[00:06] <Geth> ¦ setup-raku: 8082eeae33 | dependabot[bot]++ (committed using GitHub Web editor) | 2 files

[00:06] <Geth> ¦ setup-raku: Bump vitest from 4.0.18 to 4.1.0 (#53)

[00:06] <Geth> ¦ setup-raku: 

[00:06] <Geth> ¦ setup-raku: Bumps [vitest](https://github.com/vitest-dev/vitest/tree/HEAD/packages/vitest) from 4.0.18 to 4.1.0.

[00:06] <Geth> ¦ setup-raku: - [Release notes](https://github.com/vitest-dev/vitest/releases)

[00:06] <Geth> ¦ setup-raku: - [Changelog](https://github.com/vitest-dev/vitest/blob/main/docs/releases.md)

[00:06] <Geth> ¦ setup-raku: - [Commits](https://github.com/vitest-dev/vitest/commits/v4.1.0/packages/vitest)

[00:06] <Geth> ¦ setup-raku: 

[00:06] <Geth> ¦ setup-raku: <…commit message has 9 more lines…>

[00:06] <Geth> ¦ setup-raku: review: https://github.com/Raku/setup-raku/commit/8082eeae33

[01:02] <Geth> ¦ rakudo/main: 6a2787ef19 | (Timo Paulssen)++ | t/04-nativecall/02-simple-args.t

[01:02] <Geth> ¦ rakudo/main: TODO both problematic nativecall tests for moar+clang+dyncall

[01:02] <Geth> ¦ rakudo/main: 

[01:02] <Geth> ¦ rakudo/main: the code clang generates assumes the upper bytes of an unsigned char or

[01:02] <Geth> ¦ rakudo/main: unsigned short are zeroed out, but dyncall only has a function for passing

[01:02] <Geth> ¦ rakudo/main: signed char and short, where negative numbers end up sign-extended, messing

[01:02] <Geth> ¦ rakudo/main: clang's generated assembly for x == 0xFE up, which tries to comare the whole

[01:02] <Geth> ¦ rakudo/main: 32bit register against 0xFE.

[01:02] <Geth> ¦ rakudo/main: review: https://github.com/rakudo/rakudo/commit/6a2787ef19

[01:02] <Geth> ¦ rakudo/main: d24a012136 | (Will Coleda)++ (committed using GitHub Web editor) | t/04-nativecall/02-simple-args.t

[01:02] <Geth> ¦ rakudo/main: Merge pull request #6215 from rakudo/we_know_dyncall_plus_clang_is_broken_so_make_ci_green

[01:02] <Geth> ¦ rakudo/main: 

[01:02] <Geth> ¦ rakudo/main: TODO both problematic nativecall tests for moar+clang+dyncall

[01:02] <Geth> ¦ rakudo/main: review: https://github.com/rakudo/rakudo/commit/d24a012136

[01:18] <[Coke]> anyone know who has privs on https://dev.azure.com/MoarVM/MoarVM/ ?

[01:18] <[Coke]> I can see the runs, but if I try to rerun failed ones, I get TF400813: The user 'edb5f846-d424-4534-aab5-b9a3282ae013' is not authorized to access this resource.

[01:24] <[Coke]> ah. looks like I'm in Raku/nqp on azure.com, but not MoarVM/MoarVM - can someone with rights add me, please?

[01:50] <Geth> ¦ rakudo/main: 6 commits pushed by (Nick Logan)++

[01:50] <Geth> ¦ rakudo/main: b05721c44f | RakuAST: move Code parse time resolver to QASTContext side table

[01:50] <Geth> ¦ rakudo/main: 6131002cd9 | RakuAST: pass compile time helpers as parameters; drop Package compile time attributes

[01:50] <Geth> ¦ rakudo/main: b93ee95ec8 | RakuAST: break AST references to CompUnit so CompUnit can't reach the SC

[01:50] <Geth> ¦ rakudo/main: 96b14bc6e4 | RakuAST: extract `IMPL-COMPOSE-TYPE` helper on `Package`

[01:50] <Geth> ¦ rakudo/main: 5c72042747 | RakuAST: prevent bare meta-object reader from degrading Package compose

[01:50] <Geth> ¦ rakudo/main: 5c36a7924c | Merge pull request #6213 from ugexe/ugexe/rakuast-separate-compile-time-refs

[01:50] <Geth> ¦ rakudo/main: review: https://github.com/rakudo/rakudo/compare/d24a01213682...5c36a7924c34

[02:58] <Geth> ¦ rakudo: ugexe++ created pull request #6217: RakuAST: fix sub signature parameter and parametric role compile crashes

[02:58] <Geth> ¦ rakudo: review: https://github.com/rakudo/rakudo/pull/6217

[04:04] <Geth> ¦ rakudo: ugexe++ created pull request #6218: RakuAST: emit empty concat for modifier-only alternation branch

[04:04] <Geth> ¦ rakudo: review: https://github.com/rakudo/rakudo/pull/6218

[07:46] *** finanalyst joined
[08:29] <lizmat> ugexe++

[08:30] <lizmat> sometimes I wonder whether it would make sense to make a $*RESOLVER and $*CONTEXT dynvar instead of keeping passing them around

[08:36] *** finanalyst left
[08:43] <Geth> ¦ roast: ba1e02a602 | (Elizabeth Mattijsen)++ | S06-currying/misc.t

[08:43] <Geth> ¦ roast: Add tests for https://github.com/rakudo/rakudo/issues/6212

[08:43] <Geth> ¦ roast: review: https://github.com/Raku/roast/commit/ba1e02a602

[09:09] <Geth> ¦ roast: 867b9869f7 | (Elizabeth Mattijsen)++ | 4 files

[09:09] <Geth> ¦ roast: Untodo now passing .assuming tests

[09:09] <Geth> ¦ roast: 

[09:09] <Geth> ¦ roast: Since the RakuAST based .assuming implementation made it to the

[09:09] <Geth> ¦ roast: 2026.05 release, it's time to make these tests canonical

[09:09] <Geth> ¦ roast: review: https://github.com/Raku/roast/commit/867b9869f7

[10:32] *** lizmat left
[10:57] *** lizmat joined
[11:23] *** lizmat left
[11:24] *** lizmat joined
[12:08] <ab5tract> From a grammar standpoint, would it be possible to support 'do' in postfix statements?

[12:09] <ab5tract> m: with True do say "I'd like this"

[12:09] <camelia> rakudo-moar 5c36a7924: OUTPUT: «===SORRY!=== Error while compiling <tmp>␤Missing block␤at <tmp>:1␤------> with True<HERE> do say "I'd like this"␤    expecting any of:␤        block or pointy block␤        infix␤        infix stopper␤»

[12:11] <ab5tract> oops, not postfix statements at all. This is in place of the regular block form, as opposed to the postfix statements. 

[12:12] <ab5tract> m: with $maybe do .definitely

[12:12] <camelia> rakudo-moar 5c36a7924: OUTPUT: «===SORRY!=== Error while compiling <tmp>␤Variable '$maybe' is not declared. Perhaps you forgot a 'sub' if this␤was intended to be part of a signature?␤at <tmp>:1␤------> with <HERE>$maybe do .definitely␤»

[12:14] <ab5tract> crucially, we'd want the auto-topicalization behavior that makes postfix so useful. 

[12:15] <ab5tract> For my taste, postfix works great for if/unless. but I really struggle to enjoy it with 'for' and 'with'

[12:16] <ab5tract> `for @foo do .frob` feels like it might be a nice compromise, assuming that it is even possible

[12:28] <lizmat> .definitely with $maybe

[12:28] <lizmat> what's wrong with that ?

[12:29] <lizmat> not sure what you want to achieve  :-)

[12:49] <ab5tract> I don't like that syntax

[12:49] <ab5tract> .definitely unless maybe # makes sense as a sequence of words

[12:50] <ab5tract> .definitely for maybe # does not

[12:51] <ab5tract> This is a personal thing. I was just trying to gauge whether it is even possible.

[12:52] <lizmat> m: sub do($a) { $a }; do .say with 42

[12:52] <camelia> rakudo-moar 5c36a7924: OUTPUT: «42␤»

[12:52] <ab5tract> lizmat: that's not the same thing though

[12:53] <ab5tract> that's using postfix with still

[12:53] <ab5tract> I am simply suggesting that do be allowed to stand in for a block and have auto-topicalization while doing so

[12:54] <ab5tract> if $wrong do make-it-right

[12:54] <ab5tract> if $wrong do .&make-it-right

[12:56] <ab5tract> And suggesting might even be too strong a word. My intent is mostly to see if there's any interest beyond myself and to gauge difficulty, if possible.

[13:08] <ShimmerFairy> I don't see why `with 42 { .say }` wouldn't suffice. I get that it's aesthetically different from having an english word, but the braces use the same number of meaningful characters, and works with the language as-is (whereas 'do' in this manner would be a new syntax, just for this feature).

[13:17] <librasteve_> i kinda like it from a user pov (can’t speak to the difficulty), I like `.say for ^3` and `.say with $x` but would be happy to see `for ^3 do .say` timtowdi wise

[14:17] *** hurufu joined
[14:41] <ab5tract> Looking at it more, it does look like 'do' would be adding a sepcial semantics for this case, which makes me like the idea a lot less

[14:45] <lizmat> it would also make error reporting on improper use of "do" a lot more difficult  :-(

[15:16] <ab5tract> okay, this is more for the fun of thinking about it than anything else, but maybe pointy arrow could have the desired semantics when used without a signature *or* a block. It could take a single expression and have -> ($_) as the signature

[15:16] <ab5tract> for @all -> .say

[15:46] <lizmat> I don't see the advantage above:  .say for @all

[15:56] *** hurufu left
[16:01] <ab5tract> Again, I really don't like that syntax and rarely use it. Something about the bare method call and the out of order language use

[16:02] <ab5tract> for @all { .say } is fine. I am also trying to discuss the capabilities of the grammar here, is this feasible, let's have a discussion, etc\

[16:05] <ab5tract> *the bare method call starting off the statement list

[16:24] *** hurufu joined