AlexDaniel buggable: tags 00:35
buggable AlexDaniel, Total: 1649; 6.D: 2; 9999: 10; @LARRY: 25; ANNOYING: 8; BOOTSTRAP: 4; BUG: 599; BUILD: 12; CONC: 47; DOCS: 1; EXOTICTEST: 1; GLR: 3; IO: 22; JVM: 53; LHF: 5; LTA: 176; MATH: 4; META: 2; MOAR: 1; NATIVECALL: 22; NYI: 56; OO: 13; OPTIMIZER: 8; OSX: 2; PARSER: 5; PERF: 28; POD: 19; PRECOMP: 15; REGEX: 49; REGRESSION: 37; REPL: 6; RFC:
AlexDaniel, 62; RT: 2; SECURITY: 1; SEGV: 25; SINK: 1; SITE: 1; SPESH: 1; STAR: 7; TESTCOMMITTED: 10; TESTNEEDED: 54; TODO: 13; UNI: 26; UNTAGGED: 418; WEIRD: 2; WINDOWS: 3; See fail.rakudo.party/ for details
weabot so many jvm-related 00:36
samcv heh there's a "WEIRD" tag 00:49
AlexDaniel and some have already expressed their disapproval of [ANNOYING] :) 00:57
but REGRESSION tag is doing well I see… 00:58
releasable6: status
releasable6 AlexDaniel, Next release in 16 days and ≈18 hours. 1 blocker. Changelog for this release was not started yet
AlexDaniel, Details: gist.github.com/d606caf23ce10ec93e...25e48a2a77
yoleaux Zoffix: check if we can change X::Numeric::CannotConvert's $.target.^name to just $.target, which the .^name bit done by the user, if needed 01:12
travis-ci Rakudo build passed. Zoffix Znet 'Merge pull request #1180 from Tyil/resolve-not-found-warning 03:07
travis-ci.org/rakudo/rakudo/builds/283447604 github.com/rakudo/rakudo/compare/c...8c00a1f36e
[Tux] This is Rakudo version 2017.09-210-ga28c00a1f built on MoarVM version 2017.09.1-553-ga4fef0bd 06:25
csv-ip5xs 1.215 - 1.235
test 10.552 - 10.706
test-t 3.264 - 3.283
csv-parser 0.857 - 0.903
and CSV::Parser is currently broken.
that is up to TonyO
lizmat Files=1226, Tests=75556, 347 wallclock secs (14.82 usr 5.31 sys + 2436.34 cusr 225.72 csys = 2682.19 CPU) 07:53
jnthn morning o/ 09:06
dogbert17 good morning jnthn 09:14
jnthn m: say 114948 / 139220 09:46
camelia 0.825657
jnthn m: say 76896 / 85104 09:49
camelia 0.903553
Geth rakudo/nom: b33c2d6f85 | (Elizabeth Mattijsen)++ | src/Perl6/Actions.nqp
Remove superfluous line

The scope of 'self' is already set 3 lines earlier		 10:55
lizmat FWIW, I don't see the utility in the "will build" trait on Attributes 12:57
m: sub trait_mod:<will>(Attribute $, Block $) { say "hi" }; class A { has $.a } # doesn't say "hi" 12:59
camelia ( no output )
timotimo doesn't it have to have a name, too? 13:00
lizmat m: sub trait_mod:<will>(Attribute $, Block :$build!) { say "hi" }; class A { has $.a } # doesn't say it either
camelia ( no output )
timotimo i don't think i've heard of "will build" until today :)
lizmat well, me neither actually 13:01
but apparently when you stick your nose in Attribute initialization, you run into that
timotimo aye 13:02
lizmat afk& # emergency social meeting in Brussels 13:43
japhb hopes the 'emergency' part of that is tongue in cheek and not actual 14:11
moritz it typically is :-) 14:17
Zoffix Moving EVAL/EVALFILE later in the setting... Shoved them into src/core/ForeignCode.pm that looks appropriate enough if you squint 15:39
(moving 'cause I need it to know about `Blob` and adding stubs into src/core/control.pm is unweildy because all the role mixins and the stubs for the roles need to be added as well) 15:40
japhb is reminded of writing the standard words for a Forth implementation -- definition order being absolutely critical for that 15:42
Geth rakudo/nom: c73221225f | (Zoffix Znet)++ | 2 files
Move &EVAL/&EVALFILE further down the setting

I need it to know about `Blob` and adding stubs into src/core/control.pm is unweildy because all the role mixins and the stubs for the roles need to be added as well		 15:53
AlexDaniel_ Zoffix: can you set up a webhook for nqp/ repo? 16:14
and rakudo repo also 16:16
Zoffix AlexDaniel_: what's the content type? JSON?
AlexDaniel_ yes
Zoffix AlexDaniel_: I don't have the keys for rakudo repo
AlexDaniel_: nqp: "Okay, that hook was successfully created. We sent a ping payload to test it out! " 16:17
AlexDaniel_ is it green?
Zoffix Yeah 16:18
AlexDaniel_ Zoffix: ah! Roast also please
squashable6: status 16:19
squashable6 AlexDaniel_, ⚠🍕 Next SQUASHathon in ≈17 hours (2017-10-07 UTC-12⌁UTC+14). See github.com/rakudo/rakudo/wiki/Mont...Squash-Day
Zoffix oh damn, AlexDaniel_ I re-did the nqp hook. I accentally had it set on www-form encoded instead of JSOPN 16:20
AlexDaniel_: roast is now done too
AlexDaniel_ to be honest, squashable6 ignores everything before the SQUASHathon starts, so there's no quick way to tell if it's set up correctly right now… :S
( github.com/perl6/whateverable/issues/243 ) 16:23
hmm I think I can do it quickly right now…
ah, also, should squashable6 send stuff to #perl6 or #perl6-dev this time? 16:24
Zoffix #perl6 16:26
Keep it consistent + more people watch #perl6 and might be encouraged to join in
AlexDaniel_ ok! 16:27
Zoffix looks we have no non-deprecated way to slurp from a pipe in binary mode :) 16:32
(I mean, because IO::Pipe.encoding is busted up; gonna fix it later today) 16:34
AlexDaniel_ Zoffix: any chance you can hit “redeliver” on the ping payloads? 16:44
Zoffix: inactive? irclog.perlgeek.de/perl6/2017-10-05#i_15261350 16:46
Zoffix AlexDaniel_: I gave you admin perms. Rediliver ytourself
AlexDaniel_ Zoffix: thanks 16:47
Zoffix ZOFFLOP: t/spec/S17-procasync/kill.t t/spec/S11-modules/export.t 16:55
timotimo is S17-supply/lines.t flappy for anybody else?
fails in about 1 out of 8 runs
but i'm gc-stress-testing, so it could be an artifact from that 16:56
yeah, reset moar to its normal state and it doesn't want to fail any more 16:57
Zoffix ZOFFLOP: t/spec/S11-modules/nested.t 17:03
grr 17:37
EVAL defaults to "perl6".... and the other option is "Perl5"
ZofBot: dat casing bruh
ZofBot Zoffix, c”… :S well if people think i should just remove them I will as long as other devs are okay with it I'm OK with it
Geth rakudo/nom: 6c928d61d9 | (Zoffix Znet)++ | src/core/ForeignCode.pm
Implement &EVAL/&EVALFILE with Bufs per S29; ilmari++

Closes RT#122256: rt.perl.org/Ticket/Display.html?id=122256
  - Make EVALFILE slurp in binary and let EVAL handle encodings
  - Implement EVAL(Blob), decoding the same way source file would
   be read by $lang (per S29)
   - For Perl 6: use value of `--encoding` command line arg or utf8
   - For Perl 5: decode in utf-c8 and let perl handle the rest.
   Appears to produce the same output as running the code
   with perl directly. ilmari++ for pointers on how perl does this		 17:43
synopsebot RT#122256 [open]: rt.perl.org/Ticket/Display.html?id=122256 [NYI] Support EVAL(Buf)
roast: 6949cb13ad | (Zoffix Znet)++ | 3 files
Spec &EVAL(Blob) and &EVALFILE's encoding use

  - EVAL/EVALFILE must eval using the same encoding as reading the
   source file in $lang would, per S29
RT#122256: rt.perl.org/Ticket/Display.html?id=122256 Rakudo impl: github.com/rakudo/rakudo/commit/6c928d61d9		 17:45
Zoffix hm, there's a fudged test for Blob.EVAL, but I did't wanna add it. Str.EVAL is kinda meh as is 17:51
Geth roast: 60f4bd7850 | (Zoffix Znet)++ | S29-context/eval.t
Unfudge now-passing test

Keeping the fudge for Blob.EVAL; don't wanna spec us having this method, but the fudge exists in 6.c-errata. Will give more comment on ticket RT#122256		 17:54
synopsebot RT#122256 [open]: rt.perl.org/Ticket/Display.html?id=122256 [NYI] Support EVAL(Buf)
Geth nqp: 7954f9b0c1 | (Samantha McVey)++ | 2 files
Fix getuniname for control's. Fixes RT122470 RT122471

Also adds tests for the lowest and highest codepoint for each of the ranges of control codepoints, as well as one above and below these to ensure the range is correct.		 18:06
synopsebot RT#122470 [open]: rt.perl.org/Ticket/Display.html?id=122470 [UNI][JVM] uniname("\0") returns NULL
RT#122471 [open]: rt.perl.org/Ticket/Display.html?id=122471 [UNI][JVM] uniname("\x[80]") returns empty string
samcv yay i just fixed JVM so .uniname works on CP's higher than 16bit in size 18:21
[Coke] samcv++ 18:22
samcv and fixing "\c[BELL]" which i didn't even realize it didn't support codepoits like that previously. i mean higher than 16bits
TimToady \o/ 18:24
samcv i think i'm going to have it get the first half, and then if it doesn't find it, get the 2nd half 18:36
fill the hash with the 2nd half of codepoint names 18:37
then it won't be any slower than it was previously 18:42
Geth nqp: deb8cb03ec | (Samantha McVey)++ | 2 files
[jvm] fix getstrfromname for codepoints not fitting in 16bits

Also fixes RT#130542 where BELL would return a control character instead of U+1F514 🔔.
Add tests to cover the RT as well as test another higher than 16bit codepoint's name. ... (5 more lines)		 18:47
synopsebot RT#130542 [open]: rt.perl.org/Ticket/Display.html?id=130542 [JVM] [UNI] \c[BELL] returns the U+0007 control code not U+1F514 BELL
samcv yay 18:49
no there are no more [JVM] [UNI] tagged tickets at least 18:50
bartolin \o/ samcv++
samcv maybe there is one that isn't tagged that, or maybe i need to create a ticket 18:51
Zoffix bumps 19:25
holy crap. t/spec/S17-channel/stress.t now takes like double the time than the entire stresstest run :| 19:42
S32-io/lock.t and S32-num/power.t don't wanna finish :\ 19:43
Gonna kill and try restarting the stresstest 19:44
Ah, no, t/fudgeandrun S32-io/lock.t hangs on its own 19:45
and so does S32-num/power.t 19:46
And S17-channel/stress.t ran for 46s before I killed it. Don't think it was that slow before the bump :/ 19:47
k
☠☠☠ Do not bump NQP/MoarVM ☠☠☠
ping jnthn++ in #moarvm 19:51
Also, made a bash alias to do the bumps... I think it works... gist.github.com/zoffixznet/c778b93...c3d65a86ac
Zoffix relocates
s/ping/pinged/;
MasterDuke shouldn't github.com/perl6/roast/blob/master...wer.t#L151 be `await start ...` instead of just `start ...`? 20:12
Geth nqp/master: 7 commits pushed by pmurias++ 20:16
Zoffix MasterDuke: no, that's an exit hatch in case the power op takes too long to run 20:17
MasterDuke hm, the tests pass and don't hang if i add await (and that's the test it's hanging on) 20:19
Zoffix In such a case the test doesn't run. 20:20
Ah ok, the start is there to expect the op to "hang" as it'll take a long time to compute. The failure mode is when the op throws 20:21
And the io lock tests hang due to shared mutex or whatever. They used to hang similarly when lizmat++ was adding handle closing on exit 20:22
(I'm guessing here; I don't know if they actually do) 20:23
And there it too has similar setup with a background promise bailing outta the test
m: start 42
camelia WARNINGS for <tmp>:
Useless use of constant integer 42 in sink context (line 1)
pmurias m: say(1.0001 ** 10000000000000000000000) 20:24
camelia Numeric overflow
in block <unit> at <tmp> line 1
Zoffix m: start { sleep 2; exit }; say "meow" 20:26
camelia meow
Zoffix m: start { sleep 2; exit }; sleep 4
camelia ( no output )
Zoffix shrugs
pmurias Zoffix: testing that a power explodes due to an implementation detail seems a bit weird 20:27
Zoffix pmurias: that's not what the test we're discussing does 20:28
pmurias: but there is another one next to it that tests overflow
power.t looks totally messed up on my phone 20:30
Zoffix notices self slowly shifting to ASCII-only camp
pmurias: overflow tests are part of 6.c langauge. It'd be weird to test the actual boundary of throwage, but they test for values that'd take a huge amount of time to compute, so effevtively they test the program doesn't just hang when it's given too big an input to handle 20:34
oh wait, they're not part of 6.' 20:35
or maybe they don't know anore 20:36
phone browsing
Zoffix &
buggable New CPAN upload: Text-More-Perl6-01.8.tar.gz by TBROWDER cpan.metacpan.org/authors/id/T/TB/...1.8.tar.gz 21:22
New CPAN upload: Text-More-Perl6-01.8.zip by TBROWDER cpan.metacpan.org/authors/id/T/TB/...6-01.8.zip
Geth nqp: 1e17529bdf | (Samantha McVey)++ | 2 files
[jvm] Add support for <noncharacter-XXX> for getuniname op

Also add a test as well.		 21:23
MasterDuke samcv: a bunch of the JVM RTs with subject `Roast rakudo skip/todo test:*` look like they might be unicode/string stuff related 21:26
samcv that seems likely 21:31
thx
MasterDuke welcome 21:32
buggable New CPAN upload: Text-More-Perl6-0.1.8.tar.gz by TBROWDER cpan.metacpan.org/authors/id/T/TB/...1.8.tar.gz 21:42
Zoffix Wonder why this appers to hang :/ my $d := nqp::getattr($_, IO::Handle, '$!decoder'); say $d.consume-exactly-bytes($d.bytes-available); 21:46
Trying to grab all the bytes in the decoder
jnthn Zoffix: No idea, I'm sure there's code doing exactly that somewhere in IO::Handle 21:48
Zoffix I saw one that does .read-slow-path or whatever, but I think I don't need to be slow. Trying to make .slurp(:bin) work. I just need to drain the decoder and then slurp the rest in binary? 21:49
oh wow. It doesn't hang, but just takes ages for some reason 21:50
or maybe it's Buf's .gist that takes ages :)
Yup. It's the .gist 21:51
MasterDuke what's ages? and how much data?
buggable New CPAN upload: Text-More-Perl6-0.1.8.zip by TBROWDER cpan.metacpan.org/authors/id/T/TB/...-0.1.8.zip 21:52
New CPAN upload: Geo-Ellipsoid-Perl6-0.2.0.tar.gz by TBROWDER cpan.metacpan.org/authors/id/T/TB/...2.0.tar.gz
New CPAN upload: Geo-Ellipsoid-Perl6-0.2.0.zip by TBROWDER cpan.metacpan.org/authors/id/T/TB/...-0.2.0.zip
dogbert2 jnthn: FWIW, I suspect there's something wrong with github.com/MoarVM/MoarVM/commit/a9...41835cdfe4 21:56
jnthn dogbert2: Any more detailed analysis welcome 22:01
Re-reading the patch, nothing immediately stands out as being off 22:02
dogbert2 jnthn: I have the same problems described by Zoffix, hanging tests etc but reverting that commit seems to fix it for me
buggable New CPAN upload: META6-To-Man-Perl6-0.2.0.zip by TBROWDER cpan.metacpan.org/authors/id/T/TB/...-0.2.0.zip
jnthn Ah, ok
Then I guess no need to bisect
It's still pretty odd 22:03
dogbert2 it's probably something ridiculous :)
dogbert2 or I'm totally wrong 22:04
Zoffix jnthn: the comment reads "if fromspace was already at maximum size" but the code reads `if (tc->nursery_tospace_size < MVM_NURSERY_SIZE) {` should it be ->nursery_fromspace_size ? 22:06
timotimo i always get confused by the code because it flips fromspace and tospace at one point 22:07
Zoffix ZofBot: what are these spaces?
ZofBot Zoffix, Therefore a private attribute declared using "has" is private to the class, not to the role
dogbert2 "In (from)space no one can hear you scream" 22:08
buggable New CPAN upload: META6-To-Man-Perl6-0.2.0.tar.gz by TBROWDER cpan.metacpan.org/authors/id/T/TB/...2.0.tar.gz 22:12
jnthn Zoffix: At that point tospace is becoming fromspace, and we just tweak tospace_size in-place if needed 22:16
Zoffix Ah, OK.
jnthn Will probably sticking some logging in again and run tests to see what it's doing though 22:17
Zoffix Damn, can't wait until all this stuff will be known to my brain :)
Maybe in a year...
buggable New CPAN upload: Net-IP-Lite-Perl6-0.1.4.tar.gz by TBROWDER cpan.metacpan.org/authors/id/T/TB/...1.4.tar.gz 22:22
New CPAN upload: Net-IP-Lite-Perl6-0.1.4.zip by TBROWDER cpan.metacpan.org/authors/id/T/TB/...-0.1.4.zip
sjn plays around with Cro again. 22:27
yoleaux 1 Oct 2017 04:24Z <nine> sjn: IMHO named-anywhere is a hack that works around rather than fixes the issue. PR 688 replaces the hack by convention and still does not fix the whole thing.
Zoffix ZOFFLOP: t/spec/S17-promise/nonblocking-await.t 22:28
sjn can't get cro-zeromq to pass it's tests. is it out of commission or is it supposed to work?
Zoffix spectest on my box: Files=1226, Tests=75559, 717 wallclock secs (12.80 usr 1.93 sys + 2223.94 cusr 189.78 csys = 2428.45 CPU)
Zoffix looks at lizmat's 340s runs jealously 22:29
timotimo i'm wondering what my next notebook should be, not entirely pleased with the options i get for lenovo t470 and t570 22:32
buggable New CPAN upload: Number-More-Perl6-0.2.0.tar.gz by TBROWDER cpan.metacpan.org/authors/id/T/TB/...2.0.tar.gz
Geth rakudo/nom: ac8e5f430b | (Zoffix Znet)++ | src/core/Buf.pm
Make Blob.gist trim its guts to 100 els

Large blobs take ages to render and we already limit to 100 els for Lists and Maps.		 22:38
Zoffix prolly Setty and Baggy need that too
Uhhhh....
so Hash.gist trims to 100 els, but Map.gist doesn't 22:39
and justs uses Mu.gist
Geth roast: 1871724d47 | (Zoffix Znet)++ | S32-container/buf.t
Spec Blob.gist's trimming to 100 els

Rakudo impl: github.com/rakudo/rakudo/commit/ac8e5f430b		 22:43
buggable New CPAN upload: Proc-More-Perl6-0.3.0.tar.gz by TBROWDER cpan.metacpan.org/authors/id/T/TB/...3.0.tar.gz 22:52
New CPAN upload: Proc-More-Perl6-0.3.0.zip by TBROWDER cpan.metacpan.org/authors/id/T/TB/...-0.3.0.zip
Zoffix samcv: you were looking for uni tickets ealier. RT#132216 looks to be in that area 22:53
synopsebot RT#132216 [new]: rt.perl.org/Ticket/Display.html?id=132216 'a' coll 'A" not Same but More
Geth roast: 3513c1f600 | (Samantha McVey)++ | S32-str/tc.t
Update JVM todo reason after RT124684 merged into RT124500
 23:20
synopsebot RT#124684 [open]: rt.perl.org/Ticket/Display.html?id=124684 [JVM] [NYI] [UNI] NFC/NFG
RT#124500 [open]: rt.perl.org/Ticket/Display.html?id=124500 [JVM] [NYI] [UNI] NFC/NFG
samcv ah yeah i was going to work on that yesterday but got busy with something 23:21
timotimo so, reading the thread about <<>> on the users list 23:30
m: my $x = "123"; my @foo = <<abc{$x}xyz>>; say @foo.perl
camelia ["abc", IntStr.new(123, "123"), "xyz"]
timotimo is it really supposed to put the $x that got interpolated into a separate element?
TimToady m: my $x = "123"; my @foo = <<"abc{$x}xyz">>; say @foo.perl 23:32
camelia ["abc123xyz"]
TimToady easy enough to get the other bahavior
but arguably not much like shell quoting
timotimo yeah
TimToady m: my $x = "123 456"; my @foo = <<abc{$x}xyz>>; say @foo.perl 23:35
camelia ["abc", IntStr.new(123, "123"), IntStr.new(456, "456"), "xyz"]
timotimo cool. i get finnish spam now. the "my dead client had lots of money" kind. google translate turned the second sentence into "I'm Andrew Allen, a lawyer and a lawyer" 23:36
i find that amusing for some reason
"lakimies ja asianajaja" 23:37

Powered by Cro and the Raku® Programming Language.

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

Thank you!