[05:14] *** hurufu joined
[05:52] *** finanalyst joined
[07:19] <Geth> ¦ nqp/profile_sql_output_split_more_stuff: 33460348c4 | (Timo Paulssen)++ | src/vm/moar/HLL/Backend.nqp

[07:19] <Geth> ¦ nqp/profile_sql_output_split_more_stuff: Refactor sql output, split more of the statements we can output

[07:19] <Geth> ¦ nqp/profile_sql_output_split_more_stuff: 

[07:19] <Geth> ¦ nqp/profile_sql_output_split_more_stuff: nine++ tested sql profiler output with core setting compilation using

[07:19] <Geth> ¦ nqp/profile_sql_output_split_more_stuff: RakuAST, and the resulting profile had an "allocations" insert

[07:19] <Geth> ¦ nqp/profile_sql_output_split_more_stuff: statement that was 221 million characters long, which sqlite3 didn't

[07:19] <Geth> ¦ nqp/profile_sql_output_split_more_stuff: like parsing as a single statement.

[07:19] <Geth> ¦ nqp/profile_sql_output_split_more_stuff: 

[07:19] <Geth> ¦ nqp/profile_sql_output_split_more_stuff: This patch also has some attempts for re-using lists more.

[07:19] <Geth> ¦ nqp/profile_sql_output_split_more_stuff: review: https://github.com/Raku/nqp/commit/33460348c4

[07:19] <Geth> ¦ nqp/profile_sql_output_split_more_stuff: 47313b7a80 | (Timo Paulssen)++ | src/vm/moar/HLL/Backend.nqp

[07:19] <Geth> ¦ nqp/profile_sql_output_split_more_stuff: profiler sql: also split routines insert statements

[07:19] <Geth> ¦ nqp/profile_sql_output_split_more_stuff: review: https://github.com/Raku/nqp/commit/47313b7a80

[07:19] <timo> (just a rebase)

[07:20] <timo> the nqp azure pipeline doesn't seem quite right

[08:25] <Geth> ¦ rakudo/main: 8695c143eb | (Elizabeth Mattijsen)++ (committed using GitHub Web editor) | src/main.nqp

[08:25] <Geth> ¦ rakudo/main: Make RAKUDO_RAKUAST=0 *not* activate the RakUAST logic (#6214)

[08:25] <Geth> ¦ rakudo/main: 

[08:25] <Geth> ¦ rakudo/main: Spotted by timo++

[08:25] <Geth> ¦ rakudo/main: review: https://github.com/rakudo/rakudo/commit/8695c143eb

[08:37] *** finanalyst left
[08:43] <Geth> ¦ nqp/main: 4ff5ad7197 | (Timo Paulssen)++ (committed by timo) | src/vm/moar/HLL/Backend.nqp

[08:43] <Geth> ¦ nqp/main: Refactor sql output, split more of the statements we can output

[08:43] <Geth> ¦ nqp/main: 

[08:43] <Geth> ¦ nqp/main: nine++ tested sql profiler output with core setting compilation using

[08:43] <Geth> ¦ nqp/main: RakuAST, and the resulting profile had an "allocations" insert

[08:43] <Geth> ¦ nqp/main: statement that was 221 million characters long, which sqlite3 didn't

[08:43] <Geth> ¦ nqp/main: like parsing as a single statement.

[08:43] <Geth> ¦ nqp/main: 

[08:44] <Geth> ¦ nqp/main: This patch also has some attempts for re-using lists more.

[08:44] <Geth> ¦ nqp/main: review: https://github.com/Raku/nqp/commit/4ff5ad7197

[08:44] <Geth> ¦ nqp/main: 4043d6253b | (Timo Paulssen)++ (committed by timo) | src/vm/moar/HLL/Backend.nqp

[08:44] <Geth> ¦ nqp/main: profiler sql: also split routines insert statements

[08:44] <Geth> ¦ nqp/main: review: https://github.com/Raku/nqp/commit/4043d6253b

[09:45] *** finanalyst joined
[09:50] <Geth> ¦ rakudo/main: c15ee3dacf | (Elizabeth Mattijsen)++ | 12 files

[09:50] <Geth> ¦ rakudo/main: Change all TWEAK methods to submethods

[09:50] <Geth> ¦ rakudo/main: 

[09:50] <Geth> ¦ rakudo/main: To prevent them from being called more than once if a subclass is

[09:50] <Geth> ¦ rakudo/main: instantiated.

[09:50] <Geth> ¦ rakudo/main: 

[09:50] <Geth> ¦ rakudo/main: A special note about IO::Pipe: $.nl-out.list needed to be replaced

[09:51] <Geth> ¦ rakudo/main: with self.nl-out.list because the virtual $.nl-out attribute call

[09:51] <Geth> ¦ rakudo/main: <…commit message has 6 more lines…>

[09:51] <Geth> ¦ rakudo/main: review: https://github.com/rakudo/rakudo/commit/c15ee3dacf

[09:52] *** finanalyst left
[10:04] <Geth> ¦ rakudo/main: c46a0f13f7 | (Elizabeth Mattijsen)++ | tools/templates/NQP_REVISION

[10:04] <Geth> ¦ rakudo/main: Bump NQP to get SQL output refactor fix, timo++

[10:04] <Geth> ¦ rakudo/main: review: https://github.com/rakudo/rakudo/commit/c46a0f13f7

[10:04] *** lizmat left
[10:14] *** lizmat joined
[10:35] *** hurufu left
[12:04] *** finanalyst joined
[12:21] *** MasterDuke joined
[12:54] *** MasterDuke left
[14:14] *** librasteve_ joined
[15:21] <lizmat> bisectable6: old=2026.04 my @a; @a[1]=1; say @a.fmt

[15:21] <bisectable6> lizmat, Problem with c46a0f1 commit: Commit exists, but an executable could not be built for it

[15:21] <lizmat> bisectable6: old=2026.03 my @a; @a[1]=1; say @a.fmt

[15:21] <bisectable6> lizmat, Problem with c46a0f1 commit: Commit exists, but an executable could not be built for it

[15:22] <lizmat> bisectable6: old=2026.03 new=c15ee3dacf5 my @a; @a[1]=1; say @a.fmt

[15:22] <bisectable6> lizmat, On both starting points (old=2026.03 new=c15ee3d) the exit code is 0, exit signal is 11 (SIGSEGV) and the output is identical as well

[15:22] <bisectable6> lizmat, Output on both points: «»

[15:23] <lizmat> bisectable6: old=2025.01 new=c15ee3dacf5 my @a; @a[1]=1; say @a.fmt

[15:23] <bisectable6> lizmat, On both starting points (old=2025.01 new=c15ee3d) the exit code is 0, exit signal is 11 (SIGSEGV) and the output is identical as well

[15:23] <bisectable6> lizmat, Output on both points: «»

[15:23] <lizmat> wow

[15:23] <lizmat> bisectable6: old=2023.01 new=c15ee3dacf5 my @a; @a[1]=1; say @a.fmt

[15:23] <bisectable6> lizmat, Cannot find revision “2023.01” (did you mean “2026.01”?)

[15:23] <lizmat> bisectable6: old=2023.02 new=c15ee3dacf5 my @a; @a[1]=1; say @a.fmt

[15:23] <bisectable6> lizmat, Bisecting by exit signal (old=2023.02 new=c15ee3d). Old exit signal: 0 (None)

[15:27] <lizmat> bisectable6: help

[15:27] <lizmat> oh dear, I kiilled bisectable6 

[15:28] *** bisectable6 left
[15:31] <Geth> ¦ rakudo/main: 890f00c752 | (Nick Logan)++ (committed using GitHub Web editor) | 8 files

[15:31] <Geth> ¦ rakudo/main: RakuAST: handle custom package declarators like legacy does

[15:31] <Geth> ¦ rakudo/main: 

[15:31] <Geth> ¦ rakudo/main: `stub-package` mapped only the seven built-in declarators (`package`,

[15:31] <Geth> ¦ rakudo/main: `role`, `class`, `module`, `grammar`, `knowhow`, `native`) to their

[15:31] <Geth> ¦ rakudo/main: `RakuAST::Package` subclasses and fell back to bare `RakuAST::Package`

[15:31] <Geth> ¦ rakudo/main: otherwise. Custom declarators registered via `EXPORTHOW::DECLARE`

[15:31] <Geth> ¦ rakudo/main: (the way `OO::Monitors` adds `monitor`) became bare `RakuAST::Package`,

[15:31] <Geth> ¦ rakudo/main: <…commit message has 25 more lines…>

[15:31] <Geth> ¦ rakudo/main: review: https://github.com/rakudo/rakudo/commit/890f00c752

[15:45] *** greppable6 left
[15:45] *** evalable6 left
[15:45] *** shareable6 left
[15:45] *** bloatable6 left
[15:45] *** linkable6 left
[15:45] *** benchable6 left
[15:46] *** tellable6 left
[15:46] *** coverable6 left
[15:46] *** nativecallable6 left
[15:46] *** releasable6 left
[15:46] *** huggable6 left
[15:46] *** sourceable6 left
[15:48] *** evalable6 joined
[15:48] *** unicodable6__ joined
[15:48] *** benchable6 joined
[15:48] *** linkable6 joined
[15:48] *** quotable6 left
[15:48] *** notable6 left
[15:49] *** shareable6 joined
[15:49] *** sourceable6 joined
[15:49] *** unicodable6 left
[15:49] *** committable6 left
[15:49] *** greppable6 joined
[15:49] *** finanalyst left
[15:49] *** committable6 joined
[15:50] *** releasable6 joined
[15:50] *** bloatable6 joined
[15:50] *** bisectable6 joined
[15:50] *** coverable6 joined
[15:51] *** notable6 joined
[15:51] *** nativecallable6 joined
[15:51] *** tellable6 joined
[15:51] *** quotable6 joined
[15:52] *** huggable6 joined
[16:01] <lizmat> looks like https://github.com/rakudo/rakudo/issues/6241 is exposing a whole class of issues :-(

[16:01] <lizmat> m: my @a is default(42); @a[1] = 666; dd @a.shift

[16:02] <camelia> rakudo-moar 890f00c75: OUTPUT: «Nil  = Nil␤»

[16:02] <lizmat> m: my @a is default(42); @a[1] = 666; dd @a[0]; @a.shift

[16:02] <camelia> rakudo-moar 890f00c75: OUTPUT: «element of @a[0] = 42␤»

[16:02] <lizmat> m: my @a is default(42); @a[1] = 666; dd @a[0]; dd @a.shift

[16:02] <camelia> rakudo-moar 890f00c75: OUTPUT: «element of @a[0] = 42␤Nil  = Nil␤»

[16:04] <lizmat> m: my @a is default(42); my $b := @a[0]; @a[1] = 666; dd @a[0]; dd @a.shift

[16:04] <camelia> rakudo-moar 890f00c75: OUTPUT: «element of @a[0] = 42␤Nil  = Nil␤»

[17:09] *** finanalyst joined
[18:00] <Geth> ¦ rakudo/main: 7cb5a1939d | (Elizabeth Mattijsen)++ | src/core.c/Array.rakumod

[18:00] <Geth> ¦ rakudo/main: Create Array.fmt candidates with more precise handling

[18:00] <Geth> ¦ rakudo/main: 

[18:00] <Geth> ¦ rakudo/main: Specifically of uninitialized elements in the array, as show in

[18:00] <Geth> ¦ rakudo/main: https://github.com/rakudo/rakudo/issues/6241 .  It looks like

[18:00] <Geth> ¦ rakudo/main: 1617881dc2 introduced a shortcut that is fine for Lists, but

[18:00] <Geth> ¦ rakudo/main: are less then ideal for Arrays, which can have type constraints

[18:00] <Geth> ¦ rakudo/main: and default values.

[18:00] <Geth> ¦ rakudo/main: 

[18:00] <Geth> ¦ rakudo/main: This at least prevents the segfault.  However the underlying NQP

[18:00] <Geth> ¦ rakudo/main: code that allowed for the segfault to happen, is *not* fixed by

[18:00] <Geth> ¦ rakudo/main: this: this will require scrutiny of NQP and/or MoarVM.

[18:00] <Geth> ¦ rakudo/main: review: https://github.com/rakudo/rakudo/commit/7cb5a1939d

[18:41] *** finanalyst left
[18:56] <lizmat> m: my @a; @a[1]=1; @a.pop; @a.pop   # variation on a theme

[18:57] <camelia> rakudo-moar 7cb5a1939: OUTPUT: «(signal SEGV)»

[19:43] <Geth> ¦ rakudo/main: 415fa4b069 | (Elizabeth Mattijsen)++ | src/core.c/Array.rakumod

[19:43] <Geth> ¦ rakudo/main: Fix Array.pop on holes in arrays

[19:43] <Geth> ¦ rakudo/main: 

[19:43] <Geth> ¦ rakudo/main: The code "my @a; @a[1]=1; @a.pop; @a.pop" would actually segfault

[19:43] <Geth> ¦ rakudo/main: because of the hole in the array returning an nqp::null.  And this

[19:43] <Geth> ¦ rakudo/main: somehow triggered a segfault handling in nqp::null in HLL land.

[19:43] <Geth> ¦ rakudo/main: 

[19:43] <Geth> ¦ rakudo/main: This commit makes sure that the default value for the array is

[19:43] <Geth> ¦ rakudo/main: being returned (either from a type constraint, or the "is default"

[19:43] <Geth> ¦ rakudo/main: trait.

[19:43] <Geth> ¦ rakudo/main: review: https://github.com/rakudo/rakudo/commit/415fa4b069

[19:49] <Geth> ¦ rakudo/lizmat-11: f05f154491 | (Elizabeth Mattijsen)++ | src/core.c/Array.rakumod

[19:49] <Geth> ¦ rakudo/lizmat-11: Fix Array.shift on holes in arrays

[19:49] <Geth> ¦ rakudo/lizmat-11: 

[19:49] <Geth> ¦ rakudo/lizmat-11: Unlike the needed Array.pop fix in 415fa4b069 to prevent a segfault,

[19:49] <Geth> ¦ rakudo/lizmat-11: Array.shift was already guarded against returning an nqp::null.

[19:49] <Geth> ¦ rakudo/lizmat-11: 

[19:49] <Geth> ¦ rakudo/lizmat-11: However, it was set to return Nil, rather than the default value

[19:49] <Geth> ¦ rakudo/lizmat-11: of the array.  This commit changes that to be more in line with

[19:49] <Geth> ¦ rakudo/lizmat-11: <…commit message has 10 more lines…>

[19:49] <Geth> ¦ rakudo/lizmat-11: review: https://github.com/rakudo/rakudo/commit/f05f154491

[19:49] <Geth> ¦ rakudo: lizmat++ created pull request #6242: Fix Array.shift on holes in arrays

[19:49] <Geth> ¦ rakudo: review: https://github.com/rakudo/rakudo/pull/6242

[19:53] *** lizmat left
[19:54] *** lizmat joined
[20:04] *** sivoais joined
[20:10] <Geth> ¦ nqp: ugexe++ created pull request #852: HLL::Compiler.compile: balance backend per-compile state on early exit

[20:10] <Geth> ¦ nqp: review: https://github.com/Raku/nqp/pull/852

[21:08] <Geth> ¦ rakudo: ugexe++ created pull request #6243:  RakuAST: route string EVAL through IMPL-FIXUP and wrap nested EVAL output

[21:08] <Geth> ¦ rakudo: review: https://github.com/rakudo/rakudo/pull/6243

[21:28] <ugexe> Those two were exposed by inline perl5 

[21:34] <japhb> I wonder if Inline::Perl5 still works with latest perl5

[21:34] <ugexe> mind you inline::perl5 won't actually work without changes to the distribution itself since it uses e.g. $*W

[21:34] <japhb> Well that answers that

[21:35] <japhb> "No, but not because of not being up to date with perl5 internals"

[21:35] * lizmat hopes that nine either allows it to become a community module, or helps / adds RakuAST support

[22:08] <japhb> Both?  Both.  Both is good.

[22:18] <ugexe> it also uses a slang

[23:53] *** librasteve_ left