🦋 Welcome to the IRC channel of the core developers of the Raku Programming Language (raku.org #rakulang). This channel is logged for the purpose of history keeping about its development | evalbot usage: 'm: say 3;' or /msg camelia m: ... | Logs available at irclogs.raku.org/raku-dev/live.html | For MoarVM see #moarvm
Set by lizmat on 8 June 2022.
Geth Template6/main: e259e083c1 | (Polgár Márton)++ | 7 files
0.11.3
 00:04
Geth rakudo/main: 500ad5f637 | (Elizabeth Mattijsen)++ | src/core.c/unicodey.pm6
Make Cool.uninames actually work

It had a wrong return value type		 09:54
Nemokosch bisectable6: sub ponder { next }; for (1..3) { ponder; .say }; say "done" 10:01
bisectable6 Nemokosch, Will bisect the whole range automagically because no endpoints were provided, hang tight
tellable6 2023-03-07T22:16:17Z #raku <rf> Nemokosch: How do I clear the stash in Template6
bisectable6 Nemokosch, ¦6c (67 commits): «done␤»
Nemokosch, Nothing to bisect!
Nemokosch ew
lizmat m: sub ponder { next }; for (1..3) { ponder; .say }; say "done" 10:04
camelia done
lizmat m: sub ponder { next FOO }; FOO: for (1..3) { ponder; .say }; say "done" 10:05
camelia ===SORRY!=== Error while compiling <tmp>
Variable '&FOO' is not declared. Perhaps you forgot a 'sub' if this
was intended to be part of a signature?
at <tmp>:1
------> sub ponder { next ⏏FOO }; FOO: for (1..3) { ponder; .say };
lizmat m: FOO: for (1..3) { next FOO; .say }; say "done"
camelia done
Nemokosch now I'm looking up roast if this case was ever planned for 10:06
it seems fairly intuitive that next/last stuff should behave like return
lizmat next / last / redo are control exceptions 10:08
m: CONTROL { say .^name }; next 10:09
camelia CX::Next
next without loop construct
in block <unit> at <tmp> line 1
lizmat any loop construct basically sets up a "catcher" for such exceptions
Nemokosch I don't think the implementation should matter too much for the expectations 10:10
A Block acts inlined, a Sub is context-insensitive
lizmat which is why it is able to say "next without a loop construct" because there's an outer CONTROL block that catches them and puts out the message
well, I guess you could call that behaviour another Perl heritage 10:11
Nemokosch actually this doesn't even sound too bad 10:13
subs need a CONTROL handler by default, sounds about fixed?
lizmat except from an optimization point of view 10:14
BTW, do you want a warning or throwing an exception ? 10:16
anyways, I think this could be handled easier with a check at the receiving side, checking the outer lexical sub scope: if different warn or throw 10:17
Nemokosch > CX::Next␤next without loop construct␤ 10:18
I think this would fit perfectly for these scenarios as well
lizmat still, that also would impose a performance penalty, but at least it would be for the use of next/last/redo
afk& 10:19
Nemokosch if somebody wants to live on the edge, they can use Block
I never knew there was goto in Raku lol 10:41
once RakuAST is ready, it will become much easier to search code for certain structures 10:58
by the way, from optimization point of view - in RakuAST, it will be easy to detect if this "guarding handler" is necessary at all 11:02
One more thing: I wonder how the specification could handle a change like this. First off, it's not obvious whether this is a change because it doesn't seem to be present in the spectests. But since code might depend on it (and it does, confer doc tests), let's say it is a design-level change. 11:32
is it even possible to intervene this deeply from core at all? 11:33
Geth rakudo/main: e8f419aefb | (Stefan Seifert)++ | 4 files
RakuAST: fix colonpairs silently getting dropped from identifiers
 14:25
nine I don't like this, but it improves things...
Geth rakudo/main: af0c88cc2a | (Elizabeth Mattijsen)++ | src/core.c/unicodey.pm6
Make Cool.uniname actually match the documentation

There do not appear to be any tests :-(		 16:19
lizmat wow, that took a while
vrurg Oh, was stepping on this rake long ago... 16:26
*I was

Powered by Cro and the Raku® Programming Language.

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

Thank you!