travis-ci Rakudo build passed. Christian Bartolomäus 'Merge pull request #4044 from usev6/jvm_gh4041 01:36
travis-ci.org/rakudo/rakudo/builds/744498417 github.com/rakudo/rakudo/compare/f...74f3c56295
Geth roast: f5db1287bd | (Vadim Belman)++ | S12-coercion/coercion-types.t
Fix incorrect coercion spectest

  `1` and `e` are incorrect values for Rat(Str). Replaced with `1.0` and
  `<2.71>`. The latter as an additional measure to make sure allomorphs
are treated correctly.
Also added tests for errors when invalid coerctions are attempted.		 02:13
roast: 20db11fa7d | (Vadim Belman)++ (committed using GitHub Web editor) | S12-coercion/coercion-types.t
Merge pull request #700 from vrurg/rakudo_4040

Fix incorrect coercion spectest
travis-ci Rakudo build failed. Vadim Belman 'Merge pull request #4043 from vrurg/rakudo_4040 02:50
travis-ci.org/rakudo/rakudo/builds/744511853 github.com/rakudo/rakudo/compare/e...314947ea9e
Geth rakudo: vrurg++ created pull request #4046:
Provide more improvements to language revision support and more.		 03:48
roast: vrurg++ created pull request #701:
Test for the right exception and don't disable mixins		 03:52
Geth rakudo: Kaiepi++ created pull request #4047:
Type IO::Socket::INET family/type/protocol values		 05:41
Kaiepi there are more bits and pieces of problem-solving #111's solution like ^ that don't need to be done in one big pr 05:42
Geth roast: Kaiepi++ created pull request #702:
Add a test for invalid Int:D families for IO::Socket::INET.new		 06:16
releasable6 Next release in ≈2 days and ≈11 hours. 2 blockers. Please log your changes in the ChangeLog: github.com/rakudo/rakudo/wiki/ChangeLog-Draft 07:00
bartolin oh noes, more test breakage on the jvm backend. *looking* 07:56
Altai-man bartolin++ 07:59
bartolin looks like another case of BOOTStr blowing up where Str is expected. So hopefully not too hard to fix ;) 08:00
lizmat Files=1340, Tests=114243, 228 wallclock secs (30.24 usr 8.66 sys + 3164.45 cusr 300.10 csys = 3503.45 CPU) 08:53
Geth roast: a85a8cfcb1 | (Christian Bartolomäus)++ | S06-currying/positional.t
[JVM] Unfudge tests that pass again

Fixed with github.com/rakudo/rakudo/pull/4044		 10:05
Xliff \o 11:20
patrickb: Unfortunately, I haven't seen much improvement. Using zef 0.9.4 11:21
tellable6 Xliff, I'll pass your message to patrickb
Geth rakudo: usev6++ created pull request #4048:
Avoid BOOTStr being passed to Version.new		 11:43
rakudo: c015759bc9 | (Christian Bartolomäus)++ | src/Perl6/World.nqp
Avoid BOOTStr being passed to Version.new

This fixes 'use v6.e.PREVIEW' on the JVM backend. Without the patch it dies with X::Coerce::Impossible ("Impossible coercion from
  'BOOTStr' into 'Str': value is of unacceptable type BOOTStr").		 11:45
rakudo: 5f0da2bdf5 | (Elizabeth Mattijsen)++ (committed using GitHub Web editor) | src/Perl6/World.nqp
Merge pull request #4048 from usev6/bootstr_v6.e.preview

Avoid BOOTStr being passed to Version.new
Geth rakudo: e612214ca7 | (Elizabeth Mattijsen)++ | 2 files
Make setting array shapes more alike

Between native arrays and "normal" arrays. This probably shows there's more room for simplification, but leaving this as is for now.
This also allows 'my int @a[*]' to work, just like 'my @a[*]' did, which was the impetus for this change.		 12:07
lizmat m: my @a[5] = ^5; dd @a; @a = (); dd @a 12:26
camelia Array element = Array.new(:shape(5,), [0, 1, 2, 3, 4])
Array element = Array.new(:shape(5,), [Any, Any, Any, Any, Any])
lizmat m: my int @a[5] = ^5; dd @a; @a = (); dd @a # subtly different behaviour 12:27
camelia array[int].new(:shape(5,), [0, 1, 2, 3, 4])
array[int].new(:shape(5,), [0, 1, 2, 3, 4])
lizmat well, maybe not so subtle
m: my @a[5] = ^5; dd @a; @a = 42,666; dd @a # similarly
camelia Array element = Array.new(:shape(5,), [0, 1, 2, 3, 4])
Array element = Array.new(:shape(5,), [42, 666, Any, Any, Any])
lizmat m: my int @a[5] = ^5; dd @a; @a = 42,666; dd @a # similarly
camelia array[int].new(:shape(5,), [0, 1, 2, 3, 4])
array[int].new(:shape(5,), [42, 666, 2, 3, 4])
lizmat another set of differences :-( : 12:37
m: my @a[5]; @a = ^6
camelia Index 5 for dimension 1 out of range (must be 0..4)
in block <unit> at <tmp> line 1
lizmat m: my int @a[5]; @a = ^6
camelia This type (List) does not support positional operations
in block <unit> at <tmp> line 1
lizmat seems there is a little work to do :-(
Altai-man ^^^ probably won't be ready for this release 12:38
Altai-man lizmat, is it a regression? 12:41
lizmat don't think so
native shaped arrays are just woefully undertested atm
Altai-man bisectable6, my int @a[5]; @a = ^6
bisectable6 Altai-man, Will bisect the whole range automagically because no endpoints were provided, hang tight
lizmat is writing tests atm
bisectable6 Altai-man, Output on all releases: gist.github.com/bb4239e1875d2d9832...e68a9ed983 12:42
Altai-man, bisect log: gist.github.com/0ef22efcd87545bb14...9f657e390b
Altai-man, bisect log: gist.github.com/bac9ea59fd7e00ce82...47283bc201
Altai-man, Output on all releases and bisected commits: gist.github.com/1773081d13d08841f3...ba4bdab2df
Altai-man lizmat, not a blocker for sure. Kudos for writing tests and working with this!
lizmat I think actually that the correct behaviour before 2016.12 was because "int" was syntactically accepted, but didn't do anytihng 12:43
so I think it's been like this from the beginning
lizmat m: dd Compiler.new.version 13:50
camelia v2020.08.2.51.g.85847.d.2.f.1
lizmat wow 13:50
moritz ^^^
lizmat looks like Camelia is behind a few releases :-( 13:50
bisectable6: old=2020.08 my int @a[2,2] = (1,2),(3,4); dd @a.unique 13:51
bisectable6 lizmat, Bisecting by exit code (old=2020.08 new=e612214). Old exit code: 0
lizmat, bisect log: gist.github.com/f80c7960b7c43a6464...d4becd7835
lizmat, (2020-10-23) github.com/rakudo/rakudo/commit/c3...8004bace56
lizmat yuck 13:54
that *is* a regression
Altai-man lizmat, a released one... :) 14:00
Geth__ rakudo: 78b03b6260 | (Elizabeth Mattijsen)++ | 2 files
Disable optimized .unique/repeated/squish on native arrays

At least for now just before the 2020.11 release. Introduced in c37ffe71bd19f1a1, it broke those methods on native *shaped* arrays. Unfortunately, not caught by tests :-(		 14:25
travis-ci Rakudo build passed. Elizabeth Mattijsen 'Merge pull request #4048 from usev6/bootstr_v6.e.preview 14:37
travis-ci.org/rakudo/rakudo/builds/744637544 github.com/rakudo/rakudo/compare/6...0da2bdf50d
travis-ci Rakudo build passed. Elizabeth Mattijsen 'Make setting array shapes more alike 16:30
travis-ci.org/rakudo/rakudo/builds/744640454 github.com/rakudo/rakudo/compare/5...12214ca75c
lizmat aaw Geth awol :-( 16:47
tyil ^^ 16:50
tyil lizmat: aye, hetzner (where the servers are located) is having some troubles it looks like 16:51
I'll have to wait until the servers are back up to bring geth back
lizmat yeah, sure... no pb :-) 16:52
nine How on earth can a 'return False;' lead to a Type check failed for return value; expected Bool:D but got ForeignCode 16:54
moritz it somebody lexically override False? :D 16:55
nine if only it were something simple like that 16:56
lizmat hmmm... feels like a new coercion protocol issue? 17:03
m: dd Compiler.new.version
camelia v2020.08.2.51.g.85847.d.2.f.1
lizmat moritz ^^
seems like camelia is stuck in september :-( 17:04
moritz lizmat: I'm taking a look 17:09
lizmat ++moritz!
Geth___ roast: 9d3ca2a4a4 | (Elizabeth Mattijsen)++ | 2 files
Initial set of tests one dim native shaped num arrays
 17:10
lizmat also ba0d5fe53bf854875358
linkable6 (2020-11-19) github.com/Raku/roast/commit/ba0d5fe53b Initial set of tests one native shaped int arrays
nine It's a LEAVE phaser 17:17
timotimo make like a tree and get out of here! 17:24
nine Soon....training starts in 20 minutes 17:25
dogbert17 the new spectest files contains errors, is it WIP? 17:28
lizmat: ^^ 17:29
lizmat double checks 17:30
they're clean for me?
aaah... 17:31
oops
lizmat hopes Geth gets it 17:32
ah, it did not
db3ca28e3eaabcc8f0c479
linkable6 (2020-11-19) github.com/rakudo/rakudo/commit/db3ca28e3e Handle :exists/delete on 1dim native shaped arrays with int index
lizmat dogbert17: ^^ that should make the spectest clean again 17:33
dogbert17 that was a very quick fix :-)
lizmat dogbert17: I already had the fix, I had just forgot to push it :-) 17:34
found the issue while writing these tests 17:35
nine so spectests should be fine again? 17:36
dogbert17 it looks that way
lizmat nine: yeah, they should be fine 17:41
moritz m: dd Compiler.new.version 17:45
camelia v2020.10.130.g.78.b.03.b.626
lizmat moritz++ 17:58
Geth_ roast: 1d79e5c9e7 | (Elizabeth Mattijsen)++ | 2 files
Initial set of tests one dim native shaped str arrays
 18:00
Geth_ roast: b292749d13 | (Elizabeth Mattijsen)++ | 2 files
Fix some copy-pastos in comments
 18:16
lizmat github.com/Raku/Documentable/issues/145 # sena_kun not sure that should be considered a blocker or not 18:35
afk& 18:38
jjatria gugod: Do you think you could bump the version of Cache::LRU? I keep wanting to use associative indexing on it and being unable to, even though the change is already on master 18:41
Oh damn, that's not yours. Sorry! 18:42
nwc10 m: Num(0.777777777777777777777).nude 18:48
camelia No such method 'nude' for invocant of type 'Num'. Did you mean any of
these: 'none', 'note', 'Num'?
in block <unit> at <tmp> line 1
nwc10 oops
m: Rat(0.777777777777777777777).nude
camelia ( no output )
nwc10 pah. I'm messing this up...
timotimo nwc10: send nudes to stdout
nwc10 m: say Rat(0.777777777777777777777).nude
camelia (777777777777777777777 1000000000000000000000)
nwc10 thanks
timotimo snickers like the 13 year old that he is 18:52
moritz a 13 year old snickers likely doesn't taste that good anymore :D 18:55
nine snickers doesn't that as good as when I was 13 years old? 18:56
nwc10 I'm not sure what base he is being base in 18:58
timotimo i'm failing to parse what nine wrote
nine snickers doesn't taste as good as when I was 13 years old? 19:01
That's what I wanted to type
travis-ci Rakudo build passed. Elizabeth Mattijsen 'Disable optimized .unique/repeated/squish on native arrays 19:29
travis-ci.org/rakudo/rakudo/builds/744663480 github.com/rakudo/rakudo/compare/e...b03b6260e3
Geth_ rakudo/master: 6 commits pushed by (Vadim Belman)++ 20:07
roast: a8e49d6c95 | (Vadim Belman)++ | S14-roles/versioning.t
Test for the right exception and don't disable mixins

Follow up to rakudo/rakudo#4046. Make the test check for
  `X::Language::IncompatibleRevisions` exception instead of AdHoc.
And don't expect cross-language mixins to die because they don't now.		 20:08
linkable6 RAKUDO#4046 [closed]: github.com/rakudo/rakudo/pull/4046 Provide more improvements to language revision support and fix mixins once more.
roast: 132f095ef3 | (Vadim Belman)++ (committed using GitHub Web editor) | S14-roles/versioning.t
Merge pull request #701 from vrurg/rakudo_4021

Test for the right exception and don't disable mixins
Geth_ rakudo: usev6++ created pull request #4049:
[JVM] Test if method 'coercive' is available		 20:35
[Coke] (snickers) get the teeny ones and freeze them! 20:41
Geth_ rakudo: dumarchie++ created pull request #4050:
Optimize some array operators		 22:04

Powered by Cro and the Raku® Programming Language.

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

Thank you!