Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6-dev, 2017-07-15

| Channels | #perl6-dev index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:32 evalable6 joined #perl6-dev
00:32 coverable6 joined #perl6-dev
00:32 bloatable6 joined #perl6-dev
00:32 greppable6 joined #perl6-dev
00:32 committable6 joined #perl6-dev
00:32 unicodable6 joined #perl6-dev
00:32 bisectable6 joined #perl6-dev
00:32 quotable6 joined #perl6-dev
00:32 benchable6 joined #perl6-dev
00:32 statisfiable6 joined #perl6-dev
00:48 AlexDaniel joined #perl6-dev
01:08 ugexe `perl6 -e 'use nqp; chdir "/"; say nqp::mkdir("./", 0o777);'` mkdir("/") works on osx if its not just / apparently
01:11 geekosaur probably everywhere. iirc mkdir is trying to be too smart and make intermediate dirs when needed, but that means it can succeed when everything already exists.
01:12 geekosaur (imo those should be separate operations, or if the magical one must be regarded as default then the primitive one should be exposed somewhere for those that actually need correct behavior)
01:12 ugexe well it seems to act different for / than for any other existing directory
01:14 ugexe like its giving EEXISTS for existing directories, but for / it gives EISDIR
01:15 ugexe https://github.com/MoarVM/MoarVM/blob/a2f6e5dc66cea621057ae41a6e01bb2a8d46c3fc/src/io/dirops.c#L105
01:15 ugexe errno is 21, which should be EISDIR anyway
01:18 geekosaur not that surprising and probably related to / being a special case (normally one strips trailing /s but that is a mistake here...)
01:19 * geekosaur has been through that code before and it's an utter mess
01:31 ugexe well, one way would be to use `nqp::stat("/",nqp::const::STAT_EXISTS)` before calling mkdir
01:31 ugexe i dunno if thats a good way though
01:47 TimToady joined #perl6-dev
01:47 MasterDuke joined #perl6-dev
01:49 ilbot3 joined #perl6-dev
01:49 Topic for #perl6-dev is now Perl 6 language and compiler development 2.0 | Logs at http://irclog.perlgeek.de/perl6-dev/today | For toolchain/installation stuff see #perl6-toolchain | For MoarVM see #moarvm
05:08 ugexe MasterDuke: https://github.com/rakudo/rakudo/commit/1a4d949 stopped showing exception messages in --ll-exception output. maybe add a check if -ll-exception is enabled to do the old nqp::setmessage?
05:50 greppable6 joined #perl6-dev
07:45 [Tux] This is Rakudo version 2017.06-251-g23ad2c388 built on MoarVM version 2017.06-95-gbccdb423
07:45 [Tux] csv-ip5xs        2.537
07:45 [Tux] test            12.245
07:45 [Tux] test-t           4.079 - 4.200
07:45 [Tux] csv-parser      11.886
09:01 lizmat Files=1211, Tests=65531, 215 wallclock secs (13.32 usr  4.89 sys + 1289.72 cusr 136.84 csys = 1444.77 CPU)
09:02 lizmat only 4 more tests needed to break the 16 bit barrier  :-)
09:19 lizmat m: dd Nil.IO
09:19 camelia rakudo-moar 23ad2c: OUTPUT: «IO::Path␤»
10:17 Ven joined #perl6-dev
11:38 robertle joined #perl6-dev
11:40 Ven joined #perl6-dev
11:42 dogbert17 ok, so RT #127951 seems to have been fixed some time ago, would t/spec/S03-operators/identity.t be a good place to add a test for that?
11:42 synopsebot6 Link:  https://rt.perl.org/rt3/Public/Bug/Display.html?id=127951
11:48 TimToady joined #perl6-dev
11:51 BenGoldberg joined #perl6-dev
11:53 Geth ¦ rakudo: MasterDuke17++ created pull request #1116: Fix no message printing when using ll-exception
11:53 Geth ¦ rakudo: review: https://github.com/rakudo/rakudo/pull/1116
11:56 MasterDuke could someone check that ^^^ please? the install step seemed to be acting up, but i don't know why the change would cause that, so i'm not sure if it's just a local problem
12:05 Geth ¦ roast: dogbert17++ created pull request #283: Added test for RT #127951
12:05 Geth ¦ roast: review: https://github.com/perl6/roast/pull/283
12:05 synopsebot6 Link:  https://rt.perl.org/rt3/Public/Bug/Display.html?id=127951
12:09 Geth ¦ roast: 5e12710c90 | (Jan-Olof Hendig)++ | S03-operators/identity.t
12:09 Geth ¦ roast: Added test for RT #127951
12:09 Geth ¦ roast: review: https://github.com/perl6/roast/commit/5e12710c90
12:09 Geth ¦ roast: 39802b5b2b | dogbert17++ (committed using GitHub Web editor) | S03-operators/identity.t
12:09 Geth ¦ roast: Merge pull request #283 from dogbert17/test-rt-127951
12:09 synopsebot6 Link:  https://rt.perl.org/rt3/Public/Bug/Display.html?id=127951
12:09 Geth ¦ roast:
12:09 Geth ¦ roast: Added test for RT #127951
12:09 synopsebot6 Link:  https://rt.perl.org/rt3/Public/Bug/Display.html?id=127951
12:09 Geth ¦ roast: review: https://github.com/perl6/roast/commit/39802b5b2b
12:16 dogbert17 bisect: my int $x = 4; say $x || 2
12:16 bisectable6 dogbert17, Bisecting by output (old=2015.12 new=23ad2c3) because on both starting points the exit code is 0
12:16 bisectable6 dogbert17, bisect log: https://gist.github.com/257033bdaa722deabee0ba6a1e164080
12:16 bisectable6 dogbert17, (2017-07-10) https://github.com/rakudo/rakudo/commit/83e157012da83e32b10d3739800cdc09e2287db1
12:17 dogbert17 .seen TrollerCopter
12:17 yoleaux I haven't seen TrollerCopter around.
12:17 dogbert17 .seen LollerCopter
12:17 yoleaux I saw lollercopter 25 Jun 2017 12:21Z in #perl6: <lollercopter> samcv: not IMO. No other IO routines take their paths in arbitrary chunks like that, so adding this would be inconsistent. You can always use '/' as the separator, so it'd be $*CWD.child('foo/bar') or $*CWD.child: @parts.join: '/' and will just work and will be faster than adding a slurpy multi candidate.
12:24 dogbert17 .ask Zoffix any idea where to put tests for RT #130404 which you fixed a few days ago?
12:24 synopsebot6 Link:  https://rt.perl.org/rt3/Public/Bug/Display.html?id=130404
12:24 yoleaux dogbert17: I'll pass your message to Zoffix.
13:15 lizmat joined #perl6-dev
13:49 Geth_ ¦ perl6-lwp-simple: cpin++ created pull request #15: Get method sends headers if provided
13:49 Geth_ ¦ perl6-lwp-simple: review: https://github.com/perl6/perl6-lwp-simple/pull/15
13:53 dogbert17 bisect: Buf.new((my int $i = 10) +& 0xFF)
13:53 bisectable6 dogbert17, On both starting points (old=2015.12 new=23ad2c3) the exit code is 0 and the output is identical as well
13:53 bisectable6 dogbert17, Output on both points: «»
13:53 dogbert17 bisect: say Buf.new((my int $i = 10) +& 0xFF)
13:53 bisectable6 dogbert17, On both starting points (old=2015.12 new=23ad2c3) the exit code is 0 and the output is identical as well
13:53 bisectable6 dogbert17, Output on both points: «Buf:0x<0a>»
13:56 dogbert17 bisect: my int $i = 2482111348; $i = $i div 2; say $i
13:56 bisectable6 dogbert17, Bisecting by output (old=2015.12 new=23ad2c3) because on both starting points the exit code is 0
13:56 bisectable6 dogbert17, bisect log: https://gist.github.com/c1d74cb30b5e1e4687177ac9677d1ef1
13:56 bisectable6 dogbert17, (2017-02-03) https://github.com/rakudo/rakudo/commit/c98b3a581f481756234758e474c255414a340c97
13:59 dogbert17 commit: 2016.09 say Buf.new((my int $i = 10) +& 0xFF)
13:59 committable6 dogbert17, ¦2016.09: «Buf:0x<0a>»
13:59 dogbert17 commit: 2016.06 say Buf.new((my int $i = 10) +& 0xFF)
13:59 committable6 dogbert17, ¦2016.06: «Buf:0x<0a>»
13:59 dogbert17 commit: 2016.07 say Buf.new((my int $i = 10) +& 0xFF)
13:59 committable6 dogbert17, https://gist.github.com/6813cf760c415eca5eb1e376da4020c5
14:00 dogbert17 commit: 2016.08 say Buf.new((my int $i = 10) +& 0xFF)
14:00 committable6 dogbert17, ¦2016.08: «Buf:0x<0a>»
14:00 dogbert17 bisect: HEAD 2016.07 say Buf.new((my int $i = 10) +& 0xFF)
14:00 bisectable6 dogbert17, On both starting points (old=2015.12 new=23ad2c3) the exit code is 1 and the output is identical as well
14:00 bisectable6 dogbert17, https://gist.github.com/206717b2640ec43e8d5d7b757580ea20
14:01 MasterDuke dogbert17: you need `new=xxx old=yyy`
14:02 MasterDuke bisect: new=HEAD old=2016.07 say Buf.new((my int $i = 10) +& 0xFF)
14:02 bisectable6 MasterDuke, Bisecting by exit code (old=2016.07 new=23ad2c3). Old exit code: 1
14:03 bisectable6 MasterDuke, bisect log: https://gist.github.com/3faf5573a6db6f4723c9e5892c1cb3a2
14:03 bisectable6 MasterDuke, (2016-07-17) https://github.com/rakudo/rakudo/commit/242baf256a210f1cb643cb6324916dbc76184c9f
14:04 dogbert17 MasterDuke++, was on my way to call the bot morinic :)
14:04 dogbert17 *moronic
14:06 dogbert17 so it seems as if that commit should have fixed RT #128624 and to a certain extent RT #128655
14:06 synopsebot6 Link:  https://rt.perl.org/rt3/Public/Bug/Display.html?id=128624
14:06 synopsebot6 Link:  https://rt.perl.org/rt3/Public/Bug/Display.html?id=128655
14:08 dogbert17 bisect: new=HEAD old=2016.07 my int $i = 5; Buf.new($i +& 0xFF)
14:08 bisectable6 dogbert17, Bisecting by exit code (old=2016.07 new=23ad2c3). Old exit code: 1
14:08 bisectable6 dogbert17, bisect log: https://gist.github.com/2c6c8fb35b2801a57258b5df65b29df0
14:08 bisectable6 dogbert17, (2016-07-17) https://github.com/rakudo/rakudo/commit/242baf256a210f1cb643cb6324916dbc76184c9f
14:08 * dogbert17 yeah, it's the same :)
14:12 * dogbert17 hmm, temporary fix ...
14:15 Ven joined #perl6-dev
14:35 Geth ¦ nqp: MasterDuke17++ created pull request #366: Fix no message printing when using ll-exception
14:35 Geth ¦ nqp: review: https://github.com/perl6/nqp/pull/366
14:37 Geth ¦ nqp: 859b4441dd | MasterDuke17++ | src/HLL/Compiler.nqp
14:37 Geth ¦ nqp: Fix no message printing when using ll-exception
14:37 Geth ¦ nqp:
14:37 Geth ¦ nqp: The recent change to only generate an Exception's message when it's
14:37 Geth ¦ nqp: gisted broke using `--ll-exception`.
14:37 Geth ¦ nqp: review: https://github.com/perl6/nqp/commit/859b4441dd
14:37 Geth ¦ nqp: c234cf4efb | niner++ (committed using GitHub Web editor) | src/HLL/Compiler.nqp
14:37 Geth ¦ nqp: Merge pull request #366 from MasterDuke17/fix_no_message_when_using_ll-exception
14:37 Geth ¦ nqp:
14:37 Geth ¦ nqp: Fix no message printing when using ll-exception
14:37 Geth ¦ nqp: review: https://github.com/perl6/nqp/commit/c234cf4efb
15:15 ggoebel joined #perl6-dev
15:20 Ven joined #perl6-dev
16:14 Ven joined #perl6-dev
16:16 ugexe is there a way to get at dynamic variables inside of Perl6/Optimizer.nqp ?
16:19 moritz ugexe: the optimizer runs after grammar + actions, so the dynamic variables from Actions.pm and Grammar.pm aren't available
16:20 ugexe hmmm. well here https://github.com/rakudo/rakudo/blob/nom/src/Perl6/Optimizer.nqp#L394 it can use the wrong stderr
16:21 ugexe RT #131670
16:21 synopsebot6 Link:  https://rt.perl.org/rt3/Public/Bug/Display.html?id=131670
16:22 ugexe m: my $*ERR = class :: { method say(|) {}; method print(|) {}; }; warn(1)
16:22 camelia rakudo-moar 23ad2c: ( no output )
16:22 ugexe m: my $*ERR = class :: { method say(|) {}; method print(|) {}; }; my $a; $a;
16:22 camelia rakudo-moar 23ad2c: OUTPUT: «WARNINGS for <tmp>:␤Useless use of $a in sink context (line 1)␤»
16:23 geekosaur I don't think that difference is fixable
16:23 geekosaur unless you...
16:24 geekosaur hm, doesn't work with BEGIN either. *that* could be fixable
16:25 geekosaur but neither the grammar nor the optimizer can reach forward to a *runtime* set of $*ERR
16:28 ugexe could that warning logic go somewhere else possibly?
16:32 geekosaur it's a compile time warning, do you want it to be a runtime warning instead?
16:32 moritz the nice thing about compile time warnings is that they are independent of program inputs
16:33 geekosaur to be clear: this lets the $*ERR setting be visible at compile time
16:33 geekosaur m: BEGIN my $*ERR = class :: { method say(|) {}; method print(|) {}; }; my $a; $a;
16:33 camelia rakudo-moar 23ad2c: OUTPUT: «WARNINGS for <tmp>:␤Useless use of $a in sink context (line 1)␤»
16:34 geekosaur but it still doesn't work, probably because of the dynamic variable issue moritz already mentioned (i.e. that information is not around when the optimizer is run)
16:36 geekosaur fixing that likely requires that the $*ERR in effect at compile time be associated with the block being handed to the optimizer, which will probably introduce other issues
16:52 ugexe actually it seems like the naive solution, `s/my $err := stderr()/my $err := $*ERR/` works, and catches the compile time error as  you said if you use BEGIN my $*ERR = ...
17:00 ugexe ah but then compile time errors dont work right unless you explicitly set BEGIN my $err := ...
17:00 ugexe compile time warnings rather
17:06 ugexe same if s/stderr()/getstderr()/
17:07 Ven_ joined #perl6-dev
17:19 BenGoldberg joined #perl6-dev
17:23 dogbert17 anyone knows what this comment means, 'Don't know how to mark statement for perl6!'? It shows up when running a spectest
17:27 Zoffix dogbert17: someone wrote a fudge for #?perl6 instead of #?rakudo
17:27 yoleaux 12:24Z <dogbert17> Zoffix: any idea where to put tests for RT #130404 which you fixed a few days ago?
17:27 synopsebot6 Link:  https://rt.perl.org/rt3/Public/Bug/Display.html?id=130404
17:28 Zoffix or it just assumes "#perl6" in some of the comments are fudges
17:29 Zoffix dogbert17: there were already tests in. I forgot there was a ticket too. Closed. https://github.com/perl6/roast/commit/b1f116fdd5181274cda1a520194488da82c10a7e
17:30 dogbert17 Zoffix++
17:30 Zoffix Oh
17:30 Zoffix It's a differrent ticket
17:30 dogbert17 The comment shows up just before it starts running any tests
17:31 Zoffix yeah, that's when the fudger runs
17:31 Zoffix bisect: my int $x = 4; say $x || 2
17:31 bisectable6 Zoffix, Bisecting by output (old=2015.12 new=23ad2c3) because on both starting points the exit code is 0
17:31 bisectable6 Zoffix, bisect log: https://gist.github.com/3c694174f2b255d500377ffacc134df0
17:31 bisectable6 Zoffix, (2017-07-10) https://github.com/rakudo/rakudo/commit/83e157012da83e32b10d3739800cdc09e2287db1
17:31 dogbert17 so it is probably from something which was added very recently then
17:31 Zoffix Well, it's not the exact test for that ticket, but it's for this problem, so it counts. I guess.
17:32 dogbert17 so, can we close it or will it end with AlexDaniel reopening it :)
17:34 Zoffix can close
17:34 Zoffix already is rather
17:35 Geth_ ¦ perl6-lwp-simple: c71bb0e364 | cpin++ | 3 files
17:35 Geth_ ¦ perl6-lwp-simple: Get method sends headers if provided
17:35 Geth_ ¦ perl6-lwp-simple:
17:35 Geth_ ¦ perl6-lwp-simple: Fixes #14
17:35 Geth_ ¦ perl6-lwp-simple: review: https://github.com/perl6/perl6-lwp-simple/commit/c71bb0e364
17:35 Geth_ ¦ perl6-lwp-simple: 164f470bc1 | (Zoffix Znet)++ (committed using GitHub Web editor) | 3 files
17:35 Geth_ ¦ perl6-lwp-simple: Merge pull request #15 from cpin/get-headers
17:35 Geth_ ¦ perl6-lwp-simple:
17:35 Geth_ ¦ perl6-lwp-simple: Get method sends headers if provided
17:35 Geth_ ¦ perl6-lwp-simple: review: https://github.com/perl6/perl6-lwp-simple/commit/164f470bc1
17:35 MasterDuke dogbert17: it's from S15-nfg/concat-stable.t, the shebang causes it
17:36 Geth ¦ perl6-lwp-simple: 5e06386e1c | (Zoffix Znet)++ (committed using GitHub Web editor) | META6.json
17:36 Geth ¦ perl6-lwp-simple: Bump version
17:36 Geth ¦ perl6-lwp-simple: review: https://github.com/perl6/perl6-lwp-simple/commit/5e06386e1c
17:41 dogbert17 MasterDuke++, that was quick
17:43 dogbert17 running a spectest with sundry debug flags and small nursery, so far two fails
17:44 dogbert17 the first is t/spec/S15-nfg/concat-stable.rakudo.moar, no big surprise there I guess :)
17:44 dogbert17 the second is t/spec/S29-os/system.rakudo.moar
17:57 Geth ¦ roast: cabdf07369 | (Zoffix Znet)++ | packages/Test/Util.pm
17:57 Geth ¦ roast: Remove 2 pointless regexes from is_run
17:57 Geth ¦ roast:
17:57 Geth ¦ roast: - neitzca is no longer in use
17:57 Geth ¦ roast: - $*EXECUTABLE for ./perl6 would already have ./ in it
17:57 Geth ¦ roast: review: https://github.com/perl6/roast/commit/cabdf07369
17:59 Geth ¦ roast: dogbert17++ created pull request #284: Remove shebang line from script
17:59 Geth ¦ roast: review: https://github.com/perl6/roast/pull/284
17:59 timotimo is --target=parse broken?
18:00 Zoffix I hope not
18:00 timotimo well, this here looks like an infinite recursion to me
18:01 Geth ¦ roast: 1e793c1db3 | (Jan-Olof Hendig)++ | S15-nfg/concat-stable.t
18:01 Geth ¦ roast: Remove shebang line from script
18:01 Geth ¦ roast:
18:01 Geth ¦ roast: It confused the fudger
18:01 Geth ¦ roast: review: https://github.com/perl6/roast/commit/1e793c1db3
18:01 Geth ¦ roast: bf5034bb73 | dogbert17++ (committed using GitHub Web editor) | S15-nfg/concat-stable.t
18:01 Geth ¦ roast: Merge pull request #284 from dogbert17/fix-concat-stable
18:01 Geth ¦ roast:
18:01 Geth ¦ roast: Remove shebang line from script
18:01 Geth ¦ roast: review: https://github.com/perl6/roast/commit/bf5034bb73
18:02 Zoffix timotimo: ah, that yeah. I experienced that in the past. You're trying to parse some sub def, right?
18:02 timotimo no, it's "my ($a, $b, $c)"
18:02 Zoffix or sig, yeah
18:02 timotimo i wonder if we need a $matchdumpseen
18:03 nine That does indeed hang
18:03 ugexe m: role Fooable[::T $type] { method bar { say 111 } }; class Bar does Fooable["x"] { method foo { self.Fooable::bar } }; Bar.new.foo; # is there a way to call the full name of a method from a parameterized role?
18:03 camelia rakudo-moar 23ad2c: OUTPUT: «No concretization found for Fooable␤  in method foo at <tmp> line 1␤  in block <unit> at <tmp> line 1␤␤»
18:04 timotimo worst case you can find_method
18:05 dogbert17 Zoffix: of the four files you found three are .p6 files and the fourth was the one I fixed
18:05 Zoffix Ah
18:06 Zoffix Isn't the method gets added to the class?
18:06 Zoffix m: role Fooable[::T $type] { method bar { say 111 } }; class Bar does Fooable["x"] { method foo { self.bar } }; Bar.new.foo;
18:06 camelia rakudo-moar 23ad2c: OUTPUT: «111␤»
18:06 MasterDuke timotimo: fyi, a perf record of parsing that snippet you showed gives:   48.62%  moar     libmoar.so        [.] MVM_gc_root_add_frame_registers_to_worklist   17.16%  moar     libmoar.so        [.] MVM_gc_root_add_frame_roots_to_worklist
18:06 Zoffix m: role Fooable[::T $type] { method bar { say 111 } }; class Bar does Fooable["x"] { method foo { self.Bar::bar } }; Bar.new.foo;
18:06 camelia rakudo-moar 23ad2c: OUTPUT: «111␤»
18:08 Zoffix "<ugexe> seems like Buf.Stringy should be calling something like { self."{::T}"::Stringy }"
18:08 * Zoffix doesn't follow that one
18:08 timotimo MasterDuke: what snippet was that?
18:09 ugexe m: say "a".encode.WHAT; say "a".encode.Stringy;
18:09 camelia rakudo-moar 23ad2c: OUTPUT: «(utf8)␤a␤»
18:09 ugexe if utf8 knows how to .Stringy, should Buf[utf8] also?
18:10 dogbert17 Zoffix: what about RT #130454, can that be closed/rejected?
18:10 synopsebot6 Link:  https://rt.perl.org/rt3/Public/Bug/Display.html?id=130454
18:10 Zoffix ugexe: do we *have* Buf[utf8]?
18:10 Zoffix m: Buf[utf8].new
18:10 camelia rakudo-moar 23ad2c: OUTPUT: «concatenate requires a concrete string, but got null␤  in any protect at gen/moar/stage2/NQPCORE.setting line 1033␤  in any protect at gen/moar/stage2/NQPCORE.setting line 1033␤  in block <unit> at <tmp> line 1␤␤»
18:11 ugexe m: say Buf.new("a".encode).WHAT
18:11 camelia rakudo-moar 23ad2c: OUTPUT: «(Buf)␤»
18:11 MasterDuke timotimo: target=parse of "my ($a, $b, $c)"
18:11 ugexe m: say Buf.new("a".encode).perl
18:11 camelia rakudo-moar 23ad2c: OUTPUT: «Buf.new(97)␤»
18:11 Zoffix `m: say Buf.new("x".encode).Stringy` <-- there you go through .new(Buf) candidate that just takes all the ints
18:12 Zoffix dogbert17: closed. Thanks.
18:12 Zoffix m: dd utf8.new
18:13 camelia rakudo-moar 23ad2c: OUTPUT: «utf8.new()␤»
18:13 Zoffix s: utf8, 'new', \()
18:13 SourceBaby Zoffix, Sauce is at https://github.com/rakudo/rakudo/blob/23ad2c3/src/core/Buf.pm#L36
18:13 Zoffix `my class utf8 does Blob[uint8] is repr('VMArray') {`
18:15 * timotimo has a dumpseen for matches
18:17 * Zoffix has the golfiest yet open.t bug: https://gist.github.com/zoffixznet/b27263cb79ee4c83917db166d6f7e4b1
18:17 Zoffix run with echo "x" | ./perl6 open.t
18:19 Zoffix oh, finally got it out of is_run/shell: https://gist.github.com/zoffixznet/b2e0be37457273562edc25d9121bf487
18:19 Zoffix Run with echo "x" | ./perl6 --ll-exception open.t
18:20 Zoffix strangely, --ll-exception is what causes it to explode :o
18:20 * Zoffix verbumps and builds the latest and greatest
18:22 Zoffix Shouldn't this check with nqp::ifnull so that exceptions with empty messages get differentiated from those without any? https://github.com/perl6/nqp/commit/859b4441dd3f6fd8816759b3b1f12b2d45a0bfd2
18:23 Zoffix isnull I mean
18:23 Geth ¦ nqp/dump_nqpmatch_seen_hash: 95c62e0caa | (Timo Paulssen)++ | src/QRegex/Cursor.nqp
18:23 Geth ¦ nqp/dump_nqpmatch_seen_hash: prevent infinite recursion when dumping match objects
18:23 Geth ¦ nqp/dump_nqpmatch_seen_hash: review: https://github.com/perl6/nqp/commit/95c62e0caa
18:24 timotimo https://gist.github.com/timo/4c1c109080dad637e6a3bf5a03d804d3  -^
18:24 ugexe Zoffix: looks like it. here is the old code https://github.com/rakudo/rakudo/commit/1a4d94930c581e915a2351049e8ee12246c4c26c
18:25 Zoffix k, gonna change
18:25 timotimo it's problematic that it won't put the objectid next to the thing's first occurence
18:25 timotimo though with the chunks array the way it is, it would be possible to implement that with a splice ...
18:30 timotimo but now it's grocery sohpping time
18:31 ugexe m: sub foo {...}; role R[$s] { method s{$s} }; constant C = R[:()]; dd C.s;
18:31 camelia rakudo-moar 23ad2c: OUTPUT: «===SORRY!===␤QAST::Block with cuid 7 has not appeared␤»
18:32 timotimo also, 0xfoo the numbers rather than decimal
18:34 Zoffix ZOFFLOP: t/spec/S10-packages/precompilation.t
18:37 Zoffix looks like open.t stopped flooping after the --ll-exception fix :|
18:37 Zoffix Files=1245, Tests=161083, 129 wallclock secs (22.17 usr  3.15 sys + 2658.47 cusr 158.34 csys = 2842.13 CPU)
18:37 ugexe your fix, or the prior one?
18:39 Zoffix I'd guess prior one
18:39 Zoffix Though I don't get why. The original test doesn't use --ll-exception
18:40 ugexe that was my though... unless it has something to do with catching an exception somewhere and using its message to make some decision
18:40 Zoffix And considering merely adding comments to code that used to produce the crash, I'd guess there might be something else that crashes that just got shoved under the rug now ¯\_(ツ)_/¯
18:40 Zoffix ZOFVM: Files=1245, Tests=161083, 130 wallclock secs (21.54 usr  3.16 sys + 2653.97 cusr 161.34 csys = 2840.01 CPU)
18:41 Zoffix k, a few runs and no failures. Time for release.
18:41 Geth ¦ nqp: 8cd835ffad | (Zoffix Znet)++ | src/HLL/Compiler.nqp
18:41 Geth ¦ nqp: Exclude empty messages from check
18:41 Geth ¦ nqp: review: https://github.com/perl6/nqp/commit/8cd835ffad
18:41 Geth ¦ rakudo/nom: 83502cca8c | (Zoffix Znet)++ | tools/build/NQP_REVISION
18:41 Geth ¦ rakudo/nom: Bump NQP
18:41 Geth ¦ rakudo/nom:
18:41 Geth ¦ rakudo/nom: Changes:
18:41 Geth ¦ rakudo/nom: https://github.com/perl6/nqp/compare/2017.06-59-g61307bb...2017.06-62-g8cd835f
18:41 Geth ¦ rakudo/nom:
18:41 Geth ¦ rakudo/nom: 8cd835f Exclude empty messages from check
18:41 Geth ¦ rakudo/nom: c234cf4 Merge pull request #366 from MasterDuke17/fix_no_message_when_using_ll-exception
18:41 Geth ¦ rakudo/nom: 859b444 Fix no message printing when using ll-exception
18:41 Geth ¦ rakudo/nom: review: https://github.com/rakudo/rakudo/commit/83502cca8c
18:42 Geth ¦ rakudo/nom: version bump brought these changes: https://github.com/perl6/nqp/compare/2017.06-59-g61307bb...2017.06-62-g8cd835f
18:43 Zoffix ahhh crap
18:44 Zoffix Yeah, it doesn't flop anymore because the open.t file is striped off everything. I never reset it after messing with it during debug
18:44 ugexe what is the expected output of the gist you posted?
18:45 Zoffix anything that doesn't have "Reading from filehandle" error
18:47 ugexe and can I run that gist as its own script? echo "x" | perl6 p6.p6? you mentioned `echo "x" | ./perl6 --ll-exception open.t` but that doesn't mention the code in the gist
18:48 Zoffix yep. open.t still crashes
18:48 Zoffix ugexe: the code from the gist goes to open.t file
19:00 Zoffix k, I'm calling it a day on the open.t test.
19:01 Zoffix It's probably something in MoarVM, as removing the comment avoids a failure. The failing test is also not part of 6.c-errata and is testing a very arcane functionality (re-opening $*OUT/$*IN)
19:07 Zoffix filed as https://rt.perl.org/Ticket/Display.html?id=131755
19:07 Zoffix .tell jnthn perhaps you'd be interested in this bug: https://rt.perl.org/Ticket/Display.html?id=131755
19:07 yoleaux Zoffix: I'll pass your message to jnthn.
19:13 Geth ¦ roast: 4748a283e1 | (Zoffix Znet)++ | S32-io/open.t
19:13 Geth ¦ roast: Fudge failing test for release
19:13 Geth ¦ roast:
19:13 Geth ¦ roast: - Test is not part of 6.c-errata
19:13 Geth ¦ roast: - Test tests edgecase functionality (re-opening $*OUT/$*IN)
19:13 Geth ¦ roast: - Issue detected by the test occurs under unpredictable circumstances
19:13 Geth ¦ roast:     (e.g. even adding a comment affects whether it appears)
19:13 Geth ¦ roast: - Ticket for issue: https://rt.perl.org/Ticket/Display.html?id=131755
19:13 Geth ¦ roast:
19:13 Geth ¦ roast: fudger doesn't like the test, so I just commented it out
19:13 Geth ¦ roast: review: https://github.com/perl6/roast/commit/4748a283e1
19:14 geekosaur the thing I would wonder about is at what level that error happens. involvement of --ll-exception or change in source code makes me think it's the compiler that is unexpectedly emitting that
19:14 Zoffix Original issue doesn't have --ll-exception
19:14 Zoffix I'd guess it's reading memory it ain't supposed to be reading or something like that
19:15 Ven joined #perl6-dev
19:19 nine Zoffix: then maybe it'll show up on valgrind or asan?
19:19 timotimo dang i can't do the magic chunk thing because it turns out the chunks get flattened into a string somewhat often
19:29 Zoffix Maybe. I don't know about those tools.
19:29 Zoffix Also MoarVM has already been released.
19:32 timotimo i can't nqp::sprintf("%x", 1) inside nqp?
19:32 timotimo m: use nqp; nqp::sprintf("%x", 1)
19:33 camelia rakudo-moar 83502c: OUTPUT: «Directive x not applicable for type VMNull␤␤»
19:33 timotimo same problem inside rakudo. huh.
19:33 timotimo oh
19:33 timotimo it has to take a list, duh
19:42 samcv good *
19:42 Zoffix \o
19:43 ugexe m: $*OUT.close; $*OUT.open; say 1
19:43 camelia rakudo-moar 83502c: OUTPUT: «Failed to write bytes to filehandle: Bad file descriptor␤  in block <unit> at <tmp> line 1␤␤»
19:45 Zoffix hah
19:46 Zoffix c: 2017.06 $*OUT.close; $*OUT.open; say 1
19:46 committable6 Zoffix, ¦2017.06: «Failed to write bytes to filehandle: Bad file descriptor␤  in block <unit> at /tmp/zc4ctEqWWv line 1␤ «exit code = 1»»
19:48 Zoffix my commit busted it, dinnit
19:48 Zoffix bisect: $*OUT.close; $*OUT.open; say 1
19:48 bisectable6 Zoffix, Bisecting by exit code (old=2015.12 new=83502cc). Old exit code: 0
19:48 bisectable6 Zoffix, bisect log: https://gist.github.com/85aba0f780e5671f4fed5af913481cf2
19:48 bisectable6 Zoffix, (2017-06-07) https://github.com/rakudo/rakudo/commit/1ac7996a42d50fc64c758b6dc2c4eec64d2a6094
19:49 Zoffix huh. ok then
19:49 Zoffix c: 1ac7996a42d50f,1ac7996a42d50f~1 $*OUT.close; $*OUT.open; say 1
19:49 committable6 Zoffix, ¦1ac7996: «Failed to write bytes to filehandle: Bad file descriptor␤  in block <unit> at /tmp/0Q7XBbzFiU line 1␤ «exit code = 1»» ¦1ac7996a42d50f~1: «1»
19:57 ugexe m: note $*OUT.native-descriptor; $*OUT.close; note $*OUT.native-descriptor;
19:57 camelia rakudo-moar 83502c: OUTPUT: «1␤get native descriptor requires an object with REPR MVMOSHandle (got VMNull with REPR Null)␤  in block <unit> at <tmp> line 1␤␤»
19:58 Zoffix Windows spectest seems to hang on t\spec\S32-io\IO-Socket-INET.t :/
19:59 Zoffix Too late to dig into Windows failures now, I guess.
20:01 Zoffix huge amount of failures in spectest actually :/
20:08 ugexe that hanging test might just be slow as fuck
20:08 ugexe it does 2000 iterations and they are taking about a second each on my vm
20:10 Zoffix 0.o0.o
20:10 ugexe the tests pass otherwise... and after 400ish iterations it was still on track to pass that one
20:13 ugexe https://github.com/perl6/roast/commit/db09cd9fe2279be0e1c2d9205a5aa128906f28d4
20:13 ugexe dogbert17: ^
20:20 Ven joined #perl6-dev
20:21 Geth ¦ rakudo/nom: c5d74d8672 | (Zoffix Znet)++ | docs/ChangeLog
20:21 Geth ¦ rakudo/nom: Add all changes to changelog
20:21 Geth ¦ rakudo/nom: review: https://github.com/rakudo/rakudo/commit/c5d74d8672
20:29 Zoffix__ joined #perl6-dev
20:29 ugexe m: for ^10 -> $i { my $sock = IO::Socket::INET.new(:host<127.0.0.1>, :port($i)); CATCH {say $i; next}; }
20:29 camelia rakudo-moar 83502c: OUTPUT: «0␤1␤2␤3␤4␤5␤6␤7␤8␤9␤»
20:29 ugexe thats taking 10 seconds on windows
20:30 ugexe it goes twice as fast if the host/port is legit
20:35 NeuralAnomaly joined #perl6-dev
20:35 Zoffix NeuralAnomaly: status
20:35 NeuralAnomaly Zoffix, [✔] Next release is today. Since last release, there are 25 new still-open tickets (0 unreviewed and 0 blockers) and 0 unreviewed commits. See http://perl6.fail/release/stats for details
20:36 NeuralAnomaly joined #perl6-dev
20:36 Zoffix NeuralAnomaly: cut the release
20:37 Zoffix yo! robot!
20:37 Zoffix Do things!
20:37 geekosaur bottled...
20:37 Zoffix joined #perl6-dev
20:38 NeuralAnomaly joined #perl6-dev
20:38 Zoffix NeuralAnomaly: cut the release
20:38 NeuralAnomaly Zoffix, Will do! If you're feeling particularly naughty, you can watch me at http://perl6.fail/release/progress or go look at some cats http://icanhas.cheezburger.com/lolcats
20:38 NeuralAnomaly Zoffix, ♥♥♥♥♥♥ Prep done
20:38 dogbert17 heh, why on earth should that test be dog slow?
20:38 Geth ¦ nqp: af3a53fe51 | (Zoffix Znet)++ | tools/build/MOAR_REVISION
20:38 Geth ¦ nqp: bump MoarVM version to 2017.07
20:38 Geth ¦ nqp: review: https://github.com/perl6/nqp/commit/af3a53fe51
20:38 Geth ¦ nqp: version bump brought these changes: https://github.com/MoarVM/MoarVM/compare/2017.06-91-g146c8fc...2017.07
20:38 Geth ¦ nqp: b2b8c93b78 | (Zoffix Znet)++ | VERSION
20:38 Geth ¦ nqp: bump VERSION to 2017.07
20:38 Geth ¦ nqp: review: https://github.com/perl6/nqp/commit/b2b8c93b78
20:38 Geth ¦ nqp: version bump brought these changes: https://github.com/MoarVM/MoarVM/compare/2017.06-91-g146c8fc...2017.07
20:48 NeuralAnomaly Zoffix, ♥♥♥♥♥♥ nqp tests OK
20:57 NeuralAnomaly Zoffix, ♥♥♥♥♥♥ nqp release tarball tests OK
20:57 NeuralAnomaly Zoffix, ♥♥♥♥♥♥ nqp release DONE
20:58 Geth ¦ rakudo/nom: bb4f7a17b6 | (Zoffix Znet)++ | docs/announce/2017.07.md
20:58 Geth ¦ rakudo/nom: Generate release announcement for 2017.07
20:58 Geth ¦ rakudo/nom: review: https://github.com/rakudo/rakudo/commit/bb4f7a17b6
20:58 Geth ¦ rakudo/nom: version bump brought these changes: https://github.com/perl6/nqp/compare/2017.06-62-g8cd835f...2017.07
20:58 Geth ¦ rakudo/nom: 03d93b77bd | (Zoffix Znet)++ | tools/build/NQP_REVISION
20:58 Geth ¦ rakudo/nom: [release] bump NQP revision
20:58 Geth ¦ rakudo/nom: review: https://github.com/rakudo/rakudo/commit/03d93b77bd
20:58 Geth ¦ rakudo/nom: version bump brought these changes: https://github.com/perl6/nqp/compare/2017.06-62-g8cd835f...2017.07
20:58 Geth ¦ rakudo/nom: ee915857f5 | (Zoffix Znet)++ | VERSION
20:58 Geth ¦ rakudo/nom: [release] bump VERSION to 2017.07
20:58 Geth ¦ rakudo/nom: review: https://github.com/rakudo/rakudo/commit/ee915857f5
20:58 Geth ¦ rakudo/nom: version bump brought these changes: https://github.com/perl6/nqp/compare/2017.06-62-g8cd835f...2017.07
21:00 Zoffix damn, forgot to put back the 6.c-errata test files that were removed
21:01 Zoffix Oh well. It's not like I didn't say they shouldn't be removed more than once :/
21:07 NeuralAnomaly Zoffix, ☠☠☠☠☠☠ Rakudo: make stresstest (master)
21:07 NeuralAnomaly Zoffix, ☠☠☠☠☠☠☠☠☠☠ ABNORMAL EXIT!
21:07 Zoffix great
21:13 Geth ¦ rakudo/nom: 58d93b4630 | (Zoffix Znet)++ | t/spectest.data
21:13 Geth ¦ rakudo/nom: Restore deleted test files
21:13 Geth ¦ rakudo/nom:
21:13 Geth ¦ rakudo/nom: The files are still used by 6.c-errata and can't currently be removed,
21:13 Geth ¦ rakudo/nom: even if they were relocated on master.
21:13 Geth ¦ rakudo/nom: review: https://github.com/rakudo/rakudo/commit/58d93b4630
21:16 Zoffix What's the story behind S04-phasers/eval-in-begin.t being removed?
21:18 Geth ¦ roast: 1826cee917 | (Zoffix Znet)++ | 4 files
21:18 Geth ¦ roast: Add dummy test files
21:18 Geth ¦ roast:
21:18 Geth ¦ roast: For test files that were removed/relocated in master but still exist
21:18 Geth ¦ roast: in 6.c-errata.
21:18 Geth ¦ roast:
21:18 Geth ¦ roast: Eventually these will be tossed from master, when we have per-language-version
21:18 Geth ¦ roast: t/spectest.data files.
21:18 Geth ¦ roast: review: https://github.com/perl6/roast/commit/1826cee917
21:18 Zoffix NeuralAnomaly: run r-stress r-stress-v6c r-tar r-tar-build r-tar-p5 r-tar-stress r-tag r-tar-sign r-tar-copy post-scp
21:21 NeuralAnomaly Zoffix, ☠☠☠☠☠☠ Rakudo: make stresstest (master)
21:21 NeuralAnomaly Zoffix, ☠☠☠☠☠☠☠☠☠☠ ABNORMAL EXIT!
21:26 Zoffix NeuralAnomaly: run r-stress r-stress-v6c r-tar r-tar-build r-tar-p5 r-tar-stress r-tag r-tar-sign r-tar-copy post-scp
21:26 Zoffix oh, for fuck's sake... all the floppers
21:27 NeuralAnomaly joined #perl6-dev
21:27 Zoffix NeuralAnomaly: run r-stress r-stress-v6c r-tar r-tar-build r-tar-p5 r-tar-stress r-tag r-tar-sign r-tar-copy post-scp
21:28 NeuralAnomaly joined #perl6-dev
21:28 Zoffix Really need to spend some time with them and find what makes them flop
21:28 Zoffix NeuralAnomaly: run r-stress r-stress-v6c r-tar r-tar-build r-tar-p5 r-tar-stress r-tag r-tar-sign r-tar-copy post-scp
21:30 NeuralAnomaly joined #perl6-dev
21:30 Zoffix NeuralAnomaly: run r-stress-v6c r-tar r-tar-build r-tar-p5 r-tar-stress r-tag r-tar-sign r-tar-copy post-scp
21:32 NeuralAnomaly Zoffix, ♥♥♥♥♥♥ Rakudo stresstest (6.c-errata) OK
21:32 Zoffix "The Joel Test: 12 Steps to Better Code: #5 Do you fix bugs before writing new code?"  I think we should be doing more of that. https://www.joelonsoftware.com/2000/08/09/the-joel-test-12-steps-to-better-code/
21:32 Zoffix There's now like 4-6 flopping test files.
21:33 Zoffix There's about 10+ failing Windows test files
21:33 Zoffix buggable: tags
21:33 buggable Zoffix, Total: 1621; BUG: 1073; UNTAGGED: 371; LTA: 171; NYI: 95; JVM: 60; RFC: 57; REGEX: 56; CONC: 51; TESTNEEDED: 31; UNI: 28; PERF: 26; SEGV: 26; IO: 22; NATIVECALL: 22; POD: 21; REGRESSION: 21; @LARRY: 20; TODO: 18; PRECOMP: 14; BUILD: 11; OO: 11; TESTCOMMITTED: 11; OPTIMIZER: 10; STAR: 7; BOOTSTRAP: 5; REPL: 5; GLR: 4; MATH: 4; OSX: 4; WEIRD: 4; SPESH: 3; WINDOWS: 3; RT: 2; BELL
21:33 Zoffix And 1621 bug tickets.
21:37 Zoffix And 30 mentions of RT tickets in source, which I'm guessing are all the temporary fixes
21:39 Zoffix TDD is also good. Adding new feature? Write good tests for it first. 'cause otherwise there's all this extra wasted time with tickets created and people having to go into code and familiriaze themselves with it before being able to fix bugs that could've been caught right when the feature was added, when all the code is still fresh in the mind.
21:40 Zoffix </rant>
21:40 NeuralAnomaly Zoffix, ♥♥♥♥♥♥ Rakudo release DONE
21:40 NeuralAnomaly Zoffix, ♥♥♥♥♥♥ Post: upload tarballs to rakudo.org and rakudo.perl6.org
21:40 NeuralAnomaly Zoffix, ????????????????????????????????????????????????????????????
21:40 NeuralAnomaly Zoffix, The release of **Rakudo #113 2017.07** has now been completed
21:40 NeuralAnomaly Zoffix, ????????????????????????????????????????????????????????????
21:40 * NeuralAnomaly celebrates with an appropriate amount of fun
21:40 Zoffix yey
21:41 dogbert17 which files are flapping?
21:41 NeuralAnomaly joined #perl6-dev
21:41 Zoffix dogbert17: https://irclog.perlgeek.de/perl6-dev/search/?nick=&amp;q=ZOFFLOP
21:43 dogbert17 I guess we need to figure out why the are behaving so badly :)
21:43 Zoffix lizmat: Mission accomplished: "The following people contributed to this release: Elizabeth Mattijsen, Zoffix Znet, ..."
21:44 lizmat :-)
21:46 Geth ¦ rakudo/nom: d37be24961 | (Zoffix Znet)++ (committed using GitHub Web editor) | docs/release_guide.pod
21:46 Geth ¦ rakudo/nom: 2017.07 is now in the past
21:46 Geth ¦ rakudo/nom: review: https://github.com/rakudo/rakudo/commit/d37be24961
21:47 Zoffix I'm done.
21:48 dogbert17 t/spec/S17-promise/nonblocking-await.t is known to be broken
21:48 * Zoffix passes the torch to AlexDaniel
21:48 dogbert17 Zoffix++
21:48 Zoffix .tell AlexDaniel For release stuff, there are these docs and posts https://github.com/rakudo/rakudo/blob/nom/docs/release_guide_automated.md https://rakudo.party/post/Perl6-Fail-Release-Robots-And-Other-Goodies
21:48 yoleaux Zoffix: I'll pass your message to AlexDaniel.
21:51 Zoffix .tell AlexDaniel if you wanted to run the bot on your own system, would be good to practice with test repos. I recall it being a PITA to get it going due to stuff that needs to be installed (the rakudo.party post should detail it). Bot's repo is https://github.com/zoffixznet/na and perl6.fail repo is in https://github.com/zoffixznet/r6 On perl6.fail you can go to https://perl6.fail/release/stats and it'll
21:51 yoleaux Zoffix: I'll pass your message to AlexDaniel.
21:51 Zoffix tell you which commits aren't in chagnelog yet. After you write them down, click on the question mark button to mark them as logged. Same for tickets.
21:51 Zoffix .tell AlexDaniel tell you which commits aren't in chagnelog yet. After you write them down, click on the question mark button to mark them as logged. Same for tickets.
21:51 yoleaux Zoffix: I'll pass your message to AlexDaniel.
21:51 * Zoffix goes on a needed de-burnout from Perl 6
21:51 Zoffix Gonna be (finishing) learning C
21:51 dogbert17 relax and play some cool new 3d games
21:52 lizmat Zoffix++   # hanging in there!
22:07 ugexe m: say $*ERR.native-descriptor; $*ERR.close; $*ERR.open; say $*ERR.native-descriptor
22:07 camelia rakudo-moar d37be2: OUTPUT: «2␤-1␤»
22:07 ugexe m: say $*OUT.native-descriptor; $*OUT.close; $*OUT.open; say $*OUT.native-descriptor
22:07 camelia rakudo-moar d37be2: OUTPUT: «1␤Failed to write bytes to filehandle: Bad file descriptor␤  in block <unit> at <tmp> line 1␤␤»
22:07 ugexe that difference probably hints at something
22:13 geekosaur urgh. never close descriptors 0, 1, 2; you can't get them back
22:14 jnthn Zoffix++ # release
22:15 jnthn Fully agree on writing tests when adding new stuff, fwiw. Of course, somebody will probably think of a case you didn't consider and thus think to test...
22:26 geekosaur joined #perl6-dev
22:43 Ven joined #perl6-dev

| Channels | #perl6-dev index | Today | | Search | Google Search | Plain-Text | summary