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.
01:55 ilbot3 joined 04:53 Ven`` joined 08:52 Ven`` joined
Ven`` o 09:17
wait, where's my arm?
o/ -- here we go
masak: Your PR looks like some great work towards metacircularity. 09:20
masak that would explain why everything's so slow 09:28
I'm submiting issues at a higher sustained pace than ever before in 007's lifetime 09:30
many of them about things I've suddenly discovered that need further attention later
like hte thing in the latest commi about Q::Unquote -- either it needs to show up basically *everywhere*, in all type unions in Qnode properties, or some correspondingly poweful but more clever solution needs to be applied. 09:32
he most promising of which, briefly, was "make `class Q::Unquoe<T> extends T`, so that unquote Qtypes can masquerade as the Qtype they're understudying". took me a while to convince myself that that'd be a clear abuse of subtyping. 09:34
anyway, my vacation is coming to a close -- it seems I don't quite have enough time to finish all of the small leftovers in #242. hopefully during some travel next week, though. 09:36
Ven`` You did an insane amount of work, tbh. 09:37
I'm really lacking time ATM. Juggling with too many things... 09:38
I think I'll be able to finish the regex PR this month. But maybe I'll allocate more time to captures and that's gonna take some thinking.
masak cool.
I'll try to provide imely review on everything I see in public branches :) 09:39
timely*
Ven`` well, it's more of a private branch, since it's on my fork (I'm not member of 007 so I can't create branches anyway) 09:47
I'm not sure I feel good about removing Q::Statement::Block 09:52
how do you write { local $/; ... } then?! :P
masak heh 10:01
Ven``: just sent you an invite. feel free to work in PRs right on masak/007 if you want. 10:03
Ven``: I'm not convinced I should remove Q::Statement::Block -- it's got quite a small footprint, and doesn't bother anyone. but if the criterion is "do we need this in core for bootstrapping?", then maybe it goes away 10:04
Ven`` got it, appreciated 10:09
masak: I think block is useful only as a "everything is an expression"-kind of thing 10:10
(`do` or not)
masak I figured out one reason I'm averse to the "everything is an expression" thing, by the way 10:17
namely that I founded 007 as an experiment that would lead to a language that was *not* a Lisp, and yet would have powerful macros
and so EIaE would seem to slip back into Lispness 10:18
Ven`` yes, but then, Perl 6 has "do" :P. 10:23
I definitely see your point however. And I know we talked a bit about it on a very different topic -- Yapsi 10:38
Your effort to "separate control flow from Perl 6's" also reminded of what I do in a toy VM I wrote in Rust, where I use recursion to implement calls. But reading your issues and your changes, I now realize it's a bad idea 10:40
12:30 vendethiel- joined
masak that issue is inspired by a compiler book I have at home. it states that a common way for interpreters to handle control flow away from the host language's stack is the way it's handwaved in the issue 13:54
Ven`` what might that book be? 13:56
masak I think it's www.amazon.co.uk/Modern-Compiler-W...0471976970 14:05
Ven`` Thanks! Well, who am I kidding, I'm far behind on the books I plan to read anyway... 14:07
but it's still interesting to skirm through.
masak aye, that's exactly what I was doing 14:08
Ven`` plans to re-read Ruby Through a Microscope someday, with actual knowledge and experience
masak .oO( just read it with reading glasses, sheesh ) :P 14:11
Ven``
.oO( Though sometimes you need sunglasses to read code, because it's so bright ) :P
 14:12
20:41 Ven`` joined 21:57 Ven`` 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!