Welcome to the main channel on the development of MoarVM, a virtual machine for NQP and Rakudo (moarvm.org). This channel is being logged for historical purposes.
Set by lizmat on 24 May 2021.
Nicholas good *, #moarvm 06:08
nine Rule #1 when working on RakuAST: if you have no idea where some class might be declared, you'll find it in src/Raku/ast/scoping.rakumod 08:11
nine We're starting to resolve lexicals via HLL fallback for BEGIN time execution :) The end of those giant wrapper frames containing all lexicals from all scopes is coming closer 09:13
jnthn nine: Nice, is there a MoarVM side of that, or is it all done in HLL space? 09:36
tellable6 2022-06-19T20:38:04Z #moarvm <vrurg_> jnthn Is it intentional that method fallback doesn't respect MRO? Because if it is the discrepancy in `handles<m1 m2>` and `handles(*)` is disappointing and misleading when it comes to inheritance.
jnthn .tell vrurg Think I need an example to understand what you're asking 09:37
tellable6 jnthn, I'll pass your message to vrurg
nine jnthn: I added MVMCode *resolver and MVMCode *dynamic_resolver to MVMCompUnit, a syscall to set those and modified MVM_frame_find_lexical_by_name_outer and MVM_frame_getdynlex_with_frame_walker to call those if they couldn't find a lexical.
nine 467 spec test files passing 12:36
Altai-man wow, that's impressive 12:37
nine 474 spec test files passing! 12:50
And as so often it's the small, simple fixes that bring the most effect :D
vrurg m: class Foo { has $.d handles(*) = DateTime.now; }; class Bar is Foo { }; say Foo.new.yyyy-mm-dd; say Bar.new.yyyy-mm-dd; 13:05
camelia 2022-06-22
No such method 'yyyy-mm-dd' for invocant of type 'Bar'
in block <unit> at <tmp> line 1
tellable6 2022-06-22T09:37:07Z #moarvm <jnthn> vrurg Think I need an example to understand what you're asking
vrurg m: class Foo { has $.d handles<yyyy-mm-dd> = DateTime.now; }; class Bar is Foo { }; say Foo.new.yyyy-mm-dd; say Bar.new.yyyy-mm-dd; 13:06
camelia 2022-06-22
2022-06-22
vrurg jnthn: ^^^
jnthn vrurg: Looks like an oversight; I'd expect the first to work also 13:13
vrurg Because the implementation is based on fallbacks I think it where the fix should be done. Ok, it's now just a matter of finding some time for it. 13:14

Powered by Cro and the Raku® Programming Language.

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

Thank you!