00:25 oddp left 00:32 Altai-man_ joined 00:34 sena_kun left 00:35 vrurg_ joined, vrurg_ left 00:38 vrurg_ joined 00:47 vrurg left 00:56 vrurg_ left 00:57 vrurg joined 01:10 leont left 01:28 yuplushi left 01:52 yuplushi joined 02:13 lucasb left 02:33 sena_kun joined 02:34 Altai-man_ left 03:33 ShimmerFairy left 04:23 camelCaser left 04:26 camelCaser joined 04:32 Altai-man_ joined 04:34 sena_kun left 05:38 krunen left, krunen joined 06:28 patrickb joined 06:33 sena_kun joined 06:34 Altai-man_ left 06:35 patrickb left 06:44 Xliff joined
Xliff o/ 06:47
tellable6 2020-07-20T18:22:30Z #raku <vrurg> Xliff try PM me when available.
Xliff .tell vrurg For you? Available always. Email me at [email@hidden.address] if I'm not active on IRC
tellable6 Xliff, I'll pass your message to vrurg
Xliff What's the best way to assign a CArray[uint8] from a String? 06:53
s/String/Str/ 06:54
moritz I'm not certain, but it likely involves calling .encode on the Str 06:55
Xliff my $c = $string.encode; @!a[$_] = $c[$_] for ^$c.elems 07:05
??
Was hoping I could avoid the iteration.
07:09 patrickb joined
patrickb o/ 07:10
lizmat, ugexe, timotimo: I'm backlogging at the moment and saw yesterdays discussion wrt merges breaking master. Can the experience be destilled into a proposal for a sane policy of how we want to do merges? Something as safe as necessary and as easy as possible. 07:14
I want to reignite the discussion in that ticket in the hopes it will be fruitful. 07:15
07:22 squashable6 left 07:23 squashable6 joined 07:55 patrickb left
[Tux] Rakudo version 2020.07-26-g1eb712d5f - MoarVM version 2020.07-3-gb56e62f18
csv-ip5xs0.926 - 0.991
csv-ip5xs-208.717 - 9.108
csv-parser27.177 - 29.127
csv-test-xs-200.438 - 0.447
test9.058 - 9.297
test-t2.095 - 2.234
test-t --race0.954 - 0.997
test-t-2036.550 - 36.725
test-t-20 --race11.031 - 11.783
 07:57
Was also in a teams-meeting. Now killed teams and running again 08:05
08:32 Altai-man_ joined 08:34 sena_kun left 08:47 oddp joined
[Tux] Rakudo version 2020.07-26-g1eb712d5f - MoarVM version 2020.07-3-gb56e62f18
csv-ip5xs0.821 - 0.824
csv-ip5xs-207.827 - 7.886
csv-parser24.497 - 26.737
csv-test-xs-200.392 - 0.402
test7.526 - 7.661
test-t1.855 - 1.878
test-t --race0.897 - 0.935
test-t-2031.174 - 33.705
test-t-20 --race9.198 - 9.519
 08:48
09:45 leont joined
lizmat Files=1308, Tests=113043, 216 wallclock secs (29.36 usr 8.71 sys + 3012.16 cusr 289.13 csys = 3339.36 CPU) 10:31
10:33 sena_kun joined 10:34 Altai-man_ left
Geth_ rakudo: 4ad4bd303f | (Patrick Böker)++ (committed using GitHub Web editor) | azure-pipelines.yml
Fix release pipeline to not delete the linux build (#3814)

Fix release pipeline to not delete the linux build		 11:08
linkable6 RAKUDO#3814 [closed]: github.com/rakudo/rakudo/pull/3814 Fix release pipeline to not delete the linux build
11:10 oddp left 11:11 MasterDuke left 11:37 JJMerelo joined
JJMerelo Here's us proposing a "20th anniversary calendar", with 20 articles that pick up Perl 6 RFCs, comment them, pay homage to authors and write up how they ended up in Raku with a small example github.com/Raku/advent/tree/master/20th 11:37
Please propose yourself with a PR or directly if you've got the privs. 11:38
Geth_ roast: patrickbkr++ created pull request #655:
Test for empty string match with ignore-case and ignore-mark		 11:57
12:01 patrickb joined
patrickb o/ 12:01
lizmat: I'd like your feedback on a PR: github.com/rakudo/rakudo/pull/3810...6a12c6fR57
lizmat: I'm not sure this is the right way to do a deprecation.
lizmat patrickb this would be the first deprecation in configuration land for me as well, so I actually have no idea 12:06
patrickb lizmat: I'm not worried about our config system. I'm worried about users of that option in the wild. 12:11
To be honest I think there won't be any, because I think that option was primarily added as a tool to help in our build process, but still...
12:32 patrickb left, Altai-man_ joined 12:34 sena_kun left
lizmat it's terra incognita I guess 12:35
12:40 unclechu left
jnthn JJMerelo: ooh, that's an interesting idea 12:40
12:40 AlexDaniel` left
jnthn I'm really curious a) if there's one about macros, b) if I can really have a macros impl that's worth something by the time I have to write the article :) 12:41
12:46 unclechu joined 13:01 AlexDaniel` joined 13:08 Xliff left, MasterDuke joined 13:16 oddp joined 13:39 JJMerelo left 13:46 MasterDuke left 14:33 sena_kun joined 14:34 Altai-man_ left 15:44 JJMerelo joined 16:22 MasterDuke joined
Geth_ rakudo: b60e16367f | (Timo Paulssen)++ | src/core.c/Rakudo/Internals/JSON.pm6
import lizmat's rewrite of JSON::Fast's from-json, 4.4x as fast
 16:22
sena_kun There are no backward problems? 16:23
timotimo the example from the "fix nested arrays" pull request doesn't change behaviour 16:24
see #raku
also, we'll have an entire month to see if it causes problems in the ecosystem :) :)
rakudo tests and spec tests are clean
though spec tests of course don't test Rakudo::Internals::JSON
not directly, anyway
sena_kun I see.
timotimo m: my \a = from-json('[]'); say a.elems; say a[0].elems; 16:25
camelia 0
Saw 1 occurrence of deprecated code.
================================================================================
Sub from-json (from GLOBAL) seen at:
<tmp>, line 1
Please use JSON::Fast, JSON::Tiny or JSON::Pretty from modul…»
timotimo m: my \a = from-json('[]'); say a.elems ~ " and " ~ say a[0].elems;
camelia 1
Saw 1 occurrence of deprecated code.
================================================================================
Sub from-json (from GLOBAL) seen at:
<tmp>, line 1
Please use JSON::Fast, JSON::Tiny or JSON::Pretty from modul…»
timotimo oops
sena_kun I mean, that'd be very cool to have JSON::Fast 4x as fast, but having a non-compatible implementation which does that is not so cool.
timotimo m: my \a = from-json('[]'); say a.elems ~ " and " ~ a[0].elems;
camelia 0 and 1
Saw 1 occurrence of deprecated code.
================================================================================
Sub from-json (from GLOBAL) seen at:
<tmp>, line 1
Please use JSON::Fast, JSON::Tiny or JSON::Pretty from https:/…
timotimo just need to make sure this stays the same afterwards 16:26
the from-json already was from JSON::Fast, and the new impl passes all tests there
sena_kun releasable6, status 16:28
releasable6 sena_kun, Next release in ≈31 days and ≈2 hours. There are no known blockers. Changelog for this release was not started yet
sena_kun, Details: gist.github.com/3e7d299e688c047b35...c1fbc77f20
lizmat sena_kun: the from-json rewrite passed all JSON::Fast tests
which the previous version in Rakudo::Internals did also 16:29
timotimo++ # for importing that :-)
sena_kun lizmat, the improvements are very cool and impressive. :)
timotimo may also be able to steal newer to-json impls, if the necessary adaptations (like support for Version) are applied as well 16:30
lizmat sena_kun: most of the improvement comes from not relying on regexes anymore
timotimo can't wait for JSON::Tiny to perform as well as JSON::Fast 16:31
lizmat the new algorithm is basically: when a " is encountered, look for the next " that has an even number of \ in front of it
then convert that to NFD including the opening "
drop the first codepoint (this gets rid of the opening " but *not* of its possible decorations)
there used to be quite a lot of code to handle that, it became very easy to do once it was NFD :-) 16:32
16:32 Altai-man_ joined
lizmat then handle all escapes using integer arithmetic, and convert the resulting NFD back to a string at the end 16:32
16:34 sena_kun left
timotimo feel very free to prove me wrong about the @ = from-json('[]') thing 16:37
m: say "hi"
camelia hi
timotimo m: my \a = from-json('[]'); say a.elems ~ " and " ~ a[0].elems;
camelia 0 and 1
Saw 1 occurrence of deprecated code.
================================================================================
Sub from-json (from GLOBAL) seen at:
<tmp>, line 1
Please use JSON::Fast, JSON::Tiny or JSON::Pretty from https:/…
timotimo getting the same output between these two version 16:38
Altai-man_ weekly: twitter.com/koto_san_kana/status/1...1466088449 16:43
notable6 Altai-man_, Noted! (weekly)
nine This is a really odd build failure that crops up from time to time: build.opensuse.org/package/live_bu...RM/aarch64 16:51
lizmat Altai-man_: retwat 16:54
16:54 MasterDuke left 17:37 JJMerelo left 17:44 tyil left, tyilanmenyn joined 17:45 tyilanmenyn is now known as tyil 18:10 lucasb joined 18:27 Kaiepi left, Kaiepi joined 18:33 sena_kun joined 18:34 Altai-man_ left 18:38 Kaiepi left, Kaiepi joined 19:04 lichtkind joined 19:22 MasterDuke joined 19:41 MasterDuke left 20:29 squashable6 left 20:31 squashable6 joined, Xliff joined 20:33 Altai-man_ joined 20:34 sena_kun left
Xliff \o 20:36
21:06 Xliff left
timotimo Kaiepi: you want to elaborate a little on what exactly github.com/rakudo/rakudo/pull/2968 needed, i.e. to check if my last push to RI's from-json broke this again? maybe a test or two that we put in rakudo's t/ folder? 21:36
22:16 lichtkind left
Kaiepi timotimo, iirc the bug there was that assignments were being made where binding would've been more appropriate 22:22
my @ = from-json '[]' would return [[]] before, but [] with this 22:23
timotimo is this test code the right way to check it? 22:25
m: my \a = from-json('[]'); say a.elems ~ " and " ~ a[0].elems;
camelia 0 and 1
Saw 1 occurrence of deprecated code.
================================================================================
Sub from-json (from GLOBAL) seen at:
<tmp>, line 1
Please use JSON::Fast, JSON::Tiny or JSON::Pretty from https:/…
timotimo c: 2018.12 my \a = from-json('[]'); say a.elems ~ " and " ~ a[0].elems;
committable6 timotimo, gist.github.com/e0de404a2e5a087e16...39f859b273
timotimo hum 22:26
c: releases my \a = from-json('[]'); say a.elems ~ " and " ~ a[0].elems;
committable6 timotimo, gist.github.com/368f8876cdb6f4950b...f632e803ec
timotimo Kaiepi: this code is not sufficient to get the problem 22:27
help
c: releases my @a = from-json('[]'); say @a.elems ~ " and " ~ @a[0].elems;
committable6 timotimo, gist.github.com/876adfc2c64f1cab30...b1f1cce243
Kaiepi that last one should work 22:30
timotimo m: my @a = from-json('[]'); say @a.elems ~ " and " ~ @a[0].elems; dd @a;
camelia 0 and 1
Array @a = []
Saw 1 occurrence of deprecated code.
================================================================================
Sub from-json (from GLOBAL) seen at:
<tmp>, line 1
Please use JSON::Fast, JSON::Tiny or JSON::Pre…
timotimo Array @a = [], does that mean we have [] or [[],]?
22:35 Altai-man_ left
lizmat hmmm.. did I optimize away an itemization ? 22:35
timotimo the test suite should have shown that though?
lizmat apparently not ? 22:36
timotimo if not, that's a problem in the test suite, but also gives us carte blanche to break user code and say "we never guaranteed it to work this way!"
lizmat perhaps the test-suite needs more is-deeply
maybe we need to do a Blin run for modules depending on JSON::Fast ?
timotimo almost all of them :D
lizmat well, then that should be easy :-) 22:37

Powered by Cro and the Raku® Programming Language.

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

Thank you!