🦋 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.
.vushu @ab5tract I made some adjustment which made it look more natural with the snow 00:45
gist.github.com/vushu/ea36578a5728...26eed768d1
MasterDuke what's the point of the 'compactable' method? 01:15
.vushu: i get a 'substr out of range' error when building Raylib::Bindings, though it does install. is that expected? 01:26
.vushu @MasterDuke I dont think that is suppose to happen can you run any of the examples? 04:02
MasterDuke i get `Cannot locate native library '/home/dan/Source/perl6/install/share/perl6/site/resources/C71D39DAFB1D2C5664F76036FCEDF35E0B2DD2F3.so': /home/dan/Source/perl6/install/share/perl6/site/resources/C71D39DAFB1D2C5664F76036FCEDF35E0B2DD2F3.so: undefined symbol: DrawLineBSpline 04:24
  in method setup at /home/dan/Source/perl6/install/share/perl6/core/sources/4BEB2EC48C11F398F1B46D4DE03D243498D94F41 (NativeCall) line 319 ...`
but i'll admit i'm on a bunch of different branches of moarvm and rakudo, so that may have something to do with it 04:26
that was just doing `use Raylib::Bindings; my $white = Color.init(245, 245, 245, 255);` 04:27
ah, my raylib.h being in /usr/include/raylib.h instead of /usr/local/include/raylib.h might have something to do with it 04:31
hm. replacing the path in github.com/vushu/raylib-raku/blob/...e.raku#L22 and running `zef install . --force` gave the same errors... 04:34
ah, it was github.com/vushu/raylib-raku/blob/...e.raku#L14 and github.com/vushu/raylib-raku/blob/....raku#L35. i just commented out line 35 and let it use the corrected hard-coded path to my raylib.h and it worked fine 04:44
.vushu ah good that it worked out 😊 06:06
but the installation should find the raylib.h automatically using pkg-config so Im not sure what went wrong. 06:08
ab5tract MasterDuke: 'compactable' is to keep the size of the arrays mostly constant. that was designed to minimize GC, but something goes asymptotic there even though I'm not clear on why there would be more work for GC run #10_xxx to do than GC run #1. .vushu++'s solution works perfectly but I am surprised that it is necessary. 07:01
.vushu ab5tract isn’t the problem that you keep on making more snow flakes contra removing them from the arrays, I just noticed in your gist while printing the arrays elems out, that they just kept increasing in items. 07:26
ab5tract That’s it
I should have had coffee and done some more testing 07:27
I should have had coffee and done some more testing 07:28
(clearlu xD)
.vushu coffee is the important fuel 😄 07:29
librasteve o/ 08:33
Rajhan: if you are new to coding in general I recommend the Think Raku book (which is a sister to the Think Python one) ... it's a free pdf greenteapress.com/wp/think-perl-6/ 08:34
tellable6 librasteve, I'll pass your message to Rajhan
librasteve lizmat: draft for the 15/12 advent post is finished, please can you schedule it on wp? 08:35
lizmat librasteve will do
ab5tract .vushu: I backported the fixes to the layered, supply based renderer: gist.github.com/ab5tract/ff5373154...5fb0599c93 08:41
ab5tract so now you know I how I solved the performance issue: with a little help from some friends :) 08:42
lizmat librasteve applied some tweaks and fixed some typos and scheduled it 08:56
nemokosch ab5tract: what are you using NativeCall for? 08:59
lizmat librasteve am confused by the parse method in Contact::Address::USA
that looks wrong in a few ways 09:00
same fir ::UK
*for
why the -use- in there? 09:01
ab5tract nemokosch: nothing, thanks for reminding me to remove that 09:19
.vushu .ab5tract neat! 09:26
nemokosch ha gotcha, it happens 😄 09:35
El_Che lo 12:05
dakkar should `raku --profile` still work? 13:14
because I'm getting not great results…
(2023.10 on MoarVM) 13:15
www.thenautilus.net/~dakkar/raku-profile.html
the code is gist.github.com/dakkar/e96dcdb3c0f...5ed138b259 13:16
I was trying to see why a small application I wrote gets very very upset when I'm downloading a file into one of the directories it watches 13:17
turns out, something in the IO::Notification code is slow, and every time the watched file changes, the event gets queued inside the raku process and eventually the whole thing gets killed by the OOM killer 13:18
(yes, even when there's a no-op attached to the supply)
so I profiled that minimal code (`raku --profile watch.raku` and `perl fill.pl` in parallel) 13:19
and got that pretty useless profile ☹
suggestions?
.vushu .ab5tract I think you can improve the snowfalling on the ground, like my gist gist.githubusercontent.com/vushu/e...cycle.raku 14:03
ab5tract .vushu: I kind of like having it disappear into the ground. BTW, I found the bugs that were causing the arrays to grow and grow in my splice version. this one runs at a locked 59fps for me: gist.github.com/ab5tract/ff5373154...5fb0599c93 14:58
I'll play around some more and see if I can get it to fall on top fill line without turning into towers 14:59
.vushu ab5tract cool : 15:00
ab5tract thanks for your help! still loving raylib :D
.vushu no problem, glad I could help 🙂 15:01
[Coke] m: OUTER: for 1..9 { last OUTER } 16:20
camelia Cannot resolve caller last(OUTER:U); none of these signatures matches:
( --> Nil)
(Label:D $x --> Nil)
in block <unit> at <tmp> line 1
16:21
[Coke] m: OUTER: for 1..9 { next OUTER }
camelia Cannot resolve caller next(OUTER:U); none of these signatures matches:
( --> Nil)
(Label:D $x --> Nil)
in block <unit> at <tmp> line 1
[Coke] I thought those were implemented. 16:22
ouch. Can't have OUTER as a label. 16:24
m: NONBARF: for 1..9 { next NONBARF }
camelia ( no output )
[Coke] m: CALLER: for 1..9 { next CALLER }
camelia Cannot resolve caller next(CALLER:U); none of these signatures matches:
( --> Nil)
(Label:D $x --> Nil)
in block <unit> at <tmp> line 1
ab5tract kind of makes sense, no? but maybe could have a better error message? 16:27
[Coke] at least need a doc note explaining that not all labels are cool. 16:28
but we could probably check if the thing we found is even a lable.
*label
I feel like the error should be on label definition if we can't disambiguate between the original object and a label of the same name. 16:29
lizmat I feel that a label definition should override any other semantics 16:33
librasteve lizmat: thanks for your questions - the working code for Contact is here github.com/librasteve/raku-Contact 16:44
it is not yet a raku module since I prioritized the advent post over test and release
so I think that the way my Contact::AddressFactory uses a country param to pickup the country specific parser (which code was already covered in part 1 btw) is pretty slick ... but please feel free to code review 16:46
also I have stripped out some of the to-be- module code - but I thought good to leave in the Factory despatch example ... 16:47
just made some last minute edits on the post to clarify how this works ... happy to hear if I have misconstrued how to use use within a block. 16:59
[Coke] anyone have a 4-color version of camelia? 20:26
Was thinking convert blue and green to yellow, maybe. 20:27