arkiuat joined
disbot3 <librasteve> good golfing ... I'm just sanity checking 19:10
lnnk Is raku simular to perl 5 or completely different? 19:11
19:13 arkiuat left, rantanplan is now known as snonux
disbot3 <librasteve> [Coke]: I have golfed down a bit further ... writing a bug issue right now 19:16
<librasteve> github.com/rakudo/rakudo/issues/5966 19:24
[Coke] .... I did not expect it to be THAT short. 19:35
... OK, I see a problem with our golf.
m: role Defaults { ... } ; class Site does Defaults { ... }; 19:36
camelia The following packages were stubbed but not defined:
Site
[Coke] m: role Defaults { ... } ; class Site does Defaults { };
camelia ===SORRY!=== Error while compiling <tmp>
No appropriate parametric role variant available for 'Defaults':
Cannot resolve caller (Site); Routine does not have any candidates. Is only the proto defined?
at <tmp>:1
[Coke] m: role Defaults { } ; class Site does Defaults { };
camelia ( no output )
disbot3 <librasteve> sorry, I'm not following 19:38
[Coke] steve - also: I can't reproduce your error on t/01
the error is only triggered by the presence of the ... in the role there.
which isn't part of the original code.
Also you say in the ticket that the error doesn't occur with t/01 - .. yes, it doesn't it's on t/03* and it fails whether you run it with zef test or raku t/03... 19:39
19:45 arkiuat joined
disbot3 <librasteve> To see the error on raku -I. t/03-page.rakutest and raku -I. t/04-table.rakutest you need to rm -rf .precomp 19:48
<librasteve> yeah - that's what you said (sorry)
19:50 arkiuat left
disbot3 <librasteve> I see that your golf is different to my code, but I do use a role Defaults { ... } stub at the top so I think that the golf is a bug and that it may well be the same bug 19:51
19:52 arkiuat joined 19:56 arkiuat left, LainIwakura joined
[Coke] ahhh, good call 20:05
if you remove the ... you get an error from Html that it *isn't* parametric. 20:07
20:10 arkiuat joined 20:11 LainIwakura left 20:14 Aedil left
disbot3 <librasteve> thanks for he help - I am going &afk for tomight and will refactor tomorrow... 20:15
20:15 hulk left 20:39 arkiuat left 20:47 lnnk left 20:52 arkiuat joined 21:02 arkiuat left 21:05 arkiuat joined 21:25 MoC left 21:54 apac left 22:09 arkiuat left 22:23 arkiuat joined 22:27 arkiuat left 22:30 Pixi__ is now known as Pixi 22:47 arkiuat joined 22:51 apac joined, arkiuat left 23:09 arkiuat joined 23:14 arkiuat left 23:31 arkiuat joined 23:36 arkiuat left 23:49 arkiuat joined

Powered by Cro and the Raku® Programming Language.

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

Thank you!