MasterDuke .seen AlexDaniel 00:35
yoleaux2 I saw AlexDaniel 6 Jul 2016 10:57Z in #perl6-dev: <AlexDaniel> if given enough permissions to work on stuff, I can do it. And I think that there are others who will happily help with it. The problem is that we can't decide what we should use instead of RT
travis-ci Rakudo build errored. Elizabeth Mattijsen 'Revert "preparing for upcoming changes:" 00:54
travis-ci.org/rakudo/rakudo/builds/143451638 github.com/rakudo/rakudo/compare/8...0704d2ea8b
Zoffix .ask jnthn are you able to comment on this ticket? The proposal is to use multi-dispatch instead of separate subs to be consistent with other subs in Test.pm6 and liz's comments suggest I should ask your opinion: rt.perl.org/Ticket/Display.html?id...et-history 03:32
yoleaux2 Zoffix: I'll pass your message to jnthn.
dalek ast: 1c4a4f7 | (Zoffix Znet)++ | S02-types/version.t:
v6 versions must compare correctly against other versions

RT #128408		 03:58
synopsebot6 Link: rt.perl.org/rt3//Public/Bug/Displa...?id=128408
Zoffix For core members to review: liz already commented that fixing this would be bloat. Should the ticket be closed as wontfix? rt.perl.org/Ticket/Display.html?id...et-history 04:12
I'm actually finding quite a few tickets I'm marking as 'Needs core member decision'. I guess I'll keep a list and ping about a couple a day: i.imgur.com/aveZaUF.png 04:16
Does anyone know how to fudge stuff in roast so I could keep it outside of t/spec and still fudge? 04:32
dalek ast: 6cd55d6 | (Zoffix Znet)++ | S06-signature/introspection.t:
Callable in signature stringifies correctly

RT#128392		 04:34
synopsebot6 Link: rt.perl.org/rt3//Public/Bug/Displa...?id=128392
dalek ast: 6fd8bbb | (Zoffix Znet)++ | S03-operators/repeat.t:
xx must not sink

RT#128382		 04:51
synopsebot6 Link: rt.perl.org/rt3//Public/Bug/Displa...?id=128382
dalek kudo/nom: 340e2dc | (Zoffix Znet)++ | src/Perl6/Metamodel/PackageHOW.nqp:
Revert "Make .ver and .auth MetaModel methods available on Packages"

This reverts commit 26817b3e355c663bd12a9e0c3e1b0728befb3934.
As per discussion on github.com/rakudo/rakudo/pull/818#...-231468771 packages are stubs and must not have .ver/auth identifiers		 09:05
kudo/nom: 8382520 | niner++ | src/Perl6/Metamodel/PackageHOW.nqp:
Merge pull request #819 from zoffixznet/revert-ver

Revert "Make .ver and .auth MetaModel methods available on Packages"
kudo/nom: b11477f | peschwa++ | src/vm/jvm/CompUnit/Repository/Java.pm:
Add basic infrastructure to support JVM interop via CURs.
 09:37
kudo/nom: cbf1171 | peschwa++ | t/03-jvm/01-interop.t:
:from<java> is gone in favor of :from<Java>.

Sync the Str+Int version of Str.x with the str+int version
psch huh, 16 commits not enough for "heuristic branch merge"?
timotimo the dalek exterminates in mysterious ways 09:38
psch there also was a merge commit somewhere in there :P 09:39
anyway, yeah, we now have jvminterop that passes all currently existing tests in t/03-jvm/01-interop.t 09:41
it's only 26 tests, and they don't really test a lot, but it's something
timotimo *nice* 09:42
i was too uninspired to create more tests for that :(
psch i mean, the branch was pretty much still fine
but i didn't want to merge while we still had the whole install-core-dist.pl problem vOv
well, the current tests do cover most of what works, so idk 09:43
timotimo so only 5% of those tests are actually mine? :) 09:44
psch you wrote tests for jvminterop? 09:45
timotimo maybe? 09:46
psch git says no, blame falls squarely on me for t/03-jvm/01-interop.t
well, at least as of the recent merge
timotimo OK :) 09:47
it's probably fine
i might not have written tests for in there
kudo/nom: 8ea2ae8 | lizmat++ | src/core/Str.pm:
Merge pull request #820 from MasterDuke17/Str.pm_optimizations

Sync the Str+Int version of Str.x with the str+int version
timotimo Str.x? huh?
ah, infix:<x>, ok 09:56
[Tux] This is Rakudo version 2016.06-202-g8ea2ae8 built on MoarVM version 2016.06-9-g8fc21d5 10:25
test 16.623
test-t 9.068
csv-parser 17.143
hankache is it getting any faster [Tux]? 10:27
lizmat hankache: generally yes, with some ±.3 seconds noise 10:28
hankache great! 10:29
MasterDuke would anyone mind if i try starting a bot here? for background, see irclog.perlgeek.de/perl6/2016-07-09#i_12810559 10:40
tldr, since AlexDaniel already has a Rakudo built at every commit for bisectbot, why not make those available to run code like camelia 10:41
silence taken as consent 10:52
lizmat MasterDuke++ 10:56
timotimo yeah, that'd be cool 10:58
MasterDuke committable: 110704d2ea8b31435a68b3f01de9b2a9875dcc7b say $*PERL.compiler.version 10:59
committable MasterDuke: failure
MasterDuke awww 11:00
timotimo we're not going to have a cameliable, though? :) 11:03
MasterDuke ha. next might be benchable 11:04
timotimo well, camelia can already profile
MasterDuke i'd forgotten about that, but being able to run the same code at different commits and returning the timing might be useful 11:05
timotimo right 11:06
that'd require a machine that isn't busy otherwise
hack, for example, would be a very bad machine to use
or p6c.org in general
MasterDuke how many cores does it have?
dalek kudo/nom: 4e4039e | lizmat++ | src/core/List.pm:
Make List.AT-POS about 2.5x faster

  - don't allocate when we don't need to
  - rewrite logic in nqp, and be smarter about it
  - now also capable of handling bare Lists without $!reified
  - also fix $list.AT-POS(Int:U) handling		 11:08
timotimo 6 cores
the machine, i mean.
MasterDuke maybe partition one off for benchmarking with cgroups and cpuset? 11:12
committable: 110704d2ea8b31435a68b3f01de9b2a9875dcc7b say $*PERL.compiler.version 11:21
committable v2016.06.182.g.110704.d
MasterDuke committable: f583f22e3a60ef2d1ca035db17d7c966d53ea6db say $*PERL.compiler.version
committable v2016.06.183.gf.583.f.22
timotimo neat 11:22
MasterDuke committable: f583f22 say $*PERL.compiler.version
committable v2016.06.183.gf.583.f.22
MasterDuke i only have two versions built, but if we get AlexDaniel to run it... 11:23
lizmat MasterDuke: perhaps also mention the person who requested it ?
MasterDuke lizmat: that would be me 11:24
lizmat no, I mean in the answer of commitable
committable: f583f22 say $*PERL.compiler.version
committable v2016.06.183.gf.583.f.22
lizmat ^^^ something like: "lizmat: v2016.06.183.gf.583.f.22"
MasterDuke ahh, right, i'll see about that 11:25
lizmat MasterDuke++
psch nine: how do i allow a :path opt for use ...:from<Java>? 11:27
nine: it looks to me like i'd generally need to adjust DependencySpecification to pass that through, but that seems kinda bad..? 11:28
dalek kudo/nom: 959cd3b | lizmat++ | src/core/List.pm:
Make List.BIND-POS fail instead of throw

So we get a better runtime error. Also, make sure we only accept Int:D for the index.		 11:29
MasterDuke committable: f583f22 say $*PERL.compiler.version 11:30
committable MasterDuke: v2016.06.183.gf.583.f.22
lizmat MasterDuke++ 11:31
psch nine: i mean, i don't think having to install .jar files into a CURI is the right solution, so we have to be able to reach them some other way..
AlexDaniel committable: 280a487 say ‘hello’ 12:11
:O
committable: 280a487 say ‘hello’
committable AlexDaniel: hello
AlexDaniel committable: 2015.10 say ‘hello’ 12:14
committable AlexDaniel: hello
AlexDaniel MasterDuke: I have relaxed it a little bit
MasterDuke: I'm in a car right now, so gonna commit it a bit later :) 12:15
MasterDuke AlexDaniel: nice
psch AlexDaniel: shouldn't it also say which commit it's evaluating on?
like, people eval'ing at nearly the same time does happen
AlexDaniel psch: sure, but that's the very first version
psch right, just checking it's on the radar, no pressure :)
AlexDaniel psch: it gets back to you by using your nick, so that's not a big problem
psch oh, right, didn't see that 12:16
AlexDaniel committable: 2015.10 say ‘hello’; say ‘oops’
committable AlexDaniel: hello
..oops
AlexDaniel also this ↑ … :)
but we will fix that, of course
nine psch: write your own CompUnit::Repository implementation like CompUnit::Repository::Perl5 12:17
src/core/CompUnit/Repository/Perl5
psch nine: i already have that, it's about what statement_mod:<use> parses and what $*W does with the parse result
github.com/rakudo/rakudo/blob/nom/....nqp#L1181 to be precise 12:18
nine psch: use lib 'jar#/foo/bar/jar'; use Foo:from<Java>;
psch that seems really terrible
AlexDaniel committable: 2015.10 say ‘hello’; say ‘oops’
committable AlexDaniel: hello
..oops
nine psch: we do the same with Perl 5: use lib:from<Perl5> qw(/foo/bar); use Foo:from<Perl5>;
psch alright, i'll just go with that then 12:19
AlexDaniel committable: 2015.10 say ‘hello’; say ‘oops’
committable AlexDaniel: hello␤oops
AlexDaniel ok
committable: 2015.10 exit 5
committable AlexDaniel: AlexDaniel: ｢exit=5｣
AlexDaniel committable: 2015.10 say ‘hello’; exit 5 12:20
committable AlexDaniel: AlexDaniel: ｢exit=5｣ hello
nine psch: it's the same in pure Java, isn't it? You add the jars to the JAVAPATH or whatever it's called and then just load classes
AlexDaniel very rough but still usable ;) 12:21
psch nine: well, sorta. you can add paths to the env var or to the java invocation *or* create new classloaders with specific paths during runtime
nine: it's just that architecturally it'd be a lot more obvious to me to create a class loader per jar we want, which is easier if i know per jar where i have to look 12:22
but as an obvious solution it might also be somewhat suboptimal :)
so, if i do the <use lib 'path/to.jar'> i get a FileSystem nested inside my RepoRegistry..? 12:23
err, not Registry, Repo::Java 12:24
tbrowder [Coke]: Ref using a "writable clone of roast into t/spec inside the rakudo build dir", can I just move my existing roast repo there and use it to make changes for PRs as usual, i.e., will it continue to function as my local repo without having to have another copy? 12:44
timotimo yeah, git repos should be re-locatable without issue 12:45
psch you could also just set upstream for the repo in t/spec to your fork
potentially only for pushing, too, to make making PRs easier
timotimo or just git fetch from your local "other checkout" into t/spec and git reset --hard to your HEAD commit from that repo 12:48
the possibilities are pretty much endless
tbrowder timtimo: psch: thanks! I'll probably have more questions later... 12:51
MasterDuke committable: 2015.10 say ‘hello’ 12:57
committable MasterDuke: hello
dalek ast: 18fafbd | (Zoffix Znet)++ | S06-signature/ (2 files):
Merge two sets of tests for RT#128392 together
 14:02
synopsebot6 Link: rt.perl.org/rt3//Public/Bug/Displa...?id=128392
dalek ast: 67b9eb5 | (Zoffix Znet)++ | S06-signature/types.t:
Fix typo
Zoffix Another niggle with RT: copy-pasting copies stuff all on one line. Like trying to copy-paste this code example: rt.perl.org/Ticket/Display.html?id...et-history 14:47
MasterDuke benchable: f583f22,110704d my $a = "a" x 2**16;for ^10000 {my $b = $a.chop($_)} 15:31
dalek ast: 191c957 | (Zoffix Znet)++ | S02-magicals/ (2 files):
.WHAT on $*GROUP and $*USER after using their values first must live

RT#128099		 15:35
synopsebot6 Link: rt.perl.org/rt3//Public/Bug/Displa...?id=128099
sortiz m: ~Str 16:32
camelia rakudo-moar 959cd3: OUTPUT«WARNINGS for <tmp>:␤Useless use of "~" in expression "~Str" in sink context (line 1)␤Use of uninitialized value of type Str in string context␤Any of .^name, .perl, .gist, or .say can stringify undefined things, if needed. in block <unit> at <tmp> li…»
sortiz m: say ~Str 16:33
camelia rakudo-moar 959cd3: OUTPUT«Use of uninitialized value of type Str in string context␤Any of .^name, .perl, .gist, or .say can stringify undefined things, if needed. in block <unit> at <tmp> line 1␤␤»
psch m: Str.say
camelia rakudo-moar 959cd3: OUTPUT«(Str)␤»
sortiz m: say ~Str; say "I'm alive, that was only a warning, not an exception"; 16:34
camelia rakudo-moar 959cd3: OUTPUT«Use of uninitialized value of type Str in string context␤Any of .^name, .perl, .gist, or .say can stringify undefined things, if needed. in block <unit> at <tmp> line 1␤␤I'm alive, that was only a warning, not an exception␤»
sortiz Any way for handling those warnings? 16:35
timotimo yeah
m: quietly say ~Str
camelia rakudo-moar 959cd3: OUTPUT«␤»
timotimo other than that, you can have a CONTROL block and do nothing (but resume) when it's a CX::Warn or what it's called 16:36
m: CONTROL { .perl.say }; ~Str 16:37
camelia rakudo-moar 959cd3: OUTPUT«WARNINGS for <tmp>:␤Useless use of "~" in expression "~Str" in sink context (line 1)␤CX::Warn.new(message => "Use of uninitialized value of type Str in string context\nAny of .^name, .perl, .gist, or .say can stringify undefined things, if needed.")␤…»
sortiz Ok, thanks timotimo
timotimo yeah, CX::Warn
the other thing is a compiler warning
m: no warnings; CONTROL { .perl.say }; ~Str
camelia rakudo-moar 959cd3: OUTPUT«===SORRY!===␤Don't know how to 'no warnings' just yet␤»
timotimo hmm
psch m: CONTROL { CX::Warn { } }; ~STr
camelia rakudo-moar 959cd3: OUTPUT«===SORRY!=== Error while compiling <tmp>␤Unexpected block in infix position (missing statement control word before the expression?)␤at <tmp>:1␤------> CONTROL { CX::Warn⏏ { } }; ~STr␤ expecting any of:␤ infix␤ in…»
psch m: CONTROL { CX::Warn { } }; ~Str
camelia rakudo-moar 959cd3: OUTPUT«===SORRY!=== Error while compiling <tmp>␤Unexpected block in infix position (missing statement control word before the expression?)␤at <tmp>:1␤------> CONTROL { CX::Warn⏏ { } }; ~Str␤ expecting any of:␤ infix␤ in…»
psch m: CONTROL { when CX::Warn { } }; ~Str
camelia rakudo-moar 959cd3: OUTPUT«WARNINGS for <tmp>:␤Useless use of "~" in expression "~Str" in sink context (line 1)␤»
psch m: CONTROL { when CX::Warn { } }; say ~Str
camelia ( no output )
timotimo m: CONTROL { when CX::Warn { } }; say ~Str; say "alive" 16:38
camelia ( no output )
timotimo m: CONTROL { when CX::Warn { .resume } }; say ~Str; say "alive"
camelia rakudo-moar 959cd3: OUTPUT«␤alive␤»
timotimo ^- important difference
psch oh, right
sortiz oh!! Yep
timotimo BBIAB 16:39
MasterDuke benchable: f583f22,110704d my $a = "a" x 2**16;for ^10000 {my $b = $a.chop($_)} 16:52
benchable MasterDuke: 110704d=2.281 f583f22=2.304 16:53
Zoffix :o 17:00
MasterDuke i think it's kind of neat 17:03
Zoffix And useful 17:08
MasterDuke i hope so, that's the idea 17:12
Zoffix bisect: run($*EXECUTABLE, "--rxtrace", "-e", q/say "hello world"/, :out).out.slurp-rest.say 17:15
bisectable Zoffix: exit code is 0 on both starting points, bisecting by using the output
Zoffix: (2015-12-25) github.com/rakudo/rakudo/commit/07fecb5
AlexDaniel Zoffix: this typically means that the output is different each time 17:16
committable: HEAD run($*EXECUTABLE, "--rxtrace", "-e", q/say "hello world"/, :out).out.slurp-rest.say 17:18
committable AlexDaniel: Calling parse␤ Calling TOP␤ Calling target␤ Calling comp_unit␤hello world␤
AlexDaniel committable: 2015.12 run($*EXECUTABLE, "--rxtrace", "-e", q/say "hello world"/, :out).out.slurp-rest.say
committable AlexDaniel: Cannot invoke this object␤ at gen/moar/m-CORE.setting:2452 (/home/bisectable/git/bisectable/builds/ec386e5ff54a6e8028e74092d1a41cfccdc531d2/share/perl6/runtime/CORE.setting.moarvm:push-exactly:53)␤ from gen/moar/m-CORE.setting:3637
..(/home/bisectable/git/bisectable/builds/ec386e5ff54a6e8028e74092d1a41cfccdc531d2/share/perl6/runtime/CORE.setting.moarvm:pull-one:209)␤ from gen/moar/m-CORE.setting:2518
..(/home/bisectable/git/bisectable/builds/ec386e5ff54a6e8028e74092d1a41cfccdc531d2/share/perl6/runtime/CORE.setting.moarvm:sink-all:20)␤ from gen/moar/m-CORE.setting:12094
..(/home/bisectable/git/bisectable/builds/ec386e5ff54a6e8028e74092d1a41cfccdc531d2/share/perl6/runtime/CORE.setting.moarvm:sink:36)␤ from gen/moar/m-CORE.setting:35183
..(/home/bisectable/git/bisectable/builds/ec386e5ff54a6e8028e74092d1a41cfccdc531d2/share/perl6/runtime/CORE.setting.moarvm:resolve_repossession_conflicts:65)␤ from <unknown>:1
AlexDaniel oh geez
committable ..(/home/bisectable/git/bisectable/builds/ec386e5ff54a6e8028e74092d1a41cfccdc531d2/share/perl6/runtime/CORE.setting.moarvm:<dependencies+deserialize>:53)␤ from gen/moar/m-ModuleLoader.nqp:223
AlexDaniel oh no
committable ..(/home/bisectable/git/bisectable/builds/ec386e5ff54a6e8028e74092d1a41cfccdc531d2/share/nqp/lib/Perl6/ModuleLoader.moarvm::44)␤ from gen/moar/m-ModuleLoader.nqp:213
psch heh
committable ..(/home/bisectable/git/bisectable/builds/ec386e5ff54a6e8028e74092d1a41cfccdc531d2/share/nqp/lib/Perl6/ModuleLoader.moarvm:load_setting:33)␤ from src/Perl6/World.nqp:512 17:19
..(/home/bisectable/git/bisectable/builds/ec386e5ff54a6e8028e74092d1a41cfccdc531d2/share/nqp/lib/Perl6/World.moarvm:load_setting:55)␤ from src/Perl6/World.nqp:284
..(/home/bisectable/git/bisectable/builds/ec386e5ff54a6e8028e74092d1a41cfccdc531d2/share/nqp/lib/Perl6/World.moarvm:loading_and_symbol_setup:181)␤ from src/Perl6/Grammar.nqp:1138
pmichaud what's the general difference between #perl6 and #perl6-dev ?
committable ..(/home/bisectable/git/bisectable/builds/ec386e5ff54a6e8028e74092d1a41cfccdc531d2/share/nqp/lib/Perl6/Grammar.moarvm:comp_unit:901)␤ from gen/moar/stage2/nqpmo.nqp:1309 (/home/bisectable/git/bisectable/builds/ec386e5ff54a6e8028e74092d1a41cfccdc531d2/share/nqp/lib/nqpmo.moarvm::40)␤ from
..src/Perl6/Grammar.nqp:441 (/home/bisectable/git/bisectable/builds/ec386e5ff54a6e8028e74092d1a41cfccdc531d2/share/nqp/lib/Perl6/Grammar.moarvm:TOP:524)␤ from gen/moar/stage2/nqpmo.nqp:1309
MasterDuke whoops
committable ..(/home/bisectable/git/bisectable/builds/ec386e5ff54a6e8028e74092d1a41cfccdc531d2/share/nqp/lib/nqpmo.moarvm::40)␤ from gen/moar/stage2/QRegex.nqp:2093 (/home/bisectable/git/bisectable/builds/ec386e5ff54a6e8028e74092d1a41cfccdc531d2/share/nqp/lib/QRegex.moarvm:parse:44)␤ from
..gen/moar/stage2/nqpmo.nqp:1309 (/home/bisectable/git/bisectable/builds/ec386e5ff54a6e8028e74092d1a41cfccdc531d2/share/nqp/lib/nqpmo.moarvm::40)␤ from gen/moar/stage2/NQPHLL.nqp:1593
AlexDaniel told ya that you have to be careful with this now for now…
committable ..(/home/bisectable/git/bisectable/builds/ec386e5ff54a6e8028e74092d1a41cfccdc531d2/share/nqp/lib/NQPHLL.moarvm:parse:80)␤ from gen/moar/stage2/NQPHLL.nqp:1549 (/home/bisectable/git/bisectable/builds/ec386e5ff54a6e8028e74092d1a41cfccdc531d2/share/nqp/lib/NQPHLL.moarvm:compile:200)␤ from
..gen/moar/stage2/NQPHLL.nqp:1286 (/home/bisectable/git/bisectable/builds/ec386e5ff54a6e8028e74092d1a41cfccdc531d2/share/nqp/lib/NQPHLL.moarvm:eval:62)␤ from gen/moar/stage2/NQPHLL.nqp:1393
..(/home/bisectable/git/bisectable/builds/ec386e5ff54a6e8028e74092d1a41cfccdc531d2/share/nqp/lib/NQPHLL.moarvm::35)␤ from gen/moar/stage2/NQPHLL.nqp:1390 (/home/bisectable/git/bisectable/builds/ec386e5ff54a6e8028e74092d1a41cfccdc531d2/share/nqp/lib/NQPHLL.moarvm:command_eval:155)␤ from
..src/Perl6/Compiler.nqp:69 (/home/bisectable/git/bisectable/builds/ec386e5ff54a6e8028e74092d1a41cfccdc531d2/share/nqp/lib/Perl6/Compiler.moarvm:command_eval:93)␤ from gen/moar/stage2/NQPHLL.nqp:1374
AlexDaniel die
psch pmichaud: what dalek reports is one notable difference 17:20
MasterDuke on a related note...
AlexDaniel: github.com/perl6/bisectbot/pull/3
Zoffix pmichaud, most notable is the rakudo/nqp/roast repo commits are announced here and not in #perl6
psch pmichaud: #perl6 gets roast and ecosystem, this here gets nqp/rakudo
pmichaud what about language issues?
Zoffix #perl6 doesn't get roast
psch pmichaud: #perl6 seems more design and #perl6-dev more impl, from what i see 17:21
Zoffix: right, i recall roast commits from today now... not thinking quite clearly :)
pmichaud psch: okay, that answers my question. Thanks
MasterDuke #perl6 gets doc
Zoffix :)
AlexDaniel please don't mind committable, he is one day old :)
or she
psch MasterDuke: right, that was what i was thinking of!
pmichaud perhaps should be committed. ALthough in this case perhaps should be uncommitted.
AlexDaniel MasterDuke: why not run it on every single commit since 2015.10? 17:25
and draw a graph or something
so let's say every run takes 0.3, we run it 5 times on every commit 17:26
there are about 2000 commits from 2015.10 till now
committable: HEAD say 2000*5*0.3/60
committable AlexDaniel: 50
pmichaud afk again... bbl
AlexDaniel 50 minutes, meh! A bit too long
MasterDuke AlexDaniel: yeah, and if you try something that takes a couple seconds... 17:27
timotimo it'd be more sensible to run "daily" commits
AlexDaniel or just every fifth commit
timotimo 10 minutes is still alot 17:28
AlexDaniel but then you are not going to see when the slowdown or speedup is
MasterDuke i guess it depends if the main purpose is to show a general trend, or specifically compare a few commits
AlexDaniel timotimo: 10 minutes is not so much for a pretty graph :)
timotimo hm m 17:29
AlexDaniel but yeah, let's have two modes
every 20th commit for fast mode and every commit for slow mode
timotimo you'd get precise results by using cachegrind ... but it'll also make your stuff take 10x as long :) 17:31
AlexDaniel given that we're doing it for a lot of commits, I don't think that a little bit of noise will cause any trouble 17:32
or maybe it will, dunno, we will see
MasterDuke: ok, I'll work on it once I get back home 17:36
MasterDuke: thank you, that's very cool
MasterDuke sure, it's fun
dalek ast: 4652164 | (Zoffix Znet)++ | S19-command-line-options/06-dash-rxtrace.t:
--rxtrace does not crash

RT#128050
synopsebot6 Link: rt.perl.org/rt3//Public/Bug/Displa...?id=128050
sortiz I'm about to submit a PR, but when spectesting got a "Bad plan" in "t/spec/S06-signature/types.t", and indeed seems like a bad plan! So no related with my PR, can someone check that?, I can only count ten tests. 18:01
timotimo throws-like is 3 tests in one function 18:02
psch the number of subtests for throws-like is variable, isn't it?
timotimo right 18:03
sortiz Yep, but the plan count is at the hi level, no?
psch as in, every Attribute of the thrown Exception that is checked in the throws-like invocation is its own subtest
timotimo you can put many tests after the code
aye
hm
a subtest doesn't count into the plan with multiple tests, right?
psch m: use Test; plan 1; throws-like "a" + 5, Exception
camelia rakudo-moar 959cd3: OUTPUT«1..1␤Cannot convert string to number: base-10 number must begin with valid digits or '.' in '⏏a' (indicated by ⏏)␤ in block <unit> at <tmp> line 1␤␤Actually thrown at:␤ in block <unit> at <tmp> line 1␤␤# Looks like you planned 1 te…»
psch m: use Test; plan 1; throws-like '"a" + 5', Exception
camelia rakudo-moar 959cd3: OUTPUT«1..1␤ 1..2␤ ok 1 - '"a" + 5' died␤ ok 2 - right exception type (Exception)␤ok 1 - did we throws-like Exception?␤»
psch that's two tests right there, from the looks of it 18:04
we don't have a TAP parser in camelia though
dalek ast: 57d37a4 | (Zoffix Znet)++ | S06-signature/types.t:
Fix bad plan
psch but maybe that's a problem in our harness..?
timotimo no, it's one test for the "outer" plan
psch i don't really know TAP
timotimo the first 1..1 is the amount of tests all in all
psch ah, alright
timotimo the 1..2 is the subtests only
psch yeah, as i said, i don't really know TAP :S 18:05
Zoffix sortiz, try now. I messed up the plan in github.com/perl6/roast/commit/18fafbddef
timotimo :)
sortiz Ah! Zoffix...
timotimo i only know little bits of tap
.o( tap dancing )
Zoffix subtest counts as 1 test and you can use `plan` inside a subtest to specify the number of tests the subtest should run
and yeah, throws-like uses subtest to run two tests, but it counts as 1 test 18:07
timotimo either way, that coutns as a good catch by sortiz 18:08
Zoffix changes the workflow to avoid such fuckups :) 18:10
psch i did have a git precommit hook somewhere that runs the to-be-commited test file with make... 18:11
Zoffix I do that by hand, but I guess I just missed that file (just woke up at that time). I'm gonna put a spectest run in a loop 18:13
dalek ast: ec748a9 | (Zoffix Znet)++ | S32-str/substr.t:
Failure in .substr does not get incorrectly handled

RT#128038		 19:58
synopsebot6 Link: rt.perl.org/rt3//Public/Bug/Displa...?id=128038
tbrowder ref rakudo PR 817: please see PR 821 which fixes PR 817: github.com/rakudo/rakudo/pull/821 20:18
the spectest log is at gist gist.github.com/tbrowder/7ec37e374...718fa85297 20:19
Zoffix Those failures are OK. They will build fine with bleed rakudo/roast 20:20
tbrowder Zoffix: lizmat: note that PR 821 needs PR817 because I used PR 817 as its base. 20:28
i.e., it's the same branch 20:29
Zoffix doesn't have a commit bit ¯_(ツ)_/¯
tbrowder I'm gonna have a beer and think about the next step in this series. 20:30
Well, maybe I'll wait 1.49 hrs. 20:31
Zoffix .ask sortiz can this ticket be closed? rt.perl.org/Ticket/Display.html?id=128000 20:44
yoleaux2 Zoffix: I'll pass your message to sortiz.
Zoffix m: for (^100).race(batch=>1) { {1 + $^a}.assuming(10)() } 20:50
camelia rakudo-moar 959cd3: OUTPUT«(signal ABRT)Cannot resolve caller postfix:<++>(Int); none of these signatures match:␤ (Mu:D $a is rw)␤ (Mu:U $a is rw)␤ (Int:D $a is rw)␤ (int $a is rw)␤ (Bool:U $a is rw)␤ (Bool:D $a is rw)␤ (Num:D $a is rw)␤ (Num:…»
Zoffix m: for (^100).race(batch=>1) { {1 + $^a}.assuming(10)() }
camelia rakudo-moar 959cd3: OUTPUT«Parameter '$a' requires a type object of type Int, but an object instance was passed. Did you forget a 'multi'?␤␤Parameter '$a' requires a type object of type Int, but an object instance was passed. Did you forget a 'multi'?␤␤Parameter '$a' requi…»
Zoffix
.oO( «Parameter '$a' requires a type object of type Int ... ?)
lizmat tbrowder: so I need to revert the revert, and then apply 821 ? 20:57
tbrowder lizmat: I think so, at least that's the way it looks to me. 20:58
psch m: {1 + $^a}.assuming(10)
camelia ( no output )
psch oh, .race/batch problem 20:59
that's almost EXPR level scary :P
actually, probably more scary
dalek kudo/nom: 103e301 | lizmat++ | src/Perl6/Pod.nqp:
Revert "Revert "preparing for upcoming changes:""

This reverts commit 110704d2ea8b31435a68b3f01de9b2a9875dcc7b.
We've got a plan!		 21:00
kudo/nom: c531b8b | lizmat++ | src/core/List.pm:
We don't need no getattr in List itself
kudo/nom: c374989 | (Tom Browder)++ | src/Perl6/Grammar.nqp:
rename sub 'formatted_text' to 'normalize_text'
kudo/nom: 9bb6366 | lizmat++ | src/Perl6/Grammar.nqp:
Merge pull request #821 from tbrowder/pod-handling-cleanup

rename sub 'formatted_text' to 'normalize_text'
Zoffix Revertception 21:01
lizmat tbrowder++ # it all spectests clean for me, except the unrelated issue with t/spec/S19-command-line/repl.t 21:08
tbrowder lizmat: thanks! 21:09
Zoffix lizmat, would you gist the failures?
I'm thinking it's OSX-specific stuff :/ 21:10
lizmat gist.github.com/lizmat/8cb8c29be77...1ec6a6abac # Zoffix 21:11
Zoffix Thanks
dalek ast: 5e25756 | (Zoffix Znet)++ | S19-command-line/repl.t:
Fudge RT#70297 test

Tested REPL output is different on OSX		 21:20
synopsebot6 Link: rt.perl.org/rt3//Public/Bug/Displa...l?id=70297
dalek kudo/nom: a885088 | lizmat++ | src/core/List.pm:
Simplify List.reification-target

And another ensure-allocated bites the dust		 21:21
lizmat Zoffix: couldn't we relax the test a bit ? like just testing for the words? 21:23
Zoffix: also, why skip 2 ?
only test 5 failed ?
Zoffix lizmat, there are 2 tests in that subtest and the fudger puts the skip() inside the subset instead of before the subset 21:24
timotimo oh, interesting
lizmat huh?
that shouldn't happen ?
Zoffix ¯\_(ツ)_/¯ 21:25
As for relaxing... ¯\_(ツ)_/¯... the ticket is about testing the output after exiting. It appears on Linux it prints "> " but on OSX nothing. I just reopened the ticket. To me, it didn't look fixed from the start, but [Coke] suggested I ask masak and masak said the current behaviour is fine. 21:27
IMO, the current behaviour is missing a newline at the end: gist.github.com/zoffixznet/a1ef3fe...cf7a9d0707 21:28
A wild guess is adding `subtest` to this list will fudge the subtest instead of stuff inside of it: github.com/perl6/roast/blob/master/fudge#L126 21:36
dalek ast: 820e4f7 | (Zoffix Znet)++ | fudge:
Add Test::Util's run_is_repl to list of fudgeable stuff
 21:38
kudo/nom: 405519a | lizmat++ | src/core/List.pm:
Kill the last use of List.ensure-allocated

And the method itself with it. This means that now all List methods know how to handle the case when no $!reified attribute has been initialized, without allocating one just to be on the safe side
  (which is what ensure-allocated did).		 21:40
Zoffix Ah, I see why [possibly] subtest is not fudgeable material. If you do 2 skip fudge before the subtest, the fudger fudges the subtest... and then 1 test inside of it. Making it fudgeable would involve figuring out how to ignore stuff inside of the subtest. 21:48
So it's iffy either way. Either it's not fudgeable or you can't use numbers or you have to write logic to parse out stuff inside the subtest
timotimo oof 21:51
travis-ci Rakudo build errored. Elizabeth Mattijsen 'We don't need no getattr in List itself'
travis-ci.org/rakudo/rakudo/builds/143614775 github.com/rakudo/rakudo/compare/9...31b8bdb360
lizmat good night, #perl6-dev 21:52
Zoffix night
timotimo gnite liz 21:53
Zoffix All this talk about fudge makes me want to have some cake. 22:12
Coincidentally, I actually do have a piece of chocolate fudge cake.
I blame Perl 6 for making me fat.
timotimo today i had a little cupcake like thing that was filled with liquid chocolate on the inside 22:14
it was baked before i ate it
it was pretty fantastic
ilmari chocolate fondant? 22:19
timotimo hm? 22:20
ilmari duckduckgo.com/?q=chocolate+fondan...;ia=images
timotimo ah 22:21
some of those pictures look similar
ilmari it's yummeh 22:22
timotimo yeah
didn't have one of those before
ilmari hasn't had one in a while. not really a summer dessert
timotimo that's true 22:26
sortiz .tell Zoffix About RT#128000 I'd keep it open 'til solved in core. If only as a way to get some feedback 22:55
synopsebot6 Link: rt.perl.org/rt3//Public/Bug/Displa...?id=128000
yoleaux2 sortiz: I'll pass your message to Zoffix.
20:44Z <Zoffix> sortiz: can this ticket be closed? rt.perl.org/Ticket/Display.html?id=128000
Zoffix ack
yoleaux2 22:55Z <sortiz> Zoffix: About RT#128000 I'd keep it open 'til solved in core. If only as a way to get some feedback
synopsebot6 Link: rt.perl.org/rt3//Public/Bug/Displa...?id=128000
travis-ci Rakudo build errored. Elizabeth Mattijsen 'Simplify List.reification-target 23:03
travis-ci.org/rakudo/rakudo/builds/143617056 github.com/rakudo/rakudo/compare/9...8508822a31
gfldex m: my $a = Int; say so $a // $a === $.WHAT; 23:38
camelia rakudo-moar 405519: OUTPUT«===SORRY!=== Error while compiling <tmp>␤Variable $.WHAT used where no 'self' is available␤at <tmp>:1␤------> my $a = Int; say so $a // $a === $.WHAT⏏;␤ expecting any of:␤ term␤»
gfldex m: my $a = Int; say so $a // $a === $a.WHAT; 23:39
camelia rakudo-moar 405519: OUTPUT«True␤»
gfldex m: my class C { method WHAT(){ return Int } }; say C.new.WHAT 23:40
camelia rakudo-moar 405519: OUTPUT«(C)␤»
gfldex m: my class C { method definedWHAT(){ return Int } }; say C.new.WHAT 23:41
camelia rakudo-moar 405519: OUTPUT«(C)␤»
gfldex m: my class C { method defined(){ return Int } }; say C.neu.defined;
camelia rakudo-moar 405519: OUTPUT«Method 'neu' not found for invocant of class 'C'␤ in block <unit> at <tmp> line 1␤␤»
gfldex m: my class C { method defined(){ return Int } }; say C.new.defined;
camelia rakudo-moar 405519: OUTPUT«(Int)␤»
gfldex m: my class C { method defined(){ return False } }; say C.new.defined;
camelia rakudo-moar 405519: OUTPUT«False␤»
gfldex m: my class C { method defined(){ return False } }; say C.new // 'undefined';
camelia rakudo-moar 405519: OUTPUT«undefined␤»
gfldex m: my $a = Int; say so $a // $a === $a.WHAT; 23:42
camelia rakudo-moar 405519: OUTPUT«True␤»
gfldex m: my $a = Int; say so 42 // $a === $a.WHAT;
camelia rakudo-moar 405519: OUTPUT«True␤»

Powered by Cro and the Raku® Programming Language.

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

Thank you!