03:33 nine_ joined, nine left, camelia left 03:38 camelia joined 07:28 MasterDuke left
|Tux| Rakudo v2024.08-60-gdde756878 (v6.d) on MoarVM 2024.08-9-g570db2612
csv-ip5xs0.257 - 0.258
csv-ip5xs-201.082 - 1.096
csv-parser1.467 - 1.483
csv-test-xs-200.141 - 0.141
test1.846 - 1.899
test-t0.401 - 0.404
test-t --race0.266 - 0.268
test-t-204.776 - 4.804
test-t-20 --race1.131 - 1.194
08:02
tux.nl/Talks/CSV6/speed4-20.html / tux.nl/Talks/CSV6/speed4.html tux.nl/Talks/CSV6/speed.log
same versions same values 08:03
08:32 sena_kun joined 09:05 sena_kun left
timo m: say Metamodel::ClassHOW; Metamodel::ClassHOW.gist; Metamodel::ClassHOW.Str; 11:15
camelia (ClassHOW)
Use of uninitialized value of type Perl6::Metamodel::ClassHOW in string context.
Methods .^name, .raku, .gist, or .say can be used to stringify it to something meaningful.
in block <unit> at <tmp> line 1
timo m: given Metamodel::ClassHOW { say .Str; say .gist; } 11:16
camelia Use of uninitialized value of type Perl6::Metamodel::ClassHOW in string context.
Methods .^name, .raku, .gist, or .say can be used to stringify it to something meaningful.

(ClassHOW)
in block at <tmp> line 1
timo m: given Metamodel::CoercionHOW { say .Str; say .gist; }
camelia Method Perl6::Metamodel::CoercionHOW.Str not found
in block <unit> at <tmp> line 1
timo DIHWIDT or LTA? 11:17
gist.github.com/timo/55471f4649415...b85247ce8a 11:25
ParametricRoleHOW, CurriedRoleHOW, Primitives, and EnumHOW are okay with calling .gist on them, Metamodel::Primitives is the only one that likes having .Str called on it 11:26
lizmat LTA I'd say 11:57
notable6__: weekly
notable6__ lizmat, 1 note: 2024-09-14T16:52:29Z <ab5tract>: weekly - 5ab5traction5.bearblog.dev/introdu...-comma-20/
lizmat notable6__: weekly reset
notable6__ lizmat, Moved existing notes to “weekly_2024-09-17T11:57:44Z”
timo i was looking for documentation on COERCE and found no entry for COERCE, and the documentation in docs.raku.org/language/signatures#Coercion_type doesn't mention the ability to define a COERCE method, or that method new is also tried by default 12:04
also, the documentation claims that the method that's called by name on the incoming object doesn't have its return type checked, which doesn't seem correct: 12:05
m: class Gimme { }; class Tester { method Gimme() { "lol" } }; sub test(Gimme() $foo) { say $foo.raku }; test(Tester.new()) 12:06
camelia Impossible coercion from 'Tester' into 'Gimme': method Gimme returned an instance of Str
in sub test at <tmp> line 1
in block <unit> at <tmp> line 1
timo bisectable6: say 99 ~~ Mu:U 12:19
bisectable6 timo, Will bisect the whole range automagically because no endpoints were provided, hang tight
timo, ¦6c (83 commits): «False␤»
timo, Nothing to bisect!
timo oops, got that wrong 12:23
m: $_ = 42; say $_ ~~ Mu:U 12:24
camelia True
timo m: say $_ ~~ Mu:U
camelia True
timo m: say $_ ~~ Any:U 12:25
camelia True
timo m: $_ = 42; say $_ ~~ Any:U
camelia False
timo bisectable6: $_ = 42; say $_ ~~ Mu:U
bisectable6 timo, Will bisect the whole range automagically because no endpoints were provided, hang tight
timo, Output on all releases: gist.github.com/e974c8ed07bf500c50...237ff70bdf 12:26
timo, Bisecting by output (old=2021.12 new=2022.02) because on both starting points the exit code is 0
lizmat and yet another Rakudo Weekly News hits the Net: rakudoweekly.blog/2024/09/17/2024-...omma-plug/ 12:27
bisectable6 timo, bisect log: gist.github.com/22aefb15fea0bad22a...8571829084 12:36
timo, Output on all releases and bisected commits: gist.github.com/2dc430073cdc43aca2...97e6d1c0f5
12:52 [Tux] left 13:12 [Tux] joined 14:29 andinus2 left 14:30 andinus joined
timo linkable6: 2a3f61ef14acb864207734049128139e698d8575 14:37
linkable6 (2022-01-03) github.com/rakudo/rakudo/commit/2a3f61ef14 Additional fixes and improvements of smartmatch optimizations
timo potential candidate for the change
linkable6: f21af28e37e7ed407fdf6fb0d5f837b2ac377877 # as well 14:38
linkable6 (2022-01-09) github.com/rakudo/rakudo/commit/f21af28e37 Add ACCEPTS candidate for Junction to Bool
timo 91ec1bbf601f2b9060a3ffc481407508296a481d # does it work without addressing the bot? 14:40
linkable6 (2021-12-11) github.com/rakudo/rakudo/commit/91ec1bbf60 Implement optimization of topicalized smartmatches
timo 32401c4762a18c98b3d5b0bdd7c03b27400cb521 # this was the rewrite of smartmatch as a dispatcher, that could have introduced a change like that 14:44
linkable6 (2022-01-15) github.com/rakudo/rakudo/commit/32401c4762 Implement smartmatch as a dispatcher
timo buildd.debian.org/status/fetch.php...&raw=0 we have a rakudo build on an arm64 debian build server that got stuck during the testing phase, it looks like. the output isn't terribly enlightening 15:16
ab5tract Why is it trying to regex replace a line in Compiler.pm6? Not that I expect this to be relevant to the stalling build, but I’m curious 17:26
17:41 sena_kun joined
[Coke] is that the distribution build star was asking for? 17:54
ah, no that is probably an old workaround to get the same effect. 17:56
timo > make test || make test || make test || make test || MVM_SPESH_DISABLE=1 make test 18:25
:S
[Coke] ?? 18:36
timo from the build commands 18:44
retry it three times, and a fifth time without spesh just to make sure 18:45
ab5tract I noticeds that too :) 18:51
timo it's a little bit of a "lolsob" 19:00
Bisecting: 9 revisions left to test after this (roughly 3 steps) 19:29
91ec1bbf601f2b9060a3ffc481407508296a481d is the commit that changed how this code works: 19:38
linkable6 (2021-12-11) github.com/rakudo/rakudo/commit/91ec1bbf60 Implement optimization of topicalized smartmatches
timo m: my $_ = 42; $_ ~~ Mu:U
camelia Potential difficulties:
Redeclaration of symbol '$_'.
at <tmp>:1
------> my $_⏏ = 42; $_ ~~ Mu:U
timo m: $_ = 42; $_ ~~ Mu:U 19:39
camelia ( no output )
timo m: $_ = 42; say $_ ~~ Mu:U
camelia True
timo it flips behaiour with an up-to-date rakudo if you --optimize=off 19:41
19:58 sena_kun left
ab5tract m: Q| my $_ = 42; say $_ ~~ Mu:U |.AST.EVAL 20:48
camelia False
ab5tract #FixedInRakuAST 20:49
timo yeah because rakuast doesn't have an optimizer yet ;) ;)
ab5tract ah, I always forget about that 20:50
it's a bit unfortunate how often that's the case, though
timo it looks like somehow it thinks the LHS, i.e. $_, has a value and the value is of type Mu and undefined 21:00
22:39 sena_kun joined 22:41 sena_kun left