Perl 6 language and compiler development | Logs at colabti.org/irclogger/irclogger_log/perl6-dev | For toolchain/installation stuff see #perl6-toolchain | For MoarVM see #moarvm
Set by Zoffix on 27 July 2018.
releasable6 Next release in ≈1 day and ≈15 hours. 3 blockers. Please log your changes in the ChangeLog: github.com/rakudo/rakudo/wiki/ChangeLog-Draft 03:00
05:18 evalable6 joined 05:28 masak left 05:33 masak joined, masak is now known as Guest69029, epony joined 06:29 [TuxCM] left 07:09 patrickb joined 07:18 discord61 joined, discord61 left, discord61 joined 07:21 discord61 left 07:26 discord6 joined 07:27 synopsebot_ joined, tyilanmenyn joined 07:28 discord62 joined 07:29 discord62 left 07:34 Geth joined
lizmat Files=1275, Tests=108113, 209 wallclock secs (26.93 usr 7.33 sys + 2904.65 cusr 282.01 csys = 3220.92 CPU) 07:37
m: my $width = 4; printf("%0{$width}d", 42) 09:46
camelia Use of uninitialized value element of type Any in string context.
Methods .^name, .perl, .gist, or .say can be used to stringify it to something meaningful.
in block <unit> at <tmp> line 1
Your printf-style directives specify 0 arguments, but …
lizmat m: my $width = 4; printf("\%0{$width}d", 42)
camelia 0042
lizmat I don't think %0 as some direct access to $/ is documented anywhere 09:48
m: "abcd" ~~ m/(\w)+/; dd %0
camelia Hash % = {:a(Match.new(pos => 2, made => Any, from => 1, orig => "abcd", hash => Map.new(()), list => ())), :c(Match.new(from => 3, made => Any, pos => 4, orig => "abcd", list => (), hash => Map.new(())))}
lizmat bisectable6: "abcd" ~~ m/(\w)+/; dd %0
bisectable6 lizmat, Bisecting by output (old=2015.12 new=f05b1b0) because on both starting points the exit code is 0
lizmat, bisect log: gist.github.com/293bb71c5ee4404b81...7b2b7d6025
lizmat, (2017-02-22) github.com/rakudo/rakudo/commit/1c...24e4d0d5d3
lizmat thats.... unexpected 09:49
bisectable6: "abcd" ~~ m/(\w)+/; say %0 09:50
bisectable6 lizmat, Bisecting by output (old=2015.12 new=f05b1b0) because on both starting points the exit code is 0
lizmat, bisect log: gist.github.com/0d6e0b768b300993d6...9ace20cd83
lizmat, (2016-04-05) github.com/rakudo/rakudo/commit/6a...822dac090b
timotimo hm. i already knew about %0
lizmat bisectable6: old=2016.05 "abcd" ~~ m/(\w)+/; say %0 09:51
bisectable6 lizmat, On both starting points (old=2016.05 new=f05b1b0) the exit code is 0 and the output is identical as well
lizmat, Output on both points: «{a => ｢b｣, c => ｢d｣}␤»
timotimo it's equivalent to %($0)
which explains why it has the keys a and c
lizmat and what is the use ?
timotimo it uses the "every other element becomes a key or a value" thing
dunno :) 09:52
lizmat m: my $width = 4; printf("%0{$width}d", 42) # makes this really weird
camelia Use of uninitialized value element of type Any in string context.
Methods .^name, .perl, .gist, or .say can be used to stringify it to something meaningful.
in block <unit> at <tmp> line 1
Your printf-style directives specify 0 arguments, but …
timotimo yeah
[Tux] Rakudo version 2019.03.1-628-gf05b1b080 - MoarVM version 2019.05-87-ga601746d6
csv-ip5xs1.177 - 1.278
csv-ip5xs-2065.220 - 72.295
csv-parser24.038 - 25.108
csv-test-xs-200.432 - 0.459
test7.117 - 7.276
test-t1.769 - 1.818
test-t --race0.954 - 1.009
test-t-2029.182 - 30.871
test-t-20 --race9.742 - 9.782
timotimo may have to go with $($width) there
lizmat m: my $width = 4; printf("%0$($width)d", 42) # makes this really weird 09:53
camelia 0042
lizmat yeah, that's a workaround... but I'm not getting the raison d'être of %0 at all 09:54
timotimo probably just "consistency with @0 and $0
m: say &0
camelia Nil
timotimo this also exists it seems like
m: "abcd" ~~ /(\w)+/; say &0; say &1
camelia [｢a｣ ｢b｣ ｢c｣ ｢d｣]
Nil
timotimo that's certainly odd, too 09:55
lizmat yeah... I think we've found a wart
AlexDaniel kawaii: ? can I have a gist? 09:59
lizmat github.com/perl6/problem-solving/issues/44 10:03
kawaii AlexDaniel: away from my laptop currently, will ping you when you can have it 10:11
Guest77572 AlexDaniel: are you a Doctor or Professor now? 10:22
AlexDaniel Guest77572: just finished my Master's, so not that far yet :) 10:27
and not sure if I want to go further
Geth ¦ problem-solving: AlexDaniel assigned to jnthn Issue %0 and &0 should probably be syntax errors github.com/perl6/problem-solving/issues/44 10:28
11:06 AlexDaniel left 11:26 ggoebel left 11:33 gfldex joined
Geth ¦ problem-solving: AlexDaniel self-assigned Ecosystem issues and a corresponding dev github.com/perl6/problem-solving/issues/45 11:37
12:01 ggoebel joined 12:09 vrurg left
Guest77572 AlexDaniel: impressive, congratulations 12:19
timotimo now we have MasterDuke and MasterDaniel 12:37
so, what does Masterable do? :)
Guest77572 they are Masters of the universe :) 12:41
releasable next 12:43
releasable: next
releasable6 Guest77572, Next release in ≈1 day and ≈6 hours. 3 blockers. 257 out of 628 commits logged (⚠ 3 warnings)
Guest77572, Details: gist.github.com/a50d760c7a30685dc1...16c3183bac
Guest77572 only three blockers, interesting
M#1104 is obviously not a blocker 12:44
synopsebot_ M#1104 [open]: github.com/MoarVM/MoarVM/issues/1104 Spectest failures on 32 bit after commit 2fe2f588
12:45 vrurg joined 12:47 AlexDaniel joined
AlexDaniel` Guest77572: 12:48
you're kidding, right?
timotimo there's a fix in moarvm already, didn't that help? i believe we weren't able to test it because the windows build is broken for any number of reasons all the time 12:49
Guest77572 of course I'm kidding, I think it should be a blocker :) 13:09
oh, I now see that it is a blocker. I'll verify it when I get home from $work 13:10
I believe that I tested it after timotimo's fixes and the problems persisted but I might be mistaken 13:12
13:36 discord6 left, discord6 joined
patrickb How can such a comment possibly *not* slip through the cracks? github.com/MoarVM/MoarVM/commit/2f...t-33930846 13:43
timotimo oh, i saw that. i just didn't know how best to react to it; we'll definitely need a libtommath bump 13:45
kawaii AlexDaniel: for some reason gist doesn't like any of these files when I try drag them in, so I just bundled them up usercontent.irccloud-cdn.com/file/...out.tar.xz 13:47
probably something to do with their names
shrug
AlexDaniel I think there was a UI issue and it depended on where you drag it to
kawaii: I think it didn't manage to generate an svg because graphviz is not installed on your system 13:51
kawaii probably
13:56 vrurg left
AlexDaniel kawaii: github.com/rakudo/rakudo/issues/2998 14:00
lizmat: there are 4 modules with an issue coming from github.com/rakudo/rakudo/commit/2683b3d098, maybe you can take a look 14:02
lizmat grrr... :-)
which modules ?
AlexDaniel lizmat: see github.com/rakudo/rakudo/issues/2998 14:03
errors are like: Type check failed for return value; expected Seq but got List ($(7, 6, 5, 4, 3, 2, 1))
lizmat yeah, I think those tests are too strict 14:05
but lemme check 14:06
AlexDaniel lizmat: here's a separate ticket for that issue: github.com/rakudo/rakudo/issues/2999 14:07
14:10 lucasb joined
AlexDaniel kawaii: also if you can fix that annoying “perl: warning: Setting locale failed.” issue for the next run that'd be awesome 14:11
kawaii AlexDaniel: yeah sorry about that, I just didn't install my language on the VM I used to run the test 14:12
lizmat AlexDaniel: if I let List.reverse return a Seq, make install doesn't work: Type check failed for return value; expected List:D but got Seq ($((:E14CAA5C7DE5B1A1B...)
so in a sense, we're between a rock and a hard place there
AlexDaniel it's ok for this time, but later when the output is clean we'll use it to populate ecosystem-unbitrot, so ideally it shouldn't have any unrelated errors
lizmat: interesting :) It used to work at some point though? 14:13
lizmat AlexDaniel: I'll create PR's to fix the tests of those 4 modules
hmmm.... 14:14
maybe not...
lemme double check
AlexDaniel lizmat: 4 modules is a lot, actually, and it's not like the change is an obvious bugfix
lizmat you could argue that, but generally tests shouldn't need to worry about whether a List or a Seq is returned, imo 14:15
14:20 vrurg joined
lizmat AlexDaniel: running spectest with a fix now 14:23
AlexDaniel vrurg: hello! 14:24
vrurg AlexDaniel: hi!
AlexDaniel vrurg: if you can check HTTP::Server::Tiny that'd be awesome :)
vrurg: github.com/rakudo/rakudo/issues/2998
vrurg: Calling parse-http-request(Buf[uint8]) will never work with declared signature (Blob $req is copy)
vrurg AlexDaniel: will try tomorrow. Still at TPC, attending a talk right now. 14:25
AlexDaniel vrurg: sure, thanks 14:26
kawaii: I'll be away for a bit, so feel free to investigate into the remaining ones 14:32
Geth roast: a96a2d48f4 | (Elizabeth Mattijsen)++ | S03-metaops/hyper.t
Revert "Fix questionable test related to .reverse"

This reverts commit 2d1bd6ccf0321851c2f75c4bd726c70598073dd6.
vrurg AlexDaniel: NativeHelpers::Blob is seemingly related too.
kawaii AlexDaniel: you mean, annoy the people who made the breaking commits? ;) 14:33
AlexDaniel vrurg: yeah, and WebSocket fails because of HTTP::Server:Tiny :)
Geth rakudo: 8ddc2ad503 | (Elizabeth Mattijsen)++ | src/core/List.pm6
Make List.reverse return a Seq again

Apparently having it return a list caused ecosystem test breakage. Fixed R#2999.
synopsebot_ R#2999 [open]: github.com/rakudo/rakudo/issues/2999 [BLOCKER][regression] Some ecosystem fallout after “Make List.reverse a bit less Seqqy” commit
AlexDaniel kawaii: that only works for some specific issues, the rest needs to be golfed and ticketed separately 14:34
vrurg AlexDaniel: NativeHelpers could be fixed with R#2717 but I hesitate to merge it because it requires two wrong tests to be removed from roast. Would be helpful if somebody review it. 14:35
synopsebot_ R#2717 [open]: github.com/rakudo/rakudo/pull/2717 Fix for #2714
AlexDaniel vrurg: uhh, I see. And that means that either there needs to be some backcompat for previous language releases, or we consider these tests to be flat out wrong and remove them from errata branches 14:42
jnthn: you seem to have a grasp on that issue, can you take a look?
===SORRY!=== 14:43
internal problem: parser did not give circumfix an EXPR
O_o
vrurg AlexDaniel: where is this one from? 14:45
AlexDaniel vrurg: github.com/rakudo/rakudo/issues/3000 14:46
lizmat: fwiw you can also check Test::Deeply::Relaxed :) That one is likely some issue in tests of the module itself, although I'm not entirely sure 14:49
vrurg AlexDaniel: BTW, those controversial tests are fudged anyway: github.com/perl6/roast/pull/514/files 14:58
And marked as debatable by jnthn
15:10 [TuxCM] joined 15:12 vrurg left 15:16 discord6 left, discord6 joined 15:27 discord6 left 15:28 discord6 joined, discord6 left, discord6 joined 15:31 patrickb left 15:54 vrurg joined 16:03 discord6 left, discord6 joined 16:11 vrurg left 16:42 discord6 left, discord6 joined, discord6 left, discord6 joined 17:10 lucasb left 17:20 discord6 left, discord6 joined 17:22 tyilanmenyn is now known as tyil 17:44 vrurg joined
AlexDaniel vrurg: oh, they were fudged? 18:00
vrurg: so they never really passed? 18:01
vrurg Guess so.
AlexDaniel vrurg: just throw them away then…
vrurg Ok. The problem is – the fix is incomplete and doesn't cover the Blob/Buf case.
AlexDaniel gah :) 18:02
vrurg I tested the module on that branch and it failed as well.
vrurg have some more work to do.
AlexDaniel vrurg: any thoughts on how long it'll take you to finish it? Worst case
vrurg Wors: next week. Best case – tomorrow. Depends on how things will go tomorrow. 18:03
AlexDaniel alright, just keep us updated, there are other release things to do anyway… 18:04
vrurg Sure. 18:06
18:14 AlexDaniel left 18:15 AlexDaniel joined 18:33 vrurg left
dogbert17 AlexDaniel: I can, after a retest, confirm that many spectests fail on 32-bit (Linux) 18:39
m: say <-1/2>.perl # returns <-1/2> on 32-bit 18:43
camelia -0.5
AlexDaniel interesting 18:44
dogbert17: can you ticket it with all the results?
could be the same bug that just trips a bunch of tests…
dogbert17 AlexDaniel: in M#1104 then 18:46
synopsebot_ M#1104 [open]: github.com/MoarVM/MoarVM/issues/1104 [⚠ blocker ⚠] Spectest failures on 32 bit after commit 2fe2f588
18:47 HarmtH left
AlexDaniel dogbert17: 👍 18:47
18:48 vrurg joined
dogbert17 AlexDaniel: updated 18:53
18:53 HarmtH joined
AlexDaniel dogbert17: oh god 18:53
vrurg AlexDaniel: my bad. The R#2717 does fix NativeHelpers. My rakudo was installing in a wrong location. 19:13
synopsebot_ R#2717 [open]: github.com/rakudo/rakudo/pull/2717 Fix for #2714
vrurg I'll check if it does fix HTTP::Server::Tiny and then merge. 19:14
19:15 vrurg left
dogbert17 AlexDaniel: as suspected all spectest errors disappear if an older version of src/math/bigintops.c is used 19:20
AlexDaniel dogbert17: so it's caused by one of the latest changes to that file? github.com/MoarVM/MoarVM/commits/m...igintops.c 19:24
dogbert17 yes, the one mentioned in the report
19:25 vrurg joined
dogbert17 github.com/MoarVM/MoarVM/commit/2f...07ddab9e62 19:25
AlexDaniel aa, ok 19:26
Geth rakudo/master: 6 commits pushed by (Vadim Belman)++ 19:44
roast/master: 8 commits pushed by (Vadim Belman)++
vrurg AlexDaniel: done, both modules are ok now.
21:03 vrurg left
AlexDaniel vrurg++ 21:20
21:40 vrurg joined 22:12 vrurg left 22:29 vrurg joined
releasable6 Next release in ≈19 hours. 6 blockers. Please log your changes in the ChangeLog: github.com/rakudo/rakudo/wiki/ChangeLog-Draft 23:00
23:27 epony left 23:42 epony joined

Powered by Cro and the Raku® Programming Language.

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

Thank you!