🦋 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: ... | log inspection situation still under development | For MoarVM see #moarvm
Set by lizmat on 22 May 2021.
japhb m: my uint32 $i = 4294967295; say $i; class A { has uint32 $.i = 4294967295 }; A.new.i.say; # HUH? 00:11
camelia 4294967295
-1
moon-child japhb: probably related to github.com/rakudo/rakudo/issues/3936 00:15
japhb Hmmm, quite possibly related, but doesn't look to be precisely the same 00:17
OK, added a comment to that issue. Thanks for finding it, moon-child 00:19
melezhik . 00:27
raydiak m: my uint32 $i = 4294967295; say $i.VAR.WHAT; class A { has uint32 $.i = 4294967295 }; A.new.i.VAR.WHAT.say; 02:52
camelia (IntLexRef)
(Int)
raydiak the only difference I can find so far
Geth rakudo: a2faac4bd8 | (Elizabeth Mattijsen)++ | src/core.e/control.pm6
Oops, forgot to add the actual file, vrurg++

FWIW, it all tested ok locally :-(
08:31
sena_kun releasable6, status 11:02
releasable6 sena_kun, Next release in ≈4 days and ≈7 hours. 4 blockers. Changelog for this release was not started yet
sena_kun, Details: gist.github.com/1eb4c371b722fd7c92...553b66b954 11:03
Geth problem-solving/master: 5 commits pushed by (Daniel Sockwell)++, (Juan Julián Merelo Guervós)++, (Elizabeth Mattijsen)++ 11:17
Geth rakudo/new-disp: f7cbe9e046 | (Jonathan Worthington)++ | 2 files
Eliminate MAST::Call use in bind type checking

Also use a dispatch op here; most likely a lot of the binds are monomorphic and so the dispatch program can judge them quickly. Code size is also reduced by not needing to emit something to produce an error with every bind operation.
13:14
vrurg lizmat: of course it was building, you had the file locally. :) That's why I prefer PRs, this kind of mistake they can filter out too. 13:41
lizmat good point 13:42
|Tux| Rakudo v2021.06-81-ga2faac4bd (v6.d) on MoarVM 2021.06-22-gfb5d26a7e
csv-ip5xs0.857 - 0.996
csv-ip5xs-208.739 - 9.417
csv-parser27.681 - 27.716
csv-test-xs-200.369 - 0.370
test6.796 - 7.395
test-t1.999 - 2.043
test-t --race0.879 - 0.987
test-t-2033.067 - 33.374
test-t-20 --race9.712 - 9.883
13:49
Geth rakudo/new-disp: 068741a912 | (Jonathan Worthington)++ | 2 files
Implement p6sink using a dispatcher

The dispatcher is a no-op not only in the case that the value is not concrete or lacks a `sink` method, but also - newly - when the sink methood is exactly that in `Mu`, which we assume is a no-op. This should get rid of numerous sink method calls. It's also the last place in Rakudo itself that a legacy invoke op was emitted.
14:16
vrurg is hypnotized by all this new-disp comments... 14:17
lizmat has the same thing :-) 14:19
releasable6 Next release in ≈4 days and ≈3 hours. 4 blockers. Please log your changes in the ChangeLog: github.com/rakudo/rakudo/wiki/ChangeLog-Draft 15:00
Geth nqp/new-disp: f9a46b9c12 | (Jonathan Worthington)++ | src/vm/moar/QAST/QASTCompilerMAST.nqp
Switch immediate block invocation to dispatch op

This is probably the last case of an invoke op being produced by the QAST compiler.
15:15
lizmat notable6: weekly 15:21
notable6 lizmat, No notes for “weekly”
sena_kun bisectable6, say Uni(97) cmp Uni(97) 15:32
bisectable6 sena_kun, Will bisect the whole range automagically because no endpoints were provided, hang tight
sena_kun, Output on all releases: gist.github.com/2e883544482a416f0c...741d1e459e
sena_kun, Bisecting by output (old=2020.11 new=2020.12) because on both starting points the exit code is 1
sena_kun, bisect log: gist.github.com/f34d8a9474a912eef7...380369c0ca
sena_kun, (2020-12-08) github.com/rakudo/rakudo/commit/79...0f7bdc46f6
sena_kun, Bisecting by output (old=2016.04 new=2016.05) because on both starting points the exit code is 1
sena_kun, bisect log: gist.github.com/92cfb4f61ac9b4bd00...34dd4c994e 15:33
sena_kun, (2016-05-12) github.com/rakudo/rakudo/commit/33...d4efe221de
sena_kun, Output on all releases and bisected commits: gist.github.com/6cf77da932fa9a2de9...ffe3685caa
lizmat and yet another Rakudo Weekly News hits the Net: rakudoweekly.blog/2021/07/19/2021-...uled-to-3/ 15:39
melezhik lizmat some of my r3 test fails on the latest commit - github.com/rakudo/rakudo/issues/4467 , HTH 15:52
lizmat melezhik: that's probably because it needs "use v6.*" now ?
or "use v6.e.PREVIEW" 15:53
melezhik when did it start as requirement?
lizmat with that last commit
melezhik ah, ok
lizmat it was felt that next/last foo is a new feature that should not be available in 6.c or 6.d
afk for a few hours& 15:54
melezhik not, sure, use v6.e.PREVIEW did not help at least for one of the test 15:55
updated the issue 16:01
Geth rakudo: lizmat self-assigned r3 tests fail on the latest commit github.com/rakudo/rakudo/issues/4467
e777fc4fc0 | (Elizabeth Mattijsen)++ | src/core.c/Any-iterable-methods.pm6

Because the .grep logic lives in 6.c, it would be calling 6.c's next/last that don't accept any value. Fix this by directly calling the exception throwing logic, which *does* work in 6.c. This might actually have some performance benefit as well, as it removes one level of indirection.
Also normalize the order of signal code to NEXT/LAST.
This fixes #4467, melezhik++
18:38
Geth ¦ rakudo: lizmat self-assigned Comparison fails for Uni github.com/rakudo/rakudo/issues/4464 18:57
Geth nqp/new-disp: 3f496d0baa | (Jonathan Worthington)++ | 3 files
Switch NQP multiple dispatch over to new-disp
20:02
rakudo/new-disp: ecc266d1c2 | (Jonathan Worthington)++ | src/vm/moar/Perl6/Ops.nqp
NQP now installs proto code objects in the lexpad

This was done for the sake of new-disp multi dispatch, but as a nice side-effect we get this simplification too.
20:03
Geth nqp: 0e14b10a57 | (Daniel Green)++ | src/QRegex/P6Regex/Actions.nqp
Wrap QAST::Regex from qbuildsub in QAST::Stmts...

and add a `:node` parameter to `qbuildsub()` and pass in `$/` when it's called, this is then set as the `:node` argument to the new QAST::Stmts call. The combination of these changes means that we now get a filename set for more regex parts of a backtrace. E.g., `raku --ll-exception -e
  'grammar G { token TOP { $<foo>=(<xxx>) } }; G.parse("x")'` now reports
... (6 more lines)
20:08
nqp: ef246066be | (Elizabeth Mattijsen)++ (committed using GitHub Web editor) | src/QRegex/P6Regex/Actions.nqp
Merge pull request #732 from MasterDuke17/wrap_some_QAST__Regex_in_QAST__Stmts

Wrap QAST::Regex from qbuildsub in QAST::Stmts...
Geth roast: 98c86d6394 | (Daniel Green)++ | 2 files
Track changes cause by NQP PR #732

The TODO is now passing, and the new exception thrown is a better one.
20:10
roast: 89961ae172 | MasterDuke17++ (committed using GitHub Web editor) | 2 files
Merge pull request #750 from MasterDuke17/changes_for_nqp_pr_732

Track changes caused by NQP PR #732
Geth nqp: ca883aa21b | (Elizabeth Mattijsen)++ | tools/templates/MOAR_REVISION
Bump MoarVM for the latest goodies
20:18
rakudo: 853dc39369 | (Elizabeth Mattijsen)++ | tools/templates/NQP_REVISION
Bump NQP for the latest NQP / MoarVM goodies
20:35