tbrowder lizmat: ++ ditto 01:39
01:48 Altai-man_ joined 01:51 sena_kun left 01:59 hungrydonkey joined 02:33 hungrydonkey left 02:34 hungrydonkey joined 03:49 sena_kun joined 03:51 Altai-man_ left 05:48 Altai-man_ joined 05:51 sena_kun left 06:21 hungryd85 joined 06:23 hungrydonkey left 06:33 hungryd85 left, hungryd28 joined 06:43 ufobat_ left
lizmat Files=1306, Tests=111228, 213 wallclock secs (28.64 usr 8.00 sys + 2992.13 cusr 266.92 csys = 3295.69 CPU) 07:03
07:14 hungryd28 left 07:44 hungrydonkey joined 07:49 sena_kun joined 07:51 Altai-man_ left, hungrydonkey left 07:52 hungrydonkey joined 08:05 hungrydonkey left 08:44 [Tux] left 09:13 [Tux] joined 09:19 [Tux] left, [Tux] joined 09:20 [Tux] left 09:28 [Tux] joined 09:48 Altai-man_ joined 09:51 sena_kun left 10:05 cognomin_ joined 10:10 cognominal left 10:19 jjatria left 10:20 jjatria joined 10:21 jjatria left, jjatria joined 10:22 jjatria left, jjatria joined 11:40 MasterDuke left 11:49 sena_kun joined 11:50 Altai-man_ left 12:50 hungrydonkey joined 13:06 hungrydonkey left 13:09 hungrydonkey joined 13:13 hungrydonkey left 13:22 hungrydonkey joined 13:36 hungrydonkey left 13:48 Altai-man_ joined 13:51 sena_kun left 14:10 MasterDuke joined 14:15 hungrydonkey joined 14:24 hungryd47 joined, hungrydonkey left
vrurg greppable add_phaser 14:27
greppable6: add_phaser
greppable6 vrurg, 11 lines, 5 modules: gist.github.com/1a4aabf2bb09221e3f...b3493dcb9b
.oO( I sure hope they're all calling it at BEGIN time... )
nine So I either compile this code github.com/niner/Inline-Perl5/blob...V6.pm#L142 inside a role declaration, then it will fail because of the our scoped sub, or in a class declaration, then the methods will type check the invocant for that class and I can't use them for the actual target class. 14:35
Seems like I can't win :/
14:53 Kaiepi left 14:57 Ven`` joined
Xliff Is anyone here familiar with CompUnit::Util.re-export? 15:25
github.com/LLFourn/p6-CompUnit-Uti...l.pm6#L156 15:26
15:33 hungryd47 left 15:38 hungrydonkey joined
Geth_ roast: ace2bed002 | (Elizabeth Mattijsen)++ | S03-sequence/exhaustive.t
Some more exhaustive ... tests
15:49 sena_kun joined 15:51 Altai-man_ left 15:59 Xliff left 16:18 hungrydonkey left 16:20 hungrydonkey joined, Kaiepi joined
nine I think the actual way to solve my problem is to use ExportHOW to make it syntactically possible to create an Inline::Perl5::ClassHOW based class syntactically, i.e. wrap that code in a p5class { ... } 16:27
Of course that leads down to funny errors like "getlex: outer index out of range" 16:39
Oh oh...the trouble goes away with no precompilation; 16:57
17:03 hungrydonkey left 17:04 hungrydonkey joined, hungrydonkey left 17:42 maggotbrain joined 17:48 Altai-man_ joined 17:51 sena_kun left
tyil jjatria: do you need any help with the docker patch for rstar? 18:38
I can apply what you currently have and shave off rough edges tomorrow if that's ok with you 18:39
[Tux] Rakudo version 2020.02.1-295-gafff3b0b2 - MoarVM version 2020.02.1-76-gec53e4dfc
csv-ip5xs0.709 - 0.716
csv-ip5xs-206.038 - 6.193
csv-parser24.018 - 24.047
csv-test-xs-200.386 - 0.389
test7.574 - 7.683
test-t1.943 - 2.044
test-t --race0.960 - 0.978
test-t-2031.936 - 32.486
test-t-20 --race9.594 - 9.745
jjatria tyil: I think what I needed more than anything was time to sit down and teach my git to email :D 19:20
tyil: But yeah, go ahead. The patch on my gitlab fork is in good enough shape I think
timotimo what's the best way to check for an intarray / numarray in a signature outside of the native_arrays source file? (where the roles live inside of a class) 19:44
19:49 sena_kun joined 19:51 Altai-man_ left
timotimo checking for "array" seems like a good start. and it should be okay to check the .of in the method body then 19:51
jnthn m: multi m(int @) { say "int" }; multi m(num @) { say "num" }; m(my int @); m(my num @)multi m(int @) { say "int" }; multi m(num @) { say "num" }; m(my int @); m(my num @) 19:56
camelia 5===SORRY!5=== Error while compiling <tmp>
Two terms in a row
at <tmp>:1
------> 3 { say "num" }; m(my int @); m(my num @)7⏏5multi m(int @) { say "int" }; multi m(nu
expecting any of:
infix stopper
jnthn huh
oh, copied twice
m: multi m(int @) { say "int" }; multi m(num @) { say "num" }; m(my int @); m(my num @)
camelia int
jnthn timotimo: ^
tyil jjatria: ack, thanks! 19:57
nine So I guess to support assignable attributes objects of my custom meta class must create the scalar containers for them? 19:58
At least that's what I gather from: "assign requires a concrete object (got a VMNull type object instead)"
jnthn nine: Hm, I'm a little curious how you ended up without them, in so far as the I thought Attribute.new would set up a Scalar container and its descriptor 19:59
timotimo d'oh, that's so obvious 20:01
nine Well, I now EVAL "perl5class GLOBAL::$package { $body }" with $body being the code from: github.com/niner/Inline-Perl5/blob...V6.pm#L142
20:01 Ven`` left
timotimo m: multi m(int @) { say "int" }; multi m(num @) { say "num" }; m(my int8 @); m(my num32 @) 20:01
camelia 5===SORRY!5===
Calling m(Positional[int8]) will never work with any of these multi signatures:
(int @)
(num @)
at <tmp>:1
------> 3 "int" }; multi m(num @) { say "num" }; 7⏏5m(my int8 @); m(my num32 @)
Calling m(Position…
nine I made it past Metamodel::Primitives.compose_type and it recognizes the existence of the attribute now, but apparently it doesn't get initialized for new objects yet 20:02
timotimo hum
jnthn How did you create the attribute?
nine My meta class does Metamodel::AttributeContainer and the compiler presumably called add_attribute for me 20:03
jnthn Ah, so not sublcassing ClassHOW
You'll need to also participate in BUILDPLAN construction too
There's a role, and see ClassHOW for what to call in your compose
nine So, Perl6::Metamodel::BUILDPLAN and at least self.create_BUILDPLAN($obj); 20:04
jnthn Yeah. Maybe that sufficies even. 20:05
It's probably been close to a decade since I implemented that, so it's not very hot in the cache. :)
nine Ah, the fun :) Cannot find method 'lang-rev-before' on object of type Inline::Perl5::ClassHOW 20:06
Well self.create_BUILDPLAN was not enough by itself. Will add some more code... 20:09
Oh, it could also be the way I create those objects currently 20:10
$class.CREATE is probably just not enough (considering it's the _BUILD_PLAN that does the work) 20:11
jnthn Yeah, that's not enough. The default BUILDALL form Mu interprets the build plan 20:13
We can also compile 'em
Though you need to be calling from the compiler for that to work out.
nine Well switching to .bless which runs BUILDALL is still not enough
Oh....I guess method BUILDALLPLAN($type) { [].FLATTENABLE_LIST } just doesn't cut it anymore :) 20:15
Got rid of it as I get it from Metamodel::BUILDPLAN anyway...but still 20:16
Hm...BUILDPLAN and BUILDALLPLAN don't look that bad: 20:17
((0, Foo::Bar::TestV6, "\$!name", "name"), (0, Foo::Bar::TestV6, "\$!wrapped-perl5-object", "wrapped-perl5-object"), (10, Foo::Bar::TestV6, "\$!name"), (10, Foo::Bar::TestV6, "\$!wrapped-perl5-object"))
((0, Foo::Bar::TestV6, "\$!name", "name"), (0, Foo::Bar::TestV6, "\$!wrapped-perl5-object", "wrapped-perl5-object"))
20:33 patrickb joined
timotimo how icky would y'all think it'd be if we generated code for hyperops right in the Perl6::Optimizer? like we do with junction unfolding and for loop simplification? 21:46
21:48 Altai-man_ joined 21:51 sena_kun left 21:55 patrickb left 21:58 lucasb joined
Geth_ nqp/master: 4 commits pushed by (Patrick Böker)++ 22:01
23:32 Xliff joined
Xliff m: class A is repr<CStruct> { has guint $.a }; class GstObject is repr<CPointer> { * }; my $pa = cast(GstObject, A.new) 23:45
camelia 5===SORRY!5===
Type 'guint' is not declared. Did you mean any of these?

at <tmp>:1
------> 3class A is repr<CStruct> { has guint7⏏5 $.a }; class GstObject is repr<CPointer
Malformed has
at <tmp>:1
Xliff m: class A is repr<CStruct> { has uint32 $.a }; class GstObject is repr<CPointer> { * }; my $pa = cast(GstObject, A.new)
camelia 5===SORRY!5=== Error while compiling <tmp>
Undeclared routine:
cast used at line 1. Did you mean 'last', 'cas'?
Xliff m: class A is repr<CStruct> { has uint32 $.a }; class GstObject is repr<CPointer> { * }; my $pa = nativecast(GstObject, A.new)
camelia 5===SORRY!5=== Error while compiling <tmp>
Undeclared routine:
nativecast used at line 1
Xliff m: use NativeCall; class A is repr<CStruct> { has uint32 $.a }; class GstObject is repr<CPointer> { * }; my $pa = nativecast(GstObject, A.new)
camelia ( no output )
Xliff Hmm....
23:49 sena_kun joined 23:51 Altai-man_ left