6macros: discussing the finer points of Perl 6 macros, Qtrees, and how to stay sane | irclog: irclog.perlgeek.de/6macros/today
Set by moderator on 28 July 2015.
13:00 cog__ joined 17:54 Ven joined 18:13 Ven joined 18:34 Ven joined 18:53 Ven joined 19:13 Ven joined 19:33 Ven joined 19:53 Ven joined
masak Ven: oh! cool! 19:56
masak thinks about the question
your own reaction -- "no q types in val, ever!" -- is largely right
the only time we break that is with Val::Sub, which contains a Q::ParameterList and a Q::StatementList. but I've always considered those to be somewhat suspect and worthy of excision 19:58
Ven: in the absence of a need/use case for the opposite, I think Val::Regex can be pretty opaque, and not need to respond to "unpacking" or destructuring 20:00
the only thing we really require of it is that it match on strings
that way, I think we can avoid a proliferation of new Val::Regex::* types 20:01
instead, we can aim to be as creative as we need with the Val:: types we already have
vendethiel so I should convert the Q::Regex::* to other stuff? 20:07
(I have Q::Regex::{Identifier,Call,Group}) 20:08
20:13 Ven joined 20:33 Ven joined
masak no, the Q::Regex::* types are fine 20:39
in the Q end of things we *do* need the type fine-grainedness 20:40
I guess what I'm saying is that if a Val::Sub can be a mostly-opaque type then Val::Regex can, too
20:53 Ven joined
vendethiel well, not sure how that'll turn how to work then 21:04
I'll check on Val::Su
21:05 Ven joined
Ven > Could not locate compile-time value for symbol Val::Array 21:07
whatever did I do ._. 21:08
I'm not sure how to fix this Perl 6 error, at all... 21:19
even stashing doesn't fix it, whoops... 21:20