»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'p6: say 3;' or rakudo:, or /msg camelia p6: ... | irclog: irc.perl6.org or colabti.org/irclogger/irclogger_logs/perl6 | UTF-8 is our friend!
Set by moritz on 22 December 2015.
00:02 DarthGandalf left, skids joined 00:03 DarthGandalf joined 00:08 mulk[m] left, Matias[m] left, Matthew[m] left, tadzik left, ilmari[m] left, M-Illandan left, dp[m] left 00:11 DarthGandalf left, DarthGandalf joined 00:14 wamba joined, cpage_ left 00:15 maxp_ joined 00:20 DarthGandalf left, DarthGandalf joined
maxp_ Can I ask you a question about grammars? But I must to show you some code 00:23
MasterDuke maxp_: if it's more than one or two lines it's best to use a gist service 00:24
maxp_ thanks! I will use it
00:26 lukaramu left
maxp_ so... I have this program gist.github.com/anonymous/60831794...de975139b8 00:26
the file it read is this: gist.github.com/anonymous/e5473780...42d277084d 00:27
the grammar seems not work but the 'InsFunc' rule if I uncomment it works 00:28
can you explain me why?
MasterDuke maxp_: i'm not great with grammars, but one thing i noticed here gist.github.com/anonymous/60831794...le1-txt-L5 is you have whitespace between the \s* and the string 00:29
when using a rule, whitespace (in the rule itself) is significant 00:30
00:30 pyrimidine left
maxp_ yes but in the InsFunc rules ther is the whitespace too but it works 00:30
00:30 pyrimidine joined
MasterDuke oh, and when you have a grammar with a TOP, the entire input must match the TOP 00:32
it kind of has an implicit ^^ and $$ 00:33
(or ^ and $, not exactly sure which) 00:34
maxp_ ok i try
MasterDuke what happens if you add .* before and after <InsertFunc> in TOP?
00:35 pyrimidine left
MasterDuke if you haven't seen it before, docs.perl6.org/language/grammar_tutorial is a nice intro to Perl 6 grammars 00:36
babydrop maxp_: but the TOP is also a rule. Try changing them to tokens
maxp_ yes i've seen it but it doesn't explain the implicit ^^ $$ 00:37
00:37 labster joined
maxp_ but... it works now i changed with this: 00:37
babydrop oh doh
maxp_ gist.github.com/anonymous/516419b4...c92ec5acae 00:38
not it works
babydrop maxp_: right, .parse parses from start to end. It's implict ^ and $
maxp_ thank you very much!
MasterDuke ah, docs.perl6.org/language/grammars#TOP mentions it
welcome
maxp_ Ok I think I'm a bad reader
so
bye 00:39
MasterDuke it's not obvious
babydrop maxp_: if that works, then subparse will work instead of .*
maxp_ ok i will try
babydrop Though I don't see how what you pasted can work, since the token you're trying to match is not at the start of file :/
MasterDuke "for .lines" 00:40
SmokeMachine should I create a bug report?
www.irccloud.com/pastebin/WOe90w5h/
maxp_ because the insert new lines has whitespaces before and ';' or () characters;
.+ match them
MasterDuke SmokeMachine: what's the Perl 6 code that triggered that? 00:41
SmokeMachine its too big to post it here...
Ill try to reduce that... 00:42
maxp_ babydrop: I've changed with 00:44
this and it works too: rule TOP { .* <InsertFunc> .* }
because the whitespaces i presume 00:45
00:48 cpage_ joined 00:49 aborazmeh joined, aborazmeh left, aborazmeh joined 00:54 canopus left, bjz left 01:00 canopus joined
AlexDaniel SmokeMachine: of course… 01:05
SmokeMachine: no, don't create a bug report! Let's have this stinky bug there :P
SmokeMachine: is it reproducible each time? 01:06
SmokeMachine AlexDaniel: Im not sure that its a bug... (I think it is... but im no sure...)
it gives me different errors each time... 01:07
AlexDaniel SmokeMachine: what makes you think it's not? Are you using some NativeCall stuff?
yes, that's ok
01:07 lichtkind_ left
SmokeMachine the most common is Segmentation fault: 11 01:07
AlexDaniel yea, that's how most [SEGV] bugs behave
SmokeMachine no... Im using Suplies
AlexDaniel ok now that's interesting :) 01:08
SmokeMachine: can you post some code? I can try golfing it
SmokeMachine *Supplies 01:09
01:21 pmurias joined
SmokeMachine AlexDaniel: the code is: 01:23
www.irccloud.com/pastebin/IOFVFmBT/ 01:24
www.irccloud.com/pastebin/6ZaHz5cx/
AlexDaniel SmokeMachine: resources/classes ? 01:29
01:29 mr_ron left
SmokeMachine www.irccloud.com/pastebin/1bIasgzI/ 01:29
01:30 mr_ron joined
babydrop m: say "{7.2610255 / 0.578650}x faster" 01:30
camelia rakudo-moar 5476d6: OUTPUT«12.54821654x faster␤»
AlexDaniel SmokeMachine: I am getting “Cannot find method 'EXISTS-KEY': no method cache and no .^find_method” 01:31
any ideas? 01:32
SmokeMachine: somewhere in sub bla 01:33
babydrop Is there a trick on GitHub to view Blame for parent's commit? Right now I do parent commit -> find files -> find file -> blame again
SmokeMachine AlexDaniel: not yet...
are you running with --fuzz
?
babydrop Oh wow... the NaN thing wasn't my error... So I can actually claim the speed increase to my name \o/ 01:34
(it was added in 2014)
AlexDaniel SmokeMachine: oh 01:35
SmokeMachine: no, but doesn't help
that's a compile time error
01:36 bjz joined, wamba left
DrForr Okay, finally got back online. I take it I need to get access to the advent admin access - I'm actually worried about network connectivity out here. 01:37
AlexDaniel SmokeMachine: what rakudo version do you have? 01:38
babydrop DrForr: what's your username?
SmokeMachine AlexDaniel:$ perl6 -v 01:40
This is Rakudo version 2016.11-172-g5476d60 built on MoarVM version 2016.11-41-gd2139b5
implementing Perl 6.c.
01:40 seatek joined
babydrop that's HEAD 01:41
AlexDaniel SmokeMachine: well, that's the error I'm getting: gist.github.com/AlexDaniel/903500b...861b64e9ee 01:42
it's actually somewhere in sub fuzz I think
MasterDuke babydrop: closest thing i could find in the git blame doc was one of its examples: "git blame -C -C -f $commit^! -- foo" - annotate the change between the commit and its parents
babydrop Well, I was looking for GitHub thing....
There's probably some weird shortcut key you need to press :( 01:43
SmokeMachine AlexDaniel: now I got this:
MoarVM panic: Trying to unwind from wrong handler
AlexDaniel SmokeMachine: well that's great but how do I reproduce it if it doesn't compile :(
2016.11-156-g0428b79 here, not head but still 01:44
dalek c: 2f8edca | samcv++ | doc/Type/Cool.pod6:
Trigger doc rebuild
01:45
synopsebot6 Link: doc.perl6.org/type/Cool
01:45 Actualeyes joined 01:46 Pbv joined
SmokeMachine AlexDaniel: github.com/FCO/Test-Fuzz/tree/ERROR 01:47
babydrop samcv: you around? 01:48
samcv jyeah
AlexDaniel SmokeMachine: ok that works 01:49
SmokeMachine: segfault! Yay!
SmokeMachine \o/
babydrop samcv: oh sorry, I misread the diff. Was gonna say L<uninames|#routine_uninames> in github.com/perl6/doc/commit/37d0646eb4f won't work on generated pages like docs.perl6.org/routine/uniname and was gonna suggest that as a reason for why the doc build is failing... But now I see that link already existed before 01:50
samcv ah 01:51
01:54 pmurias left
AlexDaniel it does seem to crash without optimizations :) 01:54
ok let's see…
babydrop Travis ( travis-ci.org/perl6/doc/builds/182733730#L2191 ) says the failure is here github.com/perl6/doc/blob/master/h...fy.p6#L747 01:55
well, somewhere in that block...
We also have build logs here docs.perl6.org/build-log/ 01:56
babydrop makes a wild guess that it's the '#' in URL L<General Category|en.wikipedia.org/wiki/Unicode_char...tegory> 02:00
nope, there are others like it :/
02:02 travis-ci joined
travis-ci Doc build failed. Samantha McVey 'Trigger doc rebuild' 02:02
travis-ci.org/perl6/doc/builds/182772719 github.com/perl6/doc/compare/37d06...8edcade3d9
02:02 travis-ci left
samcv ugh 02:02
babydrop Well, the error is still there somewhere :} 02:03
ohhh
samcv: found it
m: '|'.uniname.say
camelia rakudo-moar 5476d6: OUTPUT«VERTICAL LINE␤»
babydrop samcv: you got some fancy pants editor? 02:04
m: '|'.uniname.say
camelia rakudo-moar 5476d6: OUTPUT«VERTICAL LINE␤»
babydrop Bah
Pbv how generate html from perl6?
babydrop This is normally where I say I'm too drunk, right? :)
samcv wait where is the error though
babydrop Oh, what is it?
samcv line no?
or maybe not talking to me
babydrop samcv: I was but I'm wrong :) 02:05
I had a brain fart.
The error is somewhere in this commit: github.com/perl6/doc/commit/37d064...cfb942fedd
The one before it got built and now the build shows this error (see bottom): docs.perl6.org/build-log/build-201...0+0000.log 02:06
And I'm guessing that's the link generator.
Pbv: that's a very vague question without an answer.
Pbv ok thank you 02:08
02:08 Pbv left 02:09 BenGoldberg joined
babydrop :/ 02:09
SmokeMachine AlexDaniel: I think that's some kind of race condition... 02:11
dalek c: 7130500 | samcv++ | doc/Type/Cool.pod6:
Try and get it building again
synopsebot6 Link: doc.perl6.org/type/Cool
babydrop shakes the Magic8 ball 02:14
that'll fail :)
I got a local build I'm golfing ATM 02:15
weird AF. If I rename "routine" to "method" for uniprop it builds 02:17
02:18 maxp_ left
AlexDaniel SmokeMachine: the code is kinda complex… cannot golf it past 100 lines 02:18
SmokeMachine :(
02:24 djbkd joined 02:28 pyrimidine joined 02:29 travis-ci joined
travis-ci Doc build failed. Samantha McVey 'Try and get it building again' 02:29
travis-ci.org/perl6/doc/builds/182775705 github.com/perl6/doc/compare/2f8ed...305006cd74
02:29 travis-ci left
samcv why does it hate me 02:29
babydrop is debugging it ATM 02:30
02:32 pyrimidine left
babydrop oh yeah... I can smell a juicy bug.. 02:33
SmokeMachine is this expected?
www.irccloud.com/pastebin/SIvWXJRp/
babydrop no more catfood for babydrop... tonight we dine
SmokeMachine m: loop [+] 1, 2, 3, 4 {.say}
camelia rakudo-moar 5476d6: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Missing block␤at <tmp>:1␤------> 3loop7⏏5 [+] 1, 2, 3, 4 {.say}␤ expecting any of:␤ scoped block␤»
SmokeMachine m: loop ([+] 1, 2, 3, 4) {.say}
camelia rakudo-moar 5476d6: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Malformed loop spec␤at <tmp>:1␤------> 3loop ([+] 1, 2, 3, 47⏏5) {.say}␤»
AlexDaniel what should it mean? :D
babydrop SmokeMachine: you meant `for` not `loop` 02:34
SmokeMachine: `loop` does a C-style loop
AlexDaniel what's the point of for if he has only one element after reduction
SmokeMachine m: loop ([+] 1, 2, 3, 4 > 0) {.say}
camelia rakudo-moar 5476d6: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Malformed loop spec␤at <tmp>:1␤------> 3loop ([+] 1, 2, 3, 4 > 07⏏5) {.say}␤»
babydrop Ah, right
AlexDaniel so what are you trying to do?
02:34 bjz left
SmokeMachine m: @a = 0; loop ([+] @a > 5) {@a.push: 1} 02:35
camelia rakudo-moar 5476d6: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Variable '@a' is not declared␤at <tmp>:1␤------> 3<BOL>7⏏5@a = 0; loop ([+] @a > 5) {@a.push: 1}␤»
02:35 djbkd left, lucasb joined
AlexDaniel just say what you want :) 02:35
babydrop heh
Where's first-code-block from?
SmokeMachine I just want to loop while the sum of all my array is less than 5 (for example) 02:36
02:36 djbkd joined
babydrop m: my @a = ^10; while @a.sum > 5 { "+".say; @a.shift } 02:37
camelia rakudo-moar 5476d6: OUTPUT«+␤+␤+␤+␤+␤+␤+␤+␤+␤+␤»
babydrop m: my @a = ^10; while @a.sum > 5 { @a.shift.say }
camelia rakudo-moar 5476d6: OUTPUT«0␤1␤2␤3␤4␤5␤6␤7␤8␤9␤»
lucasb babydrop: see the blame of the parent in github?
babydrop m: my @a = ^10; while @a.sum > 5 { @a.pop.say }
camelia rakudo-moar 5476d6: OUTPUT«9␤8␤7␤6␤5␤4␤3␤»
lucasb yeah, I wish it was easier than it is
AlexDaniel m: for ([\+] 1, 2, 3, 4) { last if $_ > 7; .say }
camelia rakudo-moar 5476d6: OUTPUT«1␤3␤6␤»
lucasb I do a little dance too, to get to it
babydrop star: use JSON::Tiny; say &from-json.WHO 02:38
camelia star-m 2016.10: OUTPUT«Sub␤»
babydrop Man, how do I find which of the modules is sub being exported from?
star: use JSON::Tiny; say &from-json.file
camelia star-m 2016.10: OUTPUT«site#sources/9B467EEF9267A777BB53BAA2F19BE2C9D756BEED (JSON::Tiny)␤»
babydrop good nuf I guess
lucasb blame -> wanted commit -> "View" in the desired file -> History -> second commit in the list -> find desired file and click on View again -> click on Blame again 02:39
SmokeMachine sure! while!!! whats happening to me?!!!
AlexDaniel SmokeMachine: nothing wrong. ↑ there's example with [\+] that works just as well 02:40
SmokeMachine :)
babydrop lucasb: well, I can save you a step there the parent commit is a link at the top of the page near commit sha
AlexDaniel still not loop though :P
babydrop or... I guess that's the parent's sha
lucasb babydrop: nope, that doesn't help :) 02:41
AlexDaniel SmokeMachine: as for the segfault, yeah… dunno. You can try submitting a ticket
but not sure if that's going to be very helpful
02:41 bjz joined
AlexDaniel you can also run it under valgrind 02:41
but yeah even that probably won't help anything
SmokeMachine :(
AlexDaniel: thank you very much! 02:42
AlexDaniel SmokeMachine: for example, there is stuff like this: rt.perl.org/Ticket/Display.html?id=128870
SmokeMachine: so it's hard to tell if its rakudo's fault in this case 02:43
I mean… the code is a bit too complex to see :-/ 02:44
babydrop samcv: got it
samcv tell me!
babydrop samcv: for `routine` names, the build script tries to look up the sub in the referenced class. So it looks in Cool.pm and fails, because the sub is actualyl defined in Str.pm 02:45
samcv true 02:46
babydrop samcv: so just change `routine` to `method` in the `=head2` and I guess that'll do?
02:46 ilbot3 left
lucasb ah, the bot always stop logging the channel at this time? then he comes back some minutes later? 02:46
dalek c: 258c4de | samcv++ | doc/Type/Cool.pod6:
Change routine to method
02:47
synopsebot6 Link: doc.perl6.org/type/Cool
02:47 ilbot3 joined 02:52 lucasb left
dalek c: e5ef1d0 | (Zoffix Znet)++ | htmlify.p6:
Awesomify error of htmlfy.p6 script

  ... when routine is not found in a class's file.
02:53
02:53 kyclark joined
babydrop lucasb: reminds me of www.bash.org/?42196 02:54
02:54 DarthGandalf left
dalek c: a3fd3f2 | (Zoffix Znet)++ | htmlify.p6:
Fix typo
02:56
02:57 cdg_ joined 02:59 cdg left 03:05 RubioTerra left, travis-ci joined
travis-ci Doc build passed. Samantha McVey 'Change routine to method' 03:05
travis-ci.org/perl6/doc/builds/182779578 github.com/perl6/doc/compare/71305...8c4deffaf8
03:05 travis-ci left 03:11 travis-ci joined
travis-ci Doc build passed. Zoffix Znet 'Awesomify error of htmlfy.p6 script 03:12
travis-ci.org/perl6/doc/builds/182780098 github.com/perl6/doc/compare/258c4...ef1d0b1e77
03:12 travis-ci left 03:14 noganex joined, travis-ci joined
travis-ci Doc build passed. Zoffix Znet 'Fix typo' 03:14
travis-ci.org/perl6/doc/builds/182780328 github.com/perl6/doc/compare/e5ef1...fd3f25ad65
03:14 travis-ci left
babydrop go away.... 03:14
03:17 imcsk8_ joined, noganex_ left 03:22 AlexDaniel left 03:28 xtreak joined 03:35 baest_ joined 03:36 cgfbee left, nchambers^ joined 03:37 cgfbee joined, zoosha joined, KotH_ joined, hoelzro_ joined, jantore_ joined, bartolin_ joined, konobi_ joined, hahainte1net joined, tony-o_ joined, cosimo_ joined, mst_ joined, tinita_ joined 03:38 cognominal left, kaare_ left, espadrine left, espadrine joined, avalenn joined 03:40 baest left 03:46 pyrimidine joined 03:51 pyrimidine left 03:54 kyclark left 03:55 BenGoldberg left
samcv ok so i got rakudo returning bools for applicable uniprops github.com/rakudo/rakudo/pull/944 04:08
04:11 Sgeo_ joined 04:12 Sgeo left, aborazmeh left 04:16 kyclark joined 04:18 skids left 04:19 pyrimidine joined 04:23 pyrimidine left 04:32 kyclark left 04:36 cibs left 04:38 cibs joined 04:49 khw left 04:58 aborazmeh joined, aborazmeh left, aborazmeh joined
DrForr Anyone around to point me to the way to get the perl6advent account going?... My 'nt connection is ... pretty much cr*p. 04:59
I can't even get to my linode without stalling. 05:02
05:09 mr_ron left
DrForr babydrop: (or anyone) I have no 'net connection. I can do 20th-24th because I'll be be back home. 05:17
05:21 cdg_ left 05:43 bjz left 05:44 bjz_ joined 05:52 Cabanossi left, Cabanossi joined 05:56 nchambers^ left, nchambers^ joined, nchambers^ is now known as nchambers 06:00 rurban joined, rurban left 06:01 pyrimidine joined 06:06 pyrimidine left 06:15 cpage_ left 06:17 aborazmeh left 06:18 Sgeo_ left 06:20 bjz_ left 06:25 bjz joined 06:41 cpage_ joined 06:46 telex left, Actualeyes left 06:48 telex joined 06:52 Tonik joined
samcv okay so I fixed the problem with unicode short names not working properly and fixed uniprop('Name') replying 0 with this PR 06:55
github.com/rakudo/rakudo/pull/944
moritz DrForr: /msg me an email address, and I can invite you to the blog
samcv my first substantial PR for rakudo, so hope it's already
i included a script i used to test it and generate the code from the unicode txt file 06:56
06:57 darutoko joined 06:58 hometruck joined
hometruck www.reddit.com/r/perl6/comments/5h...us_regexs/ how come it's so slow? 06:58
samcv MoarVM doesn't implement any really fast string finding method like Boyer-Moore yet 07:00
sadly :(
07:02 troys joined
samcv also can anybody direct me about who to ask if I can implement a `uniprops` function. we already have `uniprop`, and we have `uniname` and `uninames`. `uniprops` was also in the synopses (WIP), so is there 07:03
a mailing list i should ask or is here fine?
geekosaur #perl6-dev might be better. the lists are sadly useless for this kind of thing 07:06
samcv ah ok thanks 07:07
07:12 labster left, xtreak left 07:13 pyrimidine joined 07:15 xtreak joined
hometruck the boss said it would be working on it: irclog.perlgeek.de/perl6/2016-11-27#i_13637044 07:16
seatek Stuff like this is just so slick for doing types :) subset Email of Str where { Email::Valid.new.validate($_) }; 07:17
07:17 pyrimidine left 07:18 cognominal joined
seatek I've been kinda going nuts with typing everything 07:18
hometruck also the Ulti feature above that seems extremely interesting: record based parsing of complex files
don't believe the current engine can do that technically 07:19
(without hacks)
07:20 troys left
samcv nice seatek 07:20
seatek yeah simple silly stuff like that has been saving SO much time and code 07:22
samcv: did you ever get your backwards parenthesis figured out? :) 07:23
samcv yeah i did 07:31
hah
hometruck me too: somehow those silly stuff fits perfectly my ways of doing things
samcv i wanted to do an advent but i have no clue what I would do it on 07:32
hometruck I'm not an expert, but I have the impression that there are people payed to do things right, and people payed to do things fast
samcv that hasn't been done before
07:39 sjoshi joined, mohae left 07:43 labster joined 07:49 domidumont joined 07:50 pierre__ joined, wamba joined 07:54 domidumont left, bjz left, domidumont joined 07:58 bjz joined 08:07 sjoshi left 08:08 bjz_ joined 08:09 bjz left 08:13 cibs left 08:14 CIAvash joined 08:15 cibs joined 08:18 pyrimidine joined 08:20 pyrimidine left, pyrimidi_ joined, bjz_ left, djbkd__ joined, pierre__ left 08:22 wamba left, raiph left, wamba joined, dontbe^2 left 08:23 xtreak left, djbkd left, dontbe^2 joined 08:24 pyrimidi_ left 08:25 Sgeo joined 08:30 Actualeyes joined 08:36 kshannon joined 08:41 CIAvash left, cibs left, cibs joined 08:43 rurban joined 08:47 wamba left 08:49 hometruck left, pecastro joined, bjz joined 08:52 wamba joined 08:53 dontbe^2 left, revdiablo left 08:54 Tonik left, vike left, Tonik joined 08:55 revdiablo joined 09:00 Actualeyes left, cibs left 09:08 RabidGravy joined 09:10 cibs joined, geraud1 joined 09:11 dontbe^2 joined, mtj_ joined, k-man joined, raydiak joined, alnewkirk joined, charsbar______ joined 09:13 cibs left 09:14 Actualeyes joined 09:15 cibs joined, sena_kun joined 09:16 xtreak joined
nine Did you know that there's a Japanese Perl 6 advent calendar? qiita.com/advent-calendar/2016/perl6 09:19
09:22 pyrimidine joined
nine pyrimidine: found and reverted the MoarVM commit that broke Inline::Perl 5. Thanks for your help :) 09:24
RabidGravy morning! 09:28
09:32 sena_kun left, dontbe^2 left, mtj_ left, k-man left, raydiak left, alnewkirk left, charsbar______ left 09:33 sena_kun joined, dontbe^2 joined, mtj_ joined, k-man joined, raydiak joined, alnewkirk joined, charsbar______ joined 09:35 lukaramu joined 09:39 setty1 left
RabidGravy jnthn, yeah I think I added Chronic.at because someone asked for it, and it was easy with the other scheduling stuff in there, but a Promise.at would be useful if it could be implemented efficiently 09:41
09:52 cibs left, geraud1 left 09:58 cibs joined, geraud1 joined 10:02 andrzejku joined, FROGGS joined 10:10 wamba left
RabidGravy I think an order of the day might be for me to make a thing to periodically clear out the .precomp in my working directories 10:27
some of the older modules there are hundreds of files
10:39 rindolf joined 10:49 pyrimidine left 10:52 pyrimidine joined, xtreak left 10:55 xtreak joined 10:56 pyrimidine left 10:58 Sgeo left 10:59 TEttinger left, xtreak left, andrzejku left 11:00 labster left 11:02 Sgeo joined, rurban left 11:09 pierre__ joined 11:21 CIAvash joined 11:22 lucasb joined
RabidGravy ponders whether to perform the Fedora 25 upgrade *before* he does all the other things 11:25
lucasb has a eternal item in his TODO list: "upgrade the distribution" 11:27
RabidGravy I used to be much more reluctant as about 50% of the time it would result in having to re-install the whole thing anyway, but the last several times it has just worked 11:30
lucasb hey, very nice japanese advent calendar. I can't understand a single thing, but I'm looking at the code snippets and they are well formatted 11:34
11:37 pierre__ left
tbrowder .tell [Coke] Any suggestions or criticisms on my draft 12 Dec Advent blog? 11:38
titsuki lucasb: Thanks for reading posts. Qiita currently doen't have formatter for Perl6. So I used Perl5's fomatter instead of Perl6's one. 11:39
dalek c: c08a742 | gfldex++ | doc/Language/quoting.pod6:
improve index entry
11:43
synopsebot6 Link: doc.perl6.org/language/quoting
11:49 pierre__ joined 11:51 CIAvash left 11:52 domidumont left 11:55 pierre__ left 11:56 hahainte1net is now known as hahainternet 11:58 AlexDaniel joined
lucasb if methods are has-scoped. does 'my method' and 'our method' makes sense? 11:59
12:02 lizmat left
RabidGravy I'd say not no 12:03
lucasb yes, I agree
class C { our sub foo {} }; C::foo # this works fine 12:04
moritz so, it seems we don't have an advent post for today yet
lucasb but 'my method' and 'our method' doesn't make sense IMO
moritz anybody know what's up?
lucasb I wish it was invalid syntax, and raised an error 12:05
moritz or is there any article ready that could switch to today?
RabidGravy moritz, who was going to be doing it?
titsuki moritz: I can switch.
moritz titsuki: that would be awesome
RabidGravy: DrForr
RabidGravy ah, he was travelling IIRC, it may be a timezone thing 12:06
where is the schedule again?
sena_kun RabidGravy, github.com/perl6/mu/blob/master/mi...6/schedule 12:07
RabidGravy cheers
good thing I looked as I thought I was doing the 12th not the 13th 12:08
moritz titsuki: I've published your post. Thank you very much! 12:09
sena_kun titsuki++
dalek c: 54f0c34 | gfldex++ | doc/Language/typesystem.pod6:
correct our scoped methods and tell how to call them
synopsebot6 Link: doc.perl6.org/language/typesystem
12:09 wamba joined
titsuki moriz: It's my pleasure. 12:09
moritz lucasb: fwiw rakudo's setting uses 'my method...' 12:10
lucasb: it's useful for generating method to install somewhere, just not in sink context
lucasb moritz: yeah, I was confused. nevermind what i said :)
'my method' and 'our method' makes sense in some contexts, yes 12:11
moritz .tell DrForr we've switched your advent post with titsuki's, so now you have until the 11th to write/publish yours
yoleaux moritz: I'll pass your message to DrForr.
12:14 wamba left
AlexDaniel unicodable6: { .uniprop(‘Dash’) } 12:17
unicodable6 AlexDaniel, U+002D HYPHEN-MINUS [Pd] (-)
AlexDaniel, U+058A ARMENIAN HYPHEN [Pd] (֊)
AlexDaniel, U+05BE HEBREW PUNCTUATION MAQAF [Pd] (־)
AlexDaniel, gist.github.com/55af6a6d3156dddf19...1881c0dd48
sena_kun Since github.com/perl6/doc/pull/1035 doesn't get any new comments/objections, I'll merge it at evening. 12:19
AlexDaniel sena_kun: no need to wait, I'll review it now and merge 12:20
sena_kun AlexDaniel, hmm, okay.
RabidGravy titsuki++ # nice
AlexDaniel sena_kun: what is “Temporary test file deletion” ? 12:21
RabidGravy the example 3 is actually pertinent to my current interests
sena_kun AlexDaniel, it failed a build because of non-closed =table, but the enclusure is already okay in the master and merge will fix it.
nine win 13
AlexDaniel whatever
dalek c: 1308514 | Altai-man++ | doc/Type/ (124 files):
Big signature return type unification.

This commit affects all signatures that exist in the type documentation: it replaces "returns " occurences with "-->" form.
See github.com/perl6/doc/issues/1024#i...-264660360 for explanations.
12:22
c: f8448f2 | Altai-man++ | doc/Type/ (2 files):
Two typo fixes
c: 88c2f5f | Altai-man++ | t/return-type.t:
Test for return type in signatures
c: 40a97e3 | Altai-man++ | doc/Type/ (2 files):
Changes need to resolve a merge conflict.
c: ae816b9 | Altai-man++ | t/return-type.t:
Temporary test file deletion
c: 63d182a | (Aleks-Daniel Jakimenko-Aleksejev)++ | doc/Type/ (121 files):
Merge pull request #1035 from perl6/consistent-return-types

Consistent return types
sena_kun AlexDaniel, I had a conflict, because the test-file was removed in the master and edited in the branch. Git was not so smart to just delete/add, not move it, so the only solution that comes to me was to delete file in the branch and add it later in the master directly.
AlexDaniel sena_kun: 🤷 well, now everything is in master, soo ;) 12:23
sena_kun AlexDaniel, sure, the tests will be soon then!
AlexDaniel sena_kun: so should we put them into t/ or xt/ ?
sena_kun AlexDaniel, xt/, I suppose. 12:24
AlexDaniel okay
it's not like random people contribute signatures very often
but it's ok anyway
as long as we have a test, there is no issue, I think
sena_kun++ 12:25
dalek k-simple: eb2d294 | niner++ | t/01-sanity.t:
Add missing use statements in test file

The design of Perl 6 requires that use statements only have effect in a lexical scope. Otherwise it's impossible to use different (versions of) modules sharing a name in different parts of a program.
This means essentially that we need to explicitly use all modules we need in a scope.
AlexDaniel u: { .uniprop(‘Sentence_Terminal’) } 12:26
unicodable6 AlexDaniel, U+0021 EXCLAMATION MARK [Po] (!)
AlexDaniel, U+002E FULL STOP [Po] (.)
AlexDaniel, U+003F QUESTION MARK [Po] (?)
AlexDaniel, gist.github.com/45c0ef96a7edff3c4b...9f28e9af1d
AlexDaniel 𑗊 – really? 12:27
12:27 mawkish joined
RabidGravy nine, I was think re the lexical use thing that a nice thing to go with it would be to make it possible (or easier,) to re-export types or such, 12:30
babydrop Anyone got their Advent post ready already and can post today?
" DrForrbabydrop: (or anyone) I have no 'net connection. I can do 20th-24th because I'll be be back home."
RabidGravy babydrop, titsuki's got published already 12:31
babydrop oh
RabidGravy so it's probably one for tomorrow
12:31 pyrimidine joined, wamba joined
babydrop huggable: advent 12:32
huggable babydrop, github.com/perl6/mu/blob/master/mi...6/schedule
dalek c: 0e5ca87 | Altai-man++ | xt/return-type.t:
Test for return type of signatures in type documentation
12:33
12:34 kurahaupo joined
babydrop Well, I was gonna write one about bug fixing for perl6.party. I can I can fudge it around a bit and make it apropo for Advent? 12:35
12:36 pyrimidine left
moritz babydrop: sure 12:37
babydrop OK
AlexDaniel m: say 0x0F
camelia rakudo-moar a8ba26: OUTPUT«15␤»
AlexDaniel okay
for those wondering 12:38
u: { .uniprop(‘ASCII_Hex_Digit’) }
unicodable6 AlexDaniel, U+0030 DIGIT ZERO [Nd] (0)
AlexDaniel, U+0031 DIGIT ONE [Nd] (1)
AlexDaniel, U+0032 DIGIT TWO [Nd] (2)
AlexDaniel, gist.github.com/f8f23ba69663c9a7c9...112151224b
AlexDaniel u: { .uniprop(‘Hex_Digit’) }
unicodable6 AlexDaniel, U+0030 DIGIT ZERO [Nd] (0)
AlexDaniel, U+0031 DIGIT ONE [Nd] (1)
AlexDaniel, U+0032 DIGIT TWO [Nd] (2)
AlexDaniel, gist.github.com/982c23079e2eac7004...43a8bd031c
sena_kun Is it me or this example - docs.perl6.org/type/Str.html#method_indent looks LTA? 12:39
AlexDaniel improve it? Perhaps a couple of extra examples won't hurt 12:41
sena_kun Agreed. 12:42
moritz at the Perl 6 training I gave, I tried to describe indent, and after 20s stopped myself, and simply showed it on script in the REPL 12:44
12:47 MasterDuke left
lucasb m: my &to = method (&c) { c(self) }; say 2.&to(* + 3).&to(* * 2) 12:48
camelia rakudo-moar a8ba26: OUTPUT«10␤»
sena_kun moritz, the routine itself is simple, it is the example that makes me nervous. It looks like bad formatting at first glance and like a simple test at second. I don't say it is bad or non-descriptive, just somewhat LTA.
lucasb ^^ a generic method to chain the application of functions into a value 12:49
the name could be something like 'do', 'to', 'yield', 'tap' (but of course, these names conflicts with already existing keywords) 12:50
sena_kun lucasb, just use "."! :P 12:51
12:51 pecastro left, djbkd__ left, nchambers left, baest_ left, rburkholder left, Celelibi left, aindilis left, Unavowed left, cooper_ left, zilti left
babydrop weeeeeeeeeee 12:51
12:52 pecastro joined, djbkd__ joined, nchambers joined, baest_ joined, rburkholder joined, Celelibi joined, aindilis joined, Unavowed joined, cooper_ joined, zilti joined, sivoais joined, dg joined, ttkp6 joined, freeze joined, lizmat joined
RabidGravy boom bang a bang 12:52
babydrop We don't conflict with keywords, we shadow them :)
lucasb sena_kun: just "." ? but I'm not calling methods
dalek c: ea77029 | Altai-man++ | doc/Type/ (6 files):
Removing of `proto` signatures. It is an internal implementation detail, it is noisy with no gain for a reader and we don't have them too many to keep.

Fixes github.com/perl6/doc/issues/1033
babydrop m: sub do {}; sub WHAT {}; sub I {}; sub say { }; do WHAT I say
camelia rakudo-moar a8ba26: OUTPUT«5===SORRY!5===␤Argument to "say" seems to be malformed␤at <tmp>:1␤------> 3{}; sub I {}; sub say { }; do WHAT I say7⏏5<EOL>␤Other potential difficulties:␤ Unsupported use of bare "say"; in Perl 6 please use .say if you meant $_, or u…»
sena_kun lucasb, then "$" application will be good.
babydrop AAAAh
My plan for world domination foiled by grammar 12:53
12:53 tinita_ is now known as tinita
lucasb I wonder if anybody would think it's convenient to have such method into the base object, like Mu/Any ? 12:53
sena_kun: /j haskell
RabidGravy yeah the problem with Perl 6 is that there is such a rich set of well-chosen keywords that it's difficult to find things that don't conflict and still make sense ;-)
12:53 canopus left
babydrop lucasb: I don't get that method at all. What's the point of it? 12:54
m: say 2.&(* + 3).&(* * 2)
camelia rakudo-moar a8ba26: OUTPUT«10␤»
babydrop shrugs
lucasb instead of writing foo(bar(baz(x))), you can write x.to(baz).to(bar).to(foo) 12:55
babydrop you can write x.&baz.&bar.&foo right now, so you're making it worse...
lucasb lol, my method is not needed at all. the feature is built into P6 already
babydrop \o/ 12:56
Yeah, you can use a sub as a method and the invocant will be the first arg (and you can still pass additional args)
lucasb I tried just 'say 2.(* + 3).(* * 2)' and assumed it didn't work
babydrop Ah, yeah, need the & in there
AlexDaniel and you can do it other way round too, I think 12:57
lucasb on a related note, Ruby has a .tap method on the base Object 12:58
it yields self to the block given, but return self itself, not the return value of the block
12:59 lizmat left
lucasb so, it's useful to apply side effects, discard the results and get self back as a return value 12:59
AlexDaniel m: 42.base(2).chars.uniname.say 13:01
camelia rakudo-moar a8ba26: OUTPUT«ACKNOWLEDGE␤»
13:02 MasterDuke joined
AlexDaniel m: say uniname chars base 42: 2 : : : 13:02
camelia rakudo-moar a8ba26: OUTPUT«ACKNOWLEDGE␤»
RabidGravy right, going to bite the bullet and apply this fedora upgrade 13:03
AlexDaniel probably not the best example because all these things are available as subs
m: say uniname chars base 42: 2
camelia rakudo-moar a8ba26: OUTPUT«ACKNOWLEDGE␤»
AlexDaniel so this works too
RabidGravy so see you on the other side
13:03 RabidGravy left 13:04 canopus joined, ufobat joined
dalek : e4981e7 | (Zoffix Znet)++ | misc/perl6advent-2016/schedule:
Claim 11th spot

For "Perl 6 Core Hacking: Sometimes Things Slip Through The QASTs"
  DrForr as unable to do the 10th spot: "<DrForr> babydrop: (or anyone) I have no 'net connection. I can do 20th-24th because I'll be be back home"
  titsuki++ claimed it by moving 11th spot to 10th, so this post covers the now-empty 11th.
13:06
13:07 andrzejku joined 13:08 mawkish left, mawkish joined, lizmat joined
dalek c: 447e029 | gfldex++ | doc/Type/Callable.pod6:
show relationship between .& and ∘
13:09
synopsebot6 Link: doc.perl6.org/type/Callable
lucasb m: 2 ==> (*+3)() ==> (* * 2)() ==> say() 13:11
camelia rakudo-moar a8ba26: OUTPUT«10␤»
13:12 mawkish left
lucasb ^^ I tried to write using fewer parentheses, but it doesn't work without them :( 13:12
13:12 mawkish joined
moritz m: say 10 # fewer parens, same result :-) 13:13
camelia rakudo-moar a8ba26: OUTPUT«10␤»
cognominal zzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz\|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| 13:14
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||\|||\\\z\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\
\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\||||||||||||||||||||||||||||||||||||||\\\\\\\\\\\|||||||||\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\||\\||\\\\\\\\\\\\\\\\\\\\\\\\\\||||||||||\||||\\|-po7qqqp-oooppp-pppppppppppppppppoooooooooooooooooooooooooooo--p-oj o
babydrop yowser
AlexDaniel please keep calm :)
lucasb Christmas spirit is making everybody funny
babydrop That feels like a feline trying to communicate...
cognominal sorry
lucasb It surely is valid Perl 6 code 13:15
kurahaupo cognominal: cat on keyboard? phone in pocket? scamps on your WiFi?
13:17 pyrimidine joined, Actualeyes left 13:19 aindilis left, lizmat_ joined, lizmat left, aindilis joined, lizmat_ is now known as lizmat 13:21 pyrimidine left 13:22 CIAvash[m] joined
dalek : 0d6d4f4 | (Mathieu Gagnon)++ | util/Text-SmartLinks/lib/Text/SmartLinks.pm:
Change design to use POD title instead of syn_id

At design.perl6.org, pages from p5 Pod have a syn_id for title while p6 Pod have the title section.
The syn_id makes the search results more difficult.
13:22
: 41f11f0 | (Zoffix Znet)++ | util/Text-SmartLinks/lib/Text/SmartLinks.pm:
Merge pull request #14 from 4d47/design-title

Change design to use POD title instead of syn_id
13:30 pierre__ joined 13:31 Actualeyes joined 13:35 seatek left
ufobat m: "/foo" ~~ / ^ "" [$ || <?before "/"> ] /; print $/.perl; say " postmatch: { $/.postmatch }" 13:38
camelia rakudo-moar a8ba26: OUTPUT«Match.new(ast => Any, list => (), hash => Map.new(()), orig => "/foo", to => 0, from => 0) postmatch: ␤»
ufobat i would expect "/foo" as postmatch? am i wrong?
moritz ufobat: no, looks like a bug to me 13:39
ufobat oh okey :/
moritz has a local patch, need to build and test 13:40
ufobat can i see the diff? :-) i am curious 13:41
moritz m: say 'abc' ~~ /^/ and say $/.postmatch
camelia rakudo-moar a8ba26: OUTPUT«「」␤␤»
moritz ufobat: perlpunks.de/paste/show/584c0638.2d8e.1fc 13:42
I wonder if the nqp::if is really necessary 13:43
I don't think we construct match objects for failed matches anymore
also I wonder if $!orig wouldn't be even faster than $!CURSOR.target 13:44
ah, $!CURSOR.target is probably always a string, $!orig not necessarily
m: say 42 ~~ /.+/ && say $/.orig.^name 13:45
camelia rakudo-moar a8ba26: OUTPUT«Int␤True␤»
ufobat not sure what $!orig is- but assuming from and to is for example 2
moritz $!orig is the thing that was matched against
ufobat then only if to == from == 0 its $!orig 13:46
moritz ufobat: fix pushed 13:47
ufobat thank you :-) i am happy 13:48
13:50 pyrimidine joined, wamba left 13:58 pyrimidine left
dalek c: 535d70a | Altai-man++ | doc/Language/regexes.pod6:
Fixes github.com/perl6/doc/issues/1030
13:59
synopsebot6 Link: doc.perl6.org/language/regexes
c: c5d8949 | Altai-man++ | doc/Language/typesystem.pod6:
We don't need external indexing if it is already a header. The whole syntax seems wrong here, which caused 404 entry.
synopsebot6 Link: doc.perl6.org/language/typesystem
14:01 rurban joined 14:05 bjz left, bjz joined 14:08 lizmat left, lizmat joined
lizmat to make things even faster, I'm thinking about creating postcircumfix:<[;]>(\SELF, \one, \two) candidates 14:09
the optimizer should then rewrite the generic code to those cases 14:10
14:10 lizmat left 14:11 lizmat joined
lizmat to make things even faster, I'm thinking about creating postcircumfix:<[;]>(\SELF, \one, \two) candidates 14:11
the optimizer should then rewrite the generic code to those cases
14:14 pyrimidine joined 14:16 lizmat left 14:18 pyrimidine left
ufobat Bailador now supports "prefix" - github.com/ufobat/Bailador/blob/pr...9-prefix.t :-) 14:21
sena_kun ufobat++ 14:22
14:22 Sgeo left, Sgeo_ joined
tailgate m: gist.github.com/ahalbert/748b49d69...3b30bf7396 14:27
camelia rakudo-moar b9e2ff: OUTPUT«False␤Internal error: inconsistent bind result␤ in sub dl at <tmp> line 8␤ in block <unit> at <tmp> line 13␤␤»
14:28 skids joined 14:32 pat_js joined
timotimo tailgate: instead of reducing with "and", you could .all instead 14:36
and i'd recommend using eq instead of ~~ on the LHS of that junction
tailgate thanks 14:38
AlexDaniel what's up with this error message though?
committable6: 2015.07.2 gist.githubusercontent.com/ahalber...onList.pm6
committable6 AlexDaniel, Successfully fetched the code from the provided URL.
AlexDaniel, ¦«2015.07.2»: True␤U␤L␤L␤␤␤R␤R␤D␤D␤D␤␤␤L␤U␤R␤D␤L␤␤␤U␤U␤U␤U␤D␤␤
tailgate yeah, that was the thing. I can't use @, so I was trying to figure out how to iterate over it 14:39
Is subset a "proper" way to make lists that have limited stuff in them?
timotimo "inconsistent bind result" happens when the "where" clause resolves to different values in the simple binder and the slow binder 14:40
m: sub foo(Str where { Bool.pick }) { say "yup" }; foo("hi")
camelia rakudo-moar 2bef4d: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Cannot do non-typename cases of type_constraint yet␤at <tmp>:1␤------> 3sub foo(Str where { Bool.pick }7⏏5) { say "yup" }; foo("hi")␤»
timotimo m: sub foo(Str $foo where { Bool.pick }) { say "yup" }; foo("hi")
camelia rakudo-moar 2bef4d: OUTPUT«Internal error: inconsistent bind result␤ in sub foo at <tmp> line 1␤ in block <unit> at <tmp> line 1␤␤»
timotimo m: sub foo(Str $foo where { Bool.pick }) { say "yup" }; foo("hi")
camelia rakudo-moar 2bef4d: OUTPUT«Internal error: inconsistent bind result␤ in sub foo at <tmp> line 1␤ in block <unit> at <tmp> line 1␤␤»
timotimo m: sub foo(Str $foo where { Bool.pick }) { say "yup" }; foo("hi")
camelia rakudo-moar 2bef4d: OUTPUT«Internal error: inconsistent bind result␤ in sub foo at <tmp> line 1␤ in block <unit> at <tmp> line 1␤␤»
timotimo m: sub foo(Str $foo where { Bool.pick }) { say "yup" }; foo("hi")
camelia rakudo-moar 2bef4d: OUTPUT«Internal error: inconsistent bind result␤ in sub foo at <tmp> line 1␤ in block <unit> at <tmp> line 1␤␤»
timotimo m: sub foo(Str $foo where { Bool.pick }) { say "yup" }; foo("hi")
camelia rakudo-moar 2bef4d: OUTPUT«yup␤»
AlexDaniel haha
alright
well, that's LTA still… 14:41
timotimo the fast binder only generates a "fail/success"
the slow binder then does it more carefully so it can actually figure out *how* things failed
if we wanted to construct a more detailed account of how the fast and slow binder disagree, we'd have a slow and another slow binder instead
tailgate so, the slow one is my function, and the simple one checks if it's a List? 14:42
timotimo no 14:44
14:44 mawkish left
timotimo the fast binder is a piece of code that gets compiled into the head of your function 14:44
so before the "for $d.List { ... }" bit there's actually a bunch of code to make sure the passed arguments are correct
in this case it's probably just an "istype" that will then invoke your where clause for you 14:45
tbrowder tell [Coke] I'll be off the normal land-locked internet grid for Advent editing for a few hours while traveling. I just made some minor changes to the ending words and format. I will be monitoring IRC as well as my iPhone and cell carrier can handle it.
andrzejku hey guys
timotimo when that returns False (like the smartmatch in the line above it did), it'll say "binding failed. please generate a proper error message for me, dear slow binder"
andrzejku are sweden here?
tbrowder .tell [Coke] I'll be off the normal land-locked internet grid for Advent editing for a few hours while traveling. I just made some minor changes to the ending words and format. I will be monitoring IRC as well as my iPhone and cell carrier can handle it.
yoleaux tbrowder: I'll pass your message to [Coke].
andrzejku swedish*
tailgate oh, the smartmatch failed 14:47
hmm
timotimo yeah, it said "False" before the error 14:48
tailgate weird
timotimo yup, it's weird
oh, well, that's easy 14:49
ah, never mind, i'm dumb
m: my Str $test = qq{ULL\nRRDDD\n}; say $test.comb.map({ $_ ~~ Str and $_ eq any(<U D L R \n>) })
camelia rakudo-moar 2bef4d: OUTPUT«(any(True, False, False, False, False) any(False, False, True, False, False) any(False, False, True, False, False) any(False, False, False, False, False) any(False, False, False, True, False) any(False, False, False, True, False) any(False, True, False, Fa…»
timotimo m: my Str $test = qq{ULL\nRRDDD\n}; say $test.comb.map({ $_ ~~ Str and $_ eq any(<U D L R \n>) }).all 14:50
camelia rakudo-moar 2bef4d: OUTPUT«all(any(True, False, False, False, False), any(False, False, True, False, False), any(False, False, True, False, False), any(False, False, False, False, False), any(False, False, False, True, False), any(False, False, False, True, False), any(False, True, …»
timotimo m: my Str $test = qq{ULL\nRRDDD\n}; say $test.comb.map({ $_ ~~ Str and $_ eq any(<U D L R \n>) }).all.so
camelia rakudo-moar 2bef4d: OUTPUT«False␤»
timotimo i didn't do \n right
m: my Str $test = qq{ULL\nRRDDD\n}; say $test.comb.map({ $_ ~~ Str and $_ eq any('U', 'D', 'L', 'R', "\n") }).all.so
camelia rakudo-moar 2bef4d: OUTPUT«True␤»
tailgate yeah "\n" and '\n' tripped me up
14:50 pierre__ left
timotimo yup, but it's right in your code 14:51
tailgate lol, that's not something I hear often
timotimo ;) 14:52
m: my Str $test = qq{ULL\nRRDDD\n}; say $test.comb.map({ $_ ~~ Str and $_ eq any('U', 'D', 'L', 'R', "\n") }).reduce({ $^a and $^b })
camelia rakudo-moar 2bef4d: OUTPUT«any(False, False, False, False, True)␤»
timotimo m: my Str $test = qq{ULL\nRRDDD\n}; say $test.comb.map({ $_ ~~ Str and $_ eq any('U', 'D', 'L', 'R', "\n") }) 14:53
camelia rakudo-moar 2bef4d: OUTPUT«(any(True, False, False, False, False) any(False, False, True, False, False) any(False, False, True, False, False) any(False, False, False, False, True) any(False, False, False, True, False) any(False, False, False, True, False) any(False, True, False, Fal…»
timotimo counter-intuitive result of that reduce operation
but it would give True even with reduce
14:53 bjz left
timotimo m: my Str $test = qq{ULL\nRRDDD\n}; say $test.comb.map({ $_ ~~ Str and $_ eq any('U', 'D', 'L', 'R', "\n") }) ==> reduce({ $^a and $^b }) 14:54
camelia rakudo-moar 2bef4d: OUTPUT«(any(True, False, False, False, False) any(False, False, True, False, False) any(False, False, True, False, False) any(False, False, False, False, True) any(False, False, False, True, False) any(False, False, False, True, False) any(False, True, False, Fal…»
timotimo m: my Str $test = qq{ULL\nRRDDD\n}; say ($test.comb.map({ $_ ~~ Str and $_ eq any('U', 'D', 'L', 'R', "\n") }) ==> reduce({ $^a and $^b }))
camelia rakudo-moar 2bef4d: OUTPUT«any(False, False, False, False, True)␤»
timotimo m: my Str $test = qq{ULL\nRRDDD\n}; say ($test.comb.map({ $_ ~~ Str and $_ ~~ any('U', 'D', 'L', 'R', "\n") }) ==> reduce({ $^a and $^b }))
camelia rakudo-moar 2bef4d: OUTPUT«True␤»
14:54 bjz joined
timotimo i'm not sure why your code is b0rken 14:54
14:59 skids left
tailgate hmm. At one point the smartmatch returned true 15:00
15:00 pyrimidine joined
tailgate or I'm a crazy person 15:00
15:01 pat_js left
tailgate but hey, I've learned more about compilers here than in the class I took =) 15:03
15:04 pyrimidine left 15:07 Sgeo joined 15:08 Sgeo_ left
dalek c/return-ways-explanation: 3cb4189 | Altai-man++ | doc/Type/Signature.pod6:
Attempt to explain ways to declare return type in signature
15:09
c/return-ways-explanation: bd83d34 | Altai-man++ | doc/Type/Signature.pod6:
fill-region-as-paragraph
15:14
ufobat did substr() or postmatch() return a List? a while ago? 15:15
travis-ci.org/ufobat/Bailador#L2574
babydrop ufobat: just use bisectable 15:16
bisectable6: help
bisectable6 babydrop, Like this: bisectable6: old=2015.12 new=HEAD exit 1 if (^∞).grep({ last })[5] // 0 == 4 # RT128181
synopsebot6 Link: rt.perl.org/rt3//Public/Bug/Displa...?id=128181
timotimo or commitable
15:17 pierre__ joined, pierre__ left 15:18 Elena_Morova joined
Elena_Morova Helloo:3 im from Russia) does anyone want to chat with me? 15:18
timotimo hey, how's the weather in russia?
babydrop heh 15:19
AlexDaniel Elena_Morova: so what's your impression of Perl 6 so far? ;)
babydrop smells a robot.
AlexDaniel babydrop: I don't mind. We like robots here :)
15:22 Elena_Morova left
AlexDaniel well… 15:23
sena_kun any LHF from rt-queue for a beginner to try? 15:24
tailgate the important question: is it a perl6 robot
ufobat does bisectable6 use the exitcode? 15:25
15:25 RabidGravy joined
AlexDaniel ufobat: if it differs, yes 15:25
RabidGravy there that all passed unremarkably
AlexDaniel ufobat: otherwise it will compare output 15:26
ufobat bisectable6: old=2015.12 new=HEAD "/foo" ~~ /f/; exit 1 if $/.postmatch ~~ Str
bisectable6 ufobat, Bisecting by exit code (old=2015.12 new=2bef4d3). Old exit code: 1
ufobat, bisect log: gist.github.com/9bfea2f593a678bacc...8560279c5d
ufobat, (2016-12-10) github.com/rakudo/rakudo/commit/2b...ae7d1f55cf
ufobat thanks :)
timotimo it created a list? o_O 15:27
that's pretty weird 15:28
AlexDaniel committable6: 2bef4d30a^,2bef4d30a "/foo" ~~ /f/; dd $/.postmatch
committable6 AlexDaniel, ¦«2bef4d30a^»: "oo"␤¦«2bef4d3»: ("oo",)
AlexDaniel timotimo: no, it creates a list now
15:28 mst_ left, mst_ joined
MasterDuke sena_kun: perl6.fail/t/LTA 15:29
15:30 travis-ci joined
AlexDaniel moritz: ↑ ping 15:30
travis-ci Doc build passed. Altai-man 'Attempt to explain ways to declare return type in signature'
travis-ci.org/perl6/doc/builds/182851925 github.com/perl6/doc/commit/3cb41891a2f1
15:30 travis-ci left
timotimo oh! 15:30
it has the , at the end of the line
that's hilarious :D
AlexDaniel ah f
MasterDuke sena_kun: the LTA tickets are usually a bit easier
AlexDaniel can somebody fix it real quick? :-/
timotimo sure
sena_kun MasterDuke, thanks, I'll try to dig in. 15:31
AlexDaniel ufobat: thanks for reporting
15:31 rurban left
AlexDaniel sena_kun: maybe even try writing some tests? github.com/perl6/roast/issues 15:35
15:35 bjz left, mawkish joined, travis-ci joined
travis-ci Doc build passed. Altai-man 'fill-region-as-paragraph' 15:35
travis-ci.org/perl6/doc/builds/182852791 github.com/perl6/doc/compare/3cb41...83d346abd0
15:35 travis-ci left
sena_kun AlexDaniel, if I will be unable to find appropriate ticket, I suppose. 15:37
15:38 mst_ is now known as mst 15:40 domidumont joined, BenGoldberg joined 15:44 Khisanth left
jnthn sena_kun: rt.perl.org/Public/Bug/Display.html?id=130211 and rt.perl.org/Public/Bug/Display.html?id=130182 may be fairly approachable 15:44
sena_kun jnthn, thanks! 15:46
moritz AlexDaniel: pong 15:51
AlexDaniel moritz: timotimo already fixed it, nvm 15:52
moritz timotimo++
AlexDaniel moritz: but you can still write some tests… :) 15:53
moritz AlexDaniel: i've already done that; just waiting to push them until I'm able to run them 15:54
15:54 pmurias joined
AlexDaniel moritz: oh, then don't forget to close this: github.com/perl6/roast/issues/191 15:55
masak hm, nwc10 is not around, but I thought of him when I found www.naftaliharris.com/blog/why-mak...ython-2.8/ just now. 15:56
babydrop idly notes that were the two tests above an `is-deeply` they would've detected the regression github.com/perl6/roast/commit/99a9...86b00da63e 15:57
15:57 Khisanth joined 16:00 raiph joined
sena_kun I wrote a test for rt.perl.org/Public/Bug/Display.html?id=125247 - it is about unfinished multi-line comment error message. What is the best place to put it? I'm thinking about "S02-one-pass-parsing/comment.t", but such filename looks a bit too wide. 16:01
psch sena_kun: shouldn't that go in S02-lexical-conventions/comments.t with all the other comments tests? 16:04
babydrop sena_kun: just stick it into integration/error-reporting.t
sena_kun: in a is_run() test
sena_kun psch, babydrop, which way? :)
psch hm, right, error-reporting seems to fit better 'cause it's about an incomplete comment 16:05
and well, you're testing that it points at the right spot
babydrop sena_kun: github.com/perl6/roast/blob/master...il.pm#L237
sena_kun: what do you mean way?
sena_kun babydrop, you and psch noted two different files to put the test into. 16:06
babydrop sena_kun: I suggested mine because the bug is about error reporting. 16:07
sena_kun error-reporting.t seems good, thanks!
16:08 RabidGravy left
Ulti if I wanted is it possible for me to make my own class thats like Grammar which has its own renamed methods like rule/token ? 16:12
where abouts is the implementation in Rakudo, I took a look and couldn't pin it down
is it in NQP instead 16:13
babydrop Yeah, it's in NQP.
timotimo "renamed methods", do you mean you want "rule" to be named "sponge" instead?
babydrop Well, NQP has `grammar`
m: class Foo is Cursor { token TOP {} }.parse 16:14
camelia rakudo-moar bfe4ab: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Null regex not allowed␤at <tmp>:1␤------> 3class Foo is Cursor { token TOP {7⏏5} }.parse␤»
babydrop m: class Foo { token TOP {} }.parse
camelia rakudo-moar bfe4ab: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Null regex not allowed␤at <tmp>:1␤------> 3class Foo { token TOP {7⏏5} }.parse␤»
Ulti no I mean I want Frinfron controllers to have get/put/post/delete
babydrop neat
psch well, you'd need a HOW for the class, and DECLARE for the class and the new Routine keyword
timotimo this is about EXPORTHOW, right?
Ulti yeah I assumed it would come down to a HOW but couldnt find one for grammars that looked like it was doing it
psch m: grammar G { }; G.HOW.^name.say
camelia rakudo-moar bfe4ab: OUTPUT«Perl6::Metamodel::GrammarHOW␤»
psch that's just a ClassHOW that does DefaultParent 16:15
timotimo well, rule, token, regex are just funny methods
psch m: grammar G { }; G.HOW.^mro.say; G.HOW.^roles.say
camelia rakudo-moar bfe4ab: OUTPUT«((GrammarHOW) (ClassHOW) (Any) (Mu))␤Required named parameter 'local' not passed␤ in any roles at gen/moar/stage2/nqpmo.nqp line 1195␤ in block <unit> at <tmp> line 1␤␤»
psch m: grammar G { }; G.HOW.^mro.say; G.HOW.^roles(:local).say
camelia rakudo-moar bfe4ab: OUTPUT«((GrammarHOW) (ClassHOW) (Any) (Mu))␤No such method 'gist' for invocant of type 'Perl6::Metamodel::DefaultParent'␤ in block <unit> at <tmp> line 1␤␤»
Ulti such reflection 16:16
psch m: grammar G { }; G.HOW.^mro.say; G.HOW.^roles(:local).map({ $_.^name }).say
camelia rakudo-moar bfe4ab: OUTPUT«((GrammarHOW) (ClassHOW) (Any) (Mu))␤(Perl6::Metamodel::DefaultParent)␤»
psch there :P
raiph .ask TimToady `$` looks like an overlaying of `S` and `I`, a *S*ingle container containing an *I*tem. Did you invent that as a backromnemonic?
yoleaux raiph: I'll pass your message to TimToady.
16:18 RabidGravy joined
masak raiph: I think the much easier explanation is that when `$` was chosen, it meant "scalar". 16:19
and in a way, sh and others had already made the choice of symbol.
TimToady there's always more than one reason for everything :) 16:21
yoleaux 16:16Z <raiph> TimToady: `$` looks like an overlaying of `S` and `I`, a *S*ingle container containing an *I*tem. Did you invent that as a backromnemonic?
BenGoldberg I remember reading that the original version of $ was an overlaid 'U' and 'S', for United States currency, and the 'U' part was changed to a vertical line because people are lazy. 16:23
AlexDaniel is there any module that allows writing € or £ instead of $? :P 16:24
sena_kun I cannot install packages/Test/Util.pm with zef and I cannot pass this path as an argument. How can I solve "Could not find Test::Util at line 7" problem to run roast tests? 16:30
lucasb Test::Util is inside roast repo, so I guess it's the way your are calling 16:31
*you are
16:32 pyrimidine joined
babydrop sena_kun: how are you running the tests? 16:33
16:33 BenGoldberg left
babydrop sena_kun: make t/spec/integration/error-reporting.t 16:34
from rakudo's repo, with the roast checked out in t/spec
sena_kun babydrop, "perl6 integration/error-reporting.t", pwd is "/some/path/roast". Oh, wow. Thanks.
16:34 khw joined
babydrop sena_kun: the `Test::Util` is included with `use lib <t/spec/packages>`. You could technically run it by adding `-Ipackages` but that particular file has fudge in it, so you'd get failures if you run it without the fudger 16:35
(and make does do the fudge)
sena_kun I tried roast repo firstly...
16:36 pyrimidine left
sena_kun afk& 16:38
16:38 sena_kun left
raiph TimToady: Thanks. :) But I'm left None the Wiser who's really no Wiser than Socrates. My draft reply to the person who asked this question already began with "Larry likes to emphasize that there's more than one reason for everything" but I hoped you might say "partly" or "never heard that one before" or somesuch... 16:40
16:45 sena_kun joined, M_o_C joined 16:47 mohae joined
TimToady if I said that on #perl6, you can probably search the logs for it; I don't happen to have any neurons dedicated to remembering that factoid 16:52
it does sound vaguely familiar... 16:53
The big problem with mnemonics is that they're hard to remember sometimes... :) 16:54
araujo solved with meta-mnemonics 16:56
:)
16:58 pyrimidine joined
timotimo metamnomics 16:59
17:00 pyrimidine left, pyrimidi_ joined
TimToady Mna-mna! 17:00
mst TimToady: do dooo de do do 17:02
timotimo anamnomics?
Xliff do do do, do do do, do do do do do DO DO DO DO DO!
Oh, I loved that. My nostalgia bone just got tickled. 17:03
timotimo du-du-du-du-du, du-du du-du-du-du-du, du-du du-du-du-du-du ...
Xliff www.youtube.com/watch?v=9ytei6bu7k...l=Rschris6 17:05
Still getting this from modules in the ecosystem - "Cannot invoke this object (REPR: Null; VMNull)" 17:06
17:07 pyrimidi_ left
timotimo ugh 17:08
17:09 wamba joined
raiph thanks for the memories #perl6 :) 17:10
dalek c/return-ways-explanation: 8a4a26a | Altai-man++ | doc/Type/Signature.pod6:
Fix grammar
17:14
pmurias sena_kun: did you get your test to run?
sena_kun pmurias, yes, thanks. Now I'm preparing a PR.
babydrop \o/ 17:15
17:18 pierre__ joined 17:19 rurban joined 17:22 rurban1 joined, pierre__ left 17:26 rurban left 17:27 BenGoldberg joined 17:36 andrzejku left 17:39 ufobat left 17:40 pyrimidine joined 17:45 pyrimidine left
babydrop sena_kun: did you run your test to see if it passes :} 17:47
sena_kun babydrop, sure. If it didn't pass on travis... Shame on me then!
babydrop sena_kun: I don't think we run that anywhere on travis. 17:48
sena_kun Ah, right.
babydrop But I can see why it'll fail :}
sena_kun: how did you run it? Does make t/spec/integration/error-reporting.t actually pass?
sena_kun: because the plan needs to be bumped: github.com/Altai-man/roast/blob/3a...rting.t#L5 17:49
17:49 Druui joined
babydrop sena_kun: also, would you change the test to look for just 'line 1'. This way the test won't fail if we change the word of the error in the future (like "begins on line 1"... "start at line 1", etc) 17:50
*the wording of the error
sena_kun babydrop, ah, right! Thanks for notice. My memory is somewhat fragile this evening. Yes, I'll correct the test. 17:51
babydrop sena_kun++ thanks
sena_kun I always write my tests with `done-testing` and this time it betrayed me! 17:52
17:56 zakharyas joined
pmurias sena_kun: hint run the tests with "prove" 17:56
sena_kun pmurias, will it work in roast repo? I'm currently doing testing with `make spectest` in rakudo repo, but it feels like injustice. 17:58
babydrop note that that alone will fail with fudged tests. `make t/spec/blah/test.t` fudges (would create a file t/spec/blah/test.rakudo.moar) and if more detailed output is needed, I then run 6p is aliased to `prove -e './perl6' -vlr t/spec/blah/test.rakudo.moar`. There's some invocation that combines the two t/harness6 or t/fudgeall something or other. 17:59
pmurias sena_kun: prove takes a -e flag that allows an arbitrary command 18:00
18:00 mr-foobar left, pierre__ joined
sena_kun Ah, I see the error now. So this is what 255 return code means - plan wasn't updated. The phrase "All subtests passed" after it misguided me. 18:01
babydrop, done! 18:03
18:03 rurban1 left 18:05 cibs left
dalek c/return-ways-explanation: 0aaff17 | Altai-man++ | doc/Type/Signature.pod6:
Text tweaks
18:05
18:06 pierre__ left, cibs joined
AlexDaniel sena_kun: othersd 18:07
on't
:)
sena_kun Argh!
emacs is too lively today, lol. 18:08
18:09 zakharyas left 18:11 mr-foobar joined
pmurias should a nqp::fileislink be considered something that can be optimized out? 18:12
dalek c/return-ways-explanation: ce8b05c | Altai-man++ | doc/Type/Signature.pod6:
Fix grammar
18:14
18:22 eliasr joined 18:27 mawkish left
pmurias would someone do a stat system call just for the side effect? 18:27
18:35 andrzejku joined
japhb pmurias: Which side effect? 18:37
pmurias the stat sytem call itself rather than it's result 18:38
japhb pmurias: Only when interfacing with C (or maybe perl5) I would think ... because of the stat buffer caching. 18:40
18:40 mohae left 18:43 M_o_C left, mohae joined 18:44 thundergnat joined
japhb Talk about random trivia that suddenly becomes vitally important in a very specific set of circumstances -- from the stat(2) manpage: "On Linux, lstat() will generally not trigger automounter action, whereas stat() will (but see fstatat(2))." 18:44
babydrop So yesterday psch pointed out the comment in github.com/rakudo/rakudo/blob/nom/...7766-L7772 18:46
but I still don't get what the infix:<,>($/) := blah gymnastics actually accomplish 18:47
This one in particular: github.com/rakudo/rakudo/blob/nom/...7797-L7801
Like, why is the infix:<,>($/) needed instead of just using $/ ? 18:48
thundergnat Howdy #perl6
babydrop \o
thundergnat I finally have a little free time to play around with perl 6 again. 18:49
psch babydrop: did you try removing the infix:<,> call?
babydrop: maybe that sheds some light on the necessity
thundergnat It's winter in the northern hemisphere and in celebration I wrote some perl 6 to generate some flakes. Couldn't mange snowflakes. Settled for n-flakes. \o/
github.com/thundergnat/n-flake
Comments and criticisms welcome... though I would prefer something like "Maybe you should consider using this perl 6 idiom" over "Your mother was a hamster and your father smelt of elderberries!"
;-) 18:50
psch babydrop: i'd expect e.g. '"foo" ~~ s/(.)/{.uc}/' to stop working without the infix:<,> call, at least that's what comment suggests
+the
err, $0.uc probably 18:51
not $_.uc
18:54 pyrimidine joined 18:56 sena_kun left
babydrop m: say S/(.)/{$0.uc}/ with "foo" 18:58
camelia rakudo-moar bfe4ab: OUTPUT«Foo␤»
babydrop That still works, with this change: gist.github.com/zoffixznet/d13c78e...9f95685662
18:59 pyrimidine left 19:00 pyrimidine joined
babydrop m: my $result := Empty; use nqp; nqp::p6store($/, $result); dd $/ 19:00
camelia rakudo-moar bfe4ab: OUTPUT«Slip $/ = slip$()␤»
babydrop m: my $result := Empty; use nqp; nqp::p6store(infix:<,>($/), $result); dd $/
camelia rakudo-moar bfe4ab: OUTPUT«Nil $/ = Nil␤»
babydrop m: my $result := "foo"; use nqp; nqp::p6store(infix:<,>($/), $result); dd $/ 19:01
camelia rakudo-moar bfe4ab: OUTPUT«Str $/ = "foo"␤»
babydrop m: my $result := "foo"; use nqp; nqp::p6store($/, $result); dd $/
camelia rakudo-moar bfe4ab: OUTPUT«Str $/ = "foo"␤»
babydrop damn. This bit is part of my article and is the cause of a bug and I can't explain why we even have that in the first place.
RabidGravy thundergnat, looks good! Ship It! 19:02
babydrop m: my $result := <a b c >; use nqp; nqp::p6store($/, $result); dd $/
camelia rakudo-moar bfe4ab: OUTPUT«List $/ = $("a", "b", "c")␤»
psch babydrop: you still have the infix:<,> call in L7769?
babydrop m: my $result := <a b c >; use nqp; nqp::p6store(infix:<,>($/), $result); dd $/
camelia rakudo-moar bfe4ab: OUTPUT«Str $/ = "a"␤»
babydrop psch: yeah 19:03
psch babydrop: 'cause i can imagine that that's the one that does the closure closing, while the others are cargo-culted
RabidGravy I was wondering whether the algorithm for generating the points has any other usage which would warrant the generation being separated from the rendering in a separate module
psch babydrop: mind, i'm not sure on that, and maybe all of those are superfluous or something; spectest probably would show 19:04
babydrop spectesting now
19:04 pyrimidine left 19:05 Druui left
dalek : b56bcc6 | (JJ Merelo)++ | examples/functional/monads.pl:
Upgraded it to current version

And added some stuff, too.
19:07
: b3a6b8b | moritz++ | examples/functional/monads.pl:
Merge pull request #24 from JJ/master

Upgraded it to current version
babydrop no failures other than not ok 18 - .prematch for zero-width matches and .postmatch for zero-width matches which is 'cause I'm not on HEAD 19:08
And now that I stare at it... it just stores result of applied matches in $/ and Str!APPLY-MATCHES just returns a single string. 19:11
So the second one is definitely either cargoculted or just leftover from Old Days
samcv what do you guys think if I made Atom/Github's perl 6 highlighter part of the perl 6 org. or this maybe would be non-kosher idk. MadCapJake still hasn't replied
but would be cool since all p6 projects *are* on github so it affects how they show up on the web 19:12
moritz samcv: +1
samcv: if there's anything I need to do for that (create a repo, fork one, whatever), just tell me 19:13
samcv okay :)
babydrop moritz: well, I'm currently trying to read code with this glitch in it: i.imgur.com/FcMf3Ry.png
So I'm all for the highlighter being easily hackable by perl6 org :)
moritz babydrop: :-) 19:14
babydrop: I'd disable syntax hilighting in that situation
samcv moritz, gonna transfer my fork to the perl6 repo 19:15
errr org right now
babydrop \o/
samcv ok it's moving 19:16
github.com/perl6/language-perl6fe 19:18
babydrop great. So what's the process of using that instead of Madcapjake's version? 19:20
AlexDaniel instead huh?
babydrop ? 19:21
AlexDaniel I mean, just send a pull request, no?
babydrop AlexDaniel: to what?
AlexDaniel wait, you're moving it to perl6/ … what's going on here‽ 19:22
babydrop AlexDaniel: I'll give visual aid: github.com/MadcapJake/language-per...43/commits
AlexDaniel so MadcapJake is still active but we're forking his thing huh?
babydrop AlexDaniel: I don't see him "being active".
AlexDaniel .seen MadcapJake
yoleaux I saw MadcapJake 2 Dec 2016 06:23Z in #perl6: <MadcapJake> well I'm gonna fiddle with this laptop some more and then hit the hay. Glad to be back! See you this weekend!
AlexDaniel this weekend is like today 19:23
don't you guys have a bit of patience or something? :)
psch the 2th was a friday, so it's clearly ambiguous vOv
AlexDaniel does the word “guys” include all genders?
psch i think that is also ambiguous :P
TimToady it does in California, usually 19:24
well, at least in the phrase "you guys"
19:24 darutoko left
AlexDaniel
.oO( should use .oO() and /me properly too )
19:25
samcv babydrop, i'm going to publish 1.9 to the atom package repository, so whenever madcapjake gets back, he can depublish his one. and i will change the link that github uses for syntax highlighting as well to ours
babydrop samcv: sounds goods
AlexDaniel: after 3 weeks of "soons" promised on twitter and github, there was plenty of patience dispensed.
TimToady
.oO(.oO(*.oO))
19:26
babydrop Actually not soons but "tonights"
Previous tonight was 8 days ago and one before that on Nov 22
AlexDaniel well, if it's under perl6/, then I guess it doesn't matter 19:27
19:27 lizmat joined
babydrop And he promised to hang out commit bits so we could take care of stuff if he ain't got time, so I don't think he'd even mind this 19:27
TimToady: yeah, I know subparse doesn't scan....
AlexDaniel should we keep it as a “fork” then?
or just recreate the repo 19:28
babydrop shrugs
AlexDaniel it's going to show “forked from MadcapJake/language-perl6fe” forever
babydrop As long as I get highlights on GitHub and in Atom I couldn't care less :)
19:28 Tonik left
babydrop But so does a lot of stuff, like github.com/perl6/modules.perl6.org 19:29
AlexDaniel “This branch is 5483 commits ahead, 2 commits behind MadcapJake:master” and stuff
babydrop well, maybe not a lot bug some.
but
AlexDaniel well, seems like this issue can be resolved through GitHub support. So perhaps one day we can ask them to remove fork links for a list of repos with > 100 or so commits ahead? 19:32
anyway, doesn't matter
samcv++, this is great
samcv :) 19:33
19:33 mohae left
samcv do you want me to recreate the repo or no? 19:33
i can do that 19:34
psch well, duplicating is how github suggests to solve this kinda issue
samcv ah
TimToady m: say 2.&([R∘] * + 3, * * 2)
camelia rakudo-moar bfe4ab: OUTPUT«10␤»
TimToady lucasb: ^^ another way
babydrop m: say (my $x = "foo") ~~ s:g/(.)/{$0.uc}/; dd $x
camelia rakudo-moar bfe4ab: OUTPUT«(「f」␤ 0 => 「f」 「o」␤ 0 => 「o」 「o」␤ 0 => 「o」)␤Str $x = "FOO"␤»
psch i mean, sure, it links to the duplicating page when one wants to privatize a public fork
babydrop psch: I get the same result with the other infix:<,>($/) thing removed
psch but it's similarly applicable i'd say
babydrop So I don't get it at all :/
samcv moritz, is there a way I can get full admin powers for this repo?
psch babydrop: that's curious. maybe FROGGS++ remembers :)
AlexDaniel samcv: :O don't you have them already if you created the repo? 19:35
psch AlexDaniel: probably not after transfering ownership to the perl6 group
samcv i transfered it to the org
i can push but that is all
babydrop samcv: done
samcv and i also probably need to setup a release branch too
FROGGS umm, what exactly?
samcv and then have people push to master 19:36
psch FROGGS: rakudo commit 32d7b08334
FROGGS: babydrop is wondering about the infix:<,> calls
samcv and then a release branch for github/atom to source from
babydrop FROGGS: 1 sec, I'll show code
FROGGS I see it 19:37
babydrop FROGGS: what's the point of this infix:<,> thing? github.com/rakudo/rakudo/blob/nom/...7768-L7772
There's a comment, but I don't get it
FROGGS: and originally, that inquiry started with this one: github.com/rakudo/rakudo/blob/nom/...7797-L7801
But so far I came to conclusion that it's just a cargocult 19:38
No idea about the first one tho
FROGGS it is about pulling out the first thing and assigning it to $/
guess it is about the :g modifier
it most likely should prevent $/ becoming a List
you can remove it an spectest to see what it affects 19:39
19:39 mohae joined
babydrop Done that for the second one, no breakage 19:39
psch which might mean we don't have a test that covers what would break
samcv ok done
FROGGS well, maybe it made a different pre-GLR
psch m: say "foo" ~~ s:g/(.)/{say $/.perl}/ 19:40
camelia rakudo-moar bfe4ab: OUTPUT«Match.new(ast => Any, list => (Match.new(ast => Any, list => (), hash => Map.new(()), orig => "foo", to => 1, from => 0),), hash => Map.new(()), orig => "foo", to => 1, from => 0)␤Match.new(ast => Any, list => (Match.new(ast => Any, list => (), hash => M…»
psch m: "foo" ~~ s:g/(.)/{say $/.Str}/
babydrop The second one receives apply-matches result and it it's always a string
camelia rakudo-moar bfe4ab: OUTPUT«f␤o␤o␤Cannot modify an immutable Str␤ in block <unit> at <tmp> line 1␤␤»
babydrop Just removed the first one too and again no failures in spectest
moritz I guess we don't have a test for that because the semantics of $/ and mutiple matches was meant to be different 19:41
babydrop m: s"x" = { 42 } given my $x = "foo"; dd $x
camelia rakudo-moar bfe4ab: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Variable '$x' is not declared␤at <tmp>:1␤------> 3s"x" = { 42 } given my 7⏏5$x = "foo"; dd $x␤»
moritz when we've discussed that in the past, the idea seemed to have been to create a Match object that represents multiple matches, and bind that to $/ 19:42
psch m: say so Match.new
camelia rakudo-moar bfe4ab: OUTPUT«True␤»
psch but that needs different truthiness 19:43
moritz and expose the individual match objects as captures, or something
psch i mean, i had had a patch that nests Match objects somewhen last year
and TimToady decided to make :g return a List of Match objects instead
FROGGS moritz: I'm almost sure that this was not the case when writing the patch 19:44
moritz but is that list also bound to $/ ?
babydrop m: my $x = "foo"; s{.} = -> { 42 } given $x; dd $x
camelia rakudo-moar bfe4ab: OUTPUT«Block object coerced to string (please use .gist or .perl to do that)␤ in block at <tmp> line 1␤Str $x = "oo"␤»
babydrop What's the syntax to use a block with s/// ?
psch m: "foo" ~~ m:g/(.)/; say $/.WHAT
camelia rakudo-moar bfe4ab: OUTPUT«(List)␤»
psch moritz: ^^^
moritz babydrop: s[.] = 'foo'
oh
m: $_ = 'foo'; s[.] = 'bar'; .say
camelia rakudo-moar bfe4ab: OUTPUT«baroo␤»
babydrop Well, I'm just reading the comment above the infix:<,> thing: "# ($/,) = $result - We do this so the replacement closure can close" 19:45
What replacement closure?
"# over the current match."
moritz m: $_ = 'foo'; s[.] = $/.to 19:46
camelia ( no output )
moritz m: $_ = 'foo'; s[.] = $/.to; .say
camelia rakudo-moar bfe4ab: OUTPUT«1oo␤»
psch ...actually that was apparently *two* years ago that i had fiddled with :g
FROGGS m: say $_ = "foo"; $_ ~~ s[(.)] = "<$0>"; .say # babydrop: the "<$0>" is called as a closure 19:47
camelia rakudo-moar bfe4ab: OUTPUT«foo␤<f>oo␤»
FROGGS $0 used to be Any at some point
babydrop FROGGS: hmm... but that still gives me the same result on my version without those two infixes 19:48
FROGGS bisectable6: say $_ = "foo"; $_ ~~ s[(.)] = "<$0>"; .say
bisectable6 FROGGS, On both starting points (old=2015.12 new=bfe4abb) the exit code is 0 and the output is identical as well
FROGGS, Output on both points: foo␤<f>oo
babydrop with these changes: gist.github.com/zoffixznet/b36be38...4a205595e1
committable6: all say $_ = "foo"; $_ ~~ s[(.)] = "<$0>"; .say 19:49
psch 5a06ace5c5c is 'allow $x ~~ s/// to return Match or list of Match' fwiw
committable6 babydrop, gist.github.com/767b22a455149c0eb6...668a9f3708
psch so, yeah, that is probably decided for 6.c at least vOv
FROGGS the heck
psch we're not doing the .subst-mutate call anymore, but...
FROGGS babydrop: sorry, can't shed light on this 19:50
babydrop :(
lucasb TimToady: thanks for the snippet! interesting way of doing it :)
m: say ([Ro] *+3, * * 2, {$^a * $^a}, * + 11, * * 6)(2) 19:53
camelia rakudo-moar bfe4ab: OUTPUT«666␤»
FROGGS committable6: all say $_ = "foo"; $_ ~~ s:g[(.)] = "<$0>"; .say 19:54
TimToady lucasb: now you just need to work "donald trump" in there somewhere
committable6 FROGGS, gist.github.com/ad67999616decb914d...7cceb0590f
babydrop bisectable6: old=2015.07 new=HEAD say $_ = "foo"; $_ ~~ s:g[(.)] = "<$0>"; .say 19:55
bisectable6 babydrop, On both starting points (old=2015.07 new=bfe4abb) the exit code is 0 and the output is identical as well
babydrop, Output on both points: foo␤<f><o><o>
babydrop FROGGS: well, I'm gonna toss them both then... spectest is clean. 19:56
psch committable6: all say $_ = "foo"; $_ ~~ s:g[(.)] = "<$0>"; $/.WHAT.say
committable6 psch, gist.github.com/2b580d92f8cc929b30...0d3b7c9b4f
FROGGS babydrop: yeah, makes sense I guess 19:57
babydrop: look: rt.perl.org/Ticket/Display.html?id=126349 19:59
and these too: RT #125598, RT #123005
synopsebot6 Link: rt.perl.org/rt3//Public/Bug/Displa...?id=125598
Link: rt.perl.org/rt3//Public/Bug/Displa...?id=123005
20:00 pyrimidine joined
FROGGS committable6: all for 1 { my $x = 'a1'; $x ~~ s/(\d+)/<$0>/; say $x } 20:00
lucasb github.com/rakudo/rakudo/commit/32...a25af014f0 20:01
committable6 FROGGS, gist.github.com/b6793aee1c9b3358bb...5340d4a8f8
babydrop m: gist.github.com/zoffixznet/1a5cbd8...f86bfd82af
camelia rakudo-moar bfe4ab: OUTPUT«[11-22-33 44-55-66 77-88-99]␤»
lucasb FROGGS: ^^ not looking for that, no?
babydrop FROGGS: hm, now I finally see the difference.
FROGGS: in my version without infixes, I get [11--22--33 44-55-66 77-88-99] for that one
FROGGS lucasb: that's the patch in question, yes 20:02
20:02 labster joined
FROGGS babydrop: sorry for the bad comments... should have linked the RTs at least 20:03
babydrop No problem :) Glad we found something
tailgate I have a list of 3x3 matricies, and want to transpose all of them 20:04
I know I can use [Z] for one
samcv anyone know how I can create labels in github for issues?
FROGGS (found it via searching for "last changed by" and "closed at" in RT)
tailgate but I'm having trouble writing one that correctly tranposes all of them
TimToady how 'bout [Z]« or some such?
babydrop samcv: click on issues there's "Labels" button that leads to this page: github.com/perl6/language-perl6fe/labels 20:05
FROGGS samcv: is it your repo?
20:05 pyrimidine left
samcv oh hey 20:05
i have never seen this page before :)
yes it is now that i recreated it
i couldn't find it on my own projects either .thanks
tailgate should it be >>[Z]<< ?
or just [Z]<< ? 20:06
TimToady no, you want to [Z] each array separately, so it's not an infix
20:07 Ben_Goldberg joined, BenGoldberg left, Ben_Goldberg is now known as BenGoldberg, TEttinger joined
TimToady hmm, doesn't seem to parse right currently 20:07
tailgate ah, I got it to work with .map({ [Z] @_ }); 20:08
babydrop oh 20:09
I think it was just my terminal glitching, I get the same result for that [11-22-33] thing
babydrop tests other tickets
tailgate nope 20:10
lucasb m: $/ := 42; my $s = 'foo'; $s ~~ s/a/b/ 20:12
camelia rakudo-moar bfe4ab: OUTPUT«Cannot modify an immutable Int␤ in block <unit> at <tmp> line 1␤␤»
lucasb ^^ why can't $/ be rebound without touching the current value? 20:13
psch m: my $x := 42; $x = 10;
camelia rakudo-moar bfe4ab: OUTPUT«Cannot assign to an immutable value␤ in block <unit> at <tmp> line 1␤␤»
psch lucasb: because it's not binding, it's assigning
no, wait, modifying! 20:14
babydrop all those tickets have the same behaviour with my patch, so I'll push it.
psch hm
i think whatever i used to know about the .match internals proobably doesn't apply anymore so yeah :)
babydrop lucasb: s/// sets the current $/
So... you're preventing it from doing so by binding $/ to an Int 20:15
and it's crying
20:15 domidumont left
lucasb makes sense 20:15
it assumes $/ is already bound to a scalar container...
20:19 rurban joined
lucasb subs have a private $/ which blocks don't? 20:23
sub { $/ := 42 }(); 'a' ~~ /a/ # this works
-> { $/ := 42 }(); 'a' ~~ /a/ # this fails 20:24
20:29 BenGoldberg left 20:31 BenGoldberg joined 20:37 Ben_Goldberg joined, BenGoldberg left, Ben_Goldberg is now known as BenGoldberg
BenGoldberg m: -> { $/ := 42 }; say $/ 20:39
camelia rakudo-moar bfe4ab: OUTPUT«Nil␤»
BenGoldberg m: (-> { $/ := 42 })(); say $/
camelia rakudo-moar bfe4ab: OUTPUT«42␤»
BenGoldberg m: (sub { $/ := 42 })(); say $/
camelia rakudo-moar bfe4ab: OUTPUT«Nil␤»
babydrop no need to wrap block/sub in parens 20:43
m: sub { @_.say }("weee")
camelia rakudo-moar bfe4ab: OUTPUT«[weee]␤»
20:44 pyrimidine joined
samcv ok what should i name the new package on atom at least, i don't think they can have the same name as existing ones 20:44
language-perl6-fe maybe with a dash? 20:45
instead of language-perl6fe
babydrop Something that makes it easy to tell it's best one
samcv language-perl6-FUN-EDITION
babydrop Is that what fe stands for?
mst -Ofun is always turned on
samcv yeah
well there's no 'language-perl6' yet 20:46
babydrop that one looks good
samcv it will have a higher version number than the other package
ok
since it's part of perl6 org group that should prolly be fine
20:48 pyrimidine left 20:53 pyrimidine joined
samcv atom.io/packages/search?utf8=%E2%9...mp;q=perl6 20:53
\O/
so published. i'm going to edit a few things now that i have all the links for the package 20:54
babydrop woooo
samcv++ great job
samcv what should i have the repo name on the perl 6 org tho 20:55
maybe atom-language-perl6?
babydrop sure 20:57
20:57 pyrimidine left 20:59 cdg joined 21:02 sena_kun joined 21:06 lizmat left 21:09 eliasr left 21:13 FROGGS left 21:15 aindilis` joined, aindilis` left, aindilis left 21:17 aindilis joined 21:18 thundergnat left
samcv ok now it's all done :) 21:25
babydrop \o/
samcv ok cool. have a releases branch, i set it to protected so me and other admins can push to that 21:26
and i'll have that branch be the one github points to, and anybody in the org can push to master if they wish, or make a PR for review 21:27
babydrop m: my $e = Empty; ($e,) = 42; say $e 21:29
camelia rakudo-moar 772ead: OUTPUT«()␤»
babydrop m: my $e = 42; ($e,) = 42; say $e
camelia rakudo-moar 772ead: OUTPUT«42␤»
babydrop I wonder if that's a bug...
psch m: my $e = 42; ($e,) = 43; say $e 21:30
camelia rakudo-moar 772ead: OUTPUT«43␤»
babydrop m: my $e = "foo"; ($e,) = 42; say $e
camelia rakudo-moar 772ead: OUTPUT«42␤»
psch it seems weird
m: (Empty,) = "foo"
camelia ( no output )
psch well
jnthn m: ($(Empty),) = "foo" # curious 21:31
camelia ( no output )
psch could maybe be fixed with an iscont somewhere in the right spot
although i'm really not sure what Empty should do in an assignment LHS in the first place
jnthn m: my $a; (Empty,$a) = "foo", "bar"; say $a 21:32
camelia rakudo-moar 164eb4: OUTPUT«foo␤»
jnthn Disappear :)
But probably should not do that if it's in a Scalar container
babydrop m: my $e = slip 1, 2, 3; dd ($e,) = 42
camelia rakudo-moar 164eb4: OUTPUT«Cannot modify an immutable Int␤ in block <unit> at <tmp> line 1␤␤»
psch jnthn: do we still want a sink warning for "bar" in your example?
jnthn It's not in sink context there 21:33
It's an argument to list assignment
And no, we deliberately don't worry about too many things on the right
psch m: my $x; my $y; ($x, $y) = 1, 2, 3
camelia ( no output )
psch ah, okay
jnthn So `my ($first, $second) = @stuff;` won't whine
psch right
jnthn Not to mention stuff like blah xx * on the right
But yeah, I suspect scalar container should take precedence over Iterable 21:34
However
m: my %h; %h<a> = []; my ($a, %h<a>) = 1..10; say %h<a>
camelia rakudo-moar 164eb4: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Shape declaration is not yet implemented; please use whitespace if you meant something else␤at <tmp>:1␤------> 3my %h; %h<a> = []; my ($a, %h7⏏5<a>) = 1..10; say %h<a>␤ expecting any of:␤ …»
21:35 BenGoldberg left
jnthn oops 21:35
m: my %h; %h<a> = []; ($a, %h<a>) = 1..10; say %h<a>
camelia rakudo-moar 164eb4: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Variable '$a' is not declared␤at <tmp>:1␤------> 3my %h; %h<a> = []; (7⏏5$a, %h<a>) = 1..10; say %h<a>␤»
jnthn grmbl
m: my %h; %h<a> = []; my $a; ($a, %h<a>) = 1..10; say %h<a>
camelia rakudo-moar 164eb4: OUTPUT«2␤»
jnthn OK, that's odd
psch m: my %h; %h<a> = []; my $a; ($a, @%h<a>) = 1..10; say %h<a>
camelia rakudo-moar 164eb4: OUTPUT«Type List does not support associative indexing.␤ in block <unit> at <tmp> line 1␤␤Actually thrown at:␤ in block <unit> at <tmp> line 1␤␤»
psch m: my %h; %h<a> = []; my $a; ($a, @(%h<a>)) = 1..10; say %h<a>
camelia rakudo-moar 164eb4: OUTPUT«[2 3 4 5 6 7 8 9 10]␤»
jnthn It already *does* let the Scalar container take precednece there
psch hmm
so it's probably just order of checks? 21:36
as in, we check for Empty too early?
jnthn Well that's what I thought
Unless we do check for Empty specifically
psch yeah, that's what i'm thinking
jnthn What I was trying to tease out of it there was further evidence that we check for Iterable and *then* for Scalar
But I instead disproved that it's that simple 21:37
So yeah, Empty treated specially is a good next hypothesis
m: say Empty.^mro
camelia rakudo-moar 164eb4: OUTPUT«((Slip) (List) (Cool) (Any) (Mu))␤»
jnthn Possible that Slip is the thing that's treated ahead of Scalar cont too
Oh, which you tested earlier 21:38
Well, babydrop did
Yeah, I suspect that's it
Probably should check iscont ahead of Slip
21:41 lukiramu joined
babydrop m: $/ = 'foo' 21:43
camelia rakudo-moar 164eb4: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Unsupported use of $/ variable; in Perl 6 please use the filehandle's .nl-in attribute␤at <tmp>:1␤------> 3$/7⏏5 = 'foo'␤»
babydrop This is the first time I realize how nonsensical these errors look for someone who doesn't know Perl 5
psch m: $/ = 5 # the worse part is that it's only Str literals
camelia ( no output )
psch so yeah, it's confusing 21:44
but 'use perl5-warnings' would be weird, and just not having them is kind of weird too
we have a pragma that turns some of them off iirc
21:45 lukaramu left
psch well, the warnings anyway, don't think it turns of exceptions 21:45
m: $/ = Str.new(value => "foo") 21:49
camelia ( no output )
psch vOv
21:49 BenGoldberg joined 21:53 rurban1 joined 21:54 pyrimidine joined 21:55 rurban left, kaare_ joined 21:58 ufobat joined, ufobat left, pyrimidine left 22:05 kaare_ left 22:08 kaare_ joined 22:16 cdg left, RabidGravy left 22:30 bjz joined
babydrop
.oO( it's weird how writing about fixing a bug made want to play Far Cry: Primal)
22:32
Coming next year. Far Cry: Programmer's Edition
22:36 zakharyas joined 22:40 pmurias left
dogbert17 is it worth reporting (RT) things that break when running MoarVM with a smaller than normal nursery? 22:40
jnthn dogbert17: Better file them as MoarVM issues 22:41
dogbert17 jnthn: and how do I do that?
jnthn GitHub
TEttinger now that colombia's FARC is a political party instead of a rebel group, they could probably work a deal with the Far Cry publishers and make a colombia-tourism shooter, FARC Cry
dogbert17 ah, too simple
jnthn If only GCs were :P
TEttinger shoot poachers with a camera 22:42
dogbert17 jnthn: if you're bored, can you glean anything from this gist? So that I can possibly add more 'meat' to the issue: gist.github.com/dogbert17/a73858a3...21585ef345 22:43
22:44 andrzejku left 22:46 pierre_ joined
jnthn dogbert17: I suspect ASAN or Valgrind might be able to add something 22:46
psch huh, what's the last frame pointing at? 22:47
an MVMFrame?
jnthn psch: Last as in #0 or #18? :)
psch err, lowest :)
so #0
jnthn oh
I dunno, but something inside of the malloc lib or whatever it called to abort the program, I guess 22:48
My guess is that it's corruption rather than double-free
timotimo an MVMFrame would never appear in the C stack, fwiw
jnthn Or at least, if it's a double free inside of congblockingqueue then that's scary.
(And I'd expect us to have uncovered/fixed it my now, given it's the basis of all scheduling) 22:49
Maybe try MVM_GC_DEBUG
psch right, the '??' just confused me cause it's somewhat out of place with everything else pointing at some lib somewhere
timotimo yeah, it is strange, indeed 22:50
psch s/confused me/made me curious/
jnthn The ?? really just means, afaik, the debugger couldn't match it up to something it has symbols available for.
22:50 pierre_ left
timotimo are those addresses actually stack pointers? or are they instruction pointers? 22:50
psch hm, but malloc.c is libc isn't it?
what could that call into that doesn't have symbols at that point..?
jnthn psch: I dunno, some kernel abort handler, maybe? 22:51
timotimo well, since it got sigabrt, it's probably abort()
psch oh, right, that probably wouldn't have symbols
timotimo it should :\
we could try "break abort<tab tab>
jnthn timotimo: Given they are not monotonic, surely not stack addresses...
timotimo " in the gdb shell
jnthn: right, that makes sense
psch timotimo: stock kernels have debug symbols? 22:52
timotimo why would you be able to get a breakpoint inside kernel space with a user space debugger?
without elevated rights, i mean
jnthn Whatever it is, I don't think it's significant for this bug :)
Once you're inside malloc_printerr it's pretty clear you're hosed.
timotimo that's true 22:53
psch right
as i said, just wondering
jnthn Yeah, I don't know what it is, but some abort() or equivalent is a good guess.
timotimo could also have something to do with threading weirdness 22:54
22:54 rurban joined
jnthn It's the only address that starts 7f, which suggests it's from a different library than anything else we see there 22:54
timotimo because of how processes decide which thread receives what signals
psch abort() is in stdlib according to man
same as malloc
anyway 22:55
timotimo you could nativecall into malloc_printerr without having threads active
and see if that makes a difference
22:56 rurban1 left
dogbert17 valgrind doesn't want to play ball, i.e. the code 'works' 22:56
jnthn If you really want to know, I guess something under /proc there's a memory map
That can tell you what lib that instruction pointer is into
dogbert17: Did you try it a few times? :)
dogbert17: ASAN imposes less of a slow-down and may show it up better 22:57
psch shudders at "try a few times with valgrind"
timotimo right you have /proc/maps i believe
jnthn is still a /proc noob :)
timotimo don't forget ASLR will give you different pointers each time, most likely
jnthn Sure, I was assuming you're in GDB and have the PID of the still-living thing ;) 22:58
dogbert17: I guess MVM_GC_DEBUG didn't show up anything incriminating? 22:59
timotimo TIL about "man numa" 23:00
dogbert17 jnthn: that I haven't used, oops
23:01 pyrimidine joined
timotimo oh, that's funny 23:01
cat /proc/self/stack
dogbert17 will 'export MVM_GC_DEBUG=1' suffice?
timotimo no, it's a define in moarvm's source code 23:02
dogbert17 in collect.h perhaps
jnthn dogbert17: No, you have to edit src/gc/debug.h
dogbert17 will do
jnthn And re-compile
dogbert17 yup 23:03
23:03 wisti joined 23:04 rurban left 23:05 pyrimidine left
dogbert17 nothing extra is written to stdout before the crash 23:06
would a 'bt full' be more useful to you
timotimo and nothing to stderr either?
dogbert17 nope
timotimo the thing is that a backtrace will likely not tell us what caused the corruption. just at what place malloc stumbled upon it
dogbert17 well, in that case I'll tease valgrind a few more times :) 23:07
timotimo: btw did you figure something out with that frame problem you were looking at the other day? 23:08
timotimo no, i'm afraid not 23:09
i haven't put much more time into it
tuits keep slipping through between my fingers
dogbert17 maybe it's time for you to hold a beer between your fingers instead :) 23:10
jnthn Mmm :) 23:13
jnthn tracked down some Innis & Gunn Rum Finish for the first time in a while :) 23:14
dogbert17: Did you try MVM_GC_DEBUG set to 2, as well as 1?
dogbert17 nope; i'm at 1 but valgrind finally have something to say ... 23:16
jnthn ooh :)
2 can catch a bunch more but in return slows stuff down
Though, to my surprise, not as much as I might have imagined
I guess if the linked list it traverses is kept cache hot it isn't so bad... 23:17
dogbert17 ok, same gist a bit farther down: gist.github.com/dogbert17/a73858a3...21585ef345 23:18
dalek c: b7e4d53 | ronaldxs++ | doc/Language/syntax.pod6:
Examples and reference to further explanation for Unicode in Perl 6 syntax
c: c277111 | ronaldxs++ | doc/Language/syntax.pod6:
Simplification to address reasonable objection to vagueness about other Unicode refs
synopsebot6 Link: doc.perl6.org/language/syntax
c: d14eea5 | (Aleks-Daniel Jakimenko-Aleksejev)++ | doc/Language/syntax.pod6:
Merge pull request #1042 from ronaldxs/master

Examples and ref. to further explanation for Unicode in Perl 6 syntax
synopsebot6 Link: doc.perl6.org/language/syntax
23:18 zakharyas left
jnthn Wow. 23:18
That is incriminating indeed.
psch kind of wants to assume architecture as the culprit 23:20
buuut that's really just 4 vs 8
which, honestly, doesn't quite make sense
...also, i don't even really know C nor moar so yeah, i'll shut up :P
23:21 kurahaupo left
dogbert17 sips on a 15 y o Springbank 23:21
jnthn Learn about memory barriers and GC safepoints while wokring on those two, since at least one of the two will probably be dragged in to explain the problem also...
psch uhm, i have heaps of bring-jvm-to-moar-standard tyvm :) 23:22
+to do
(also of course not the jvm itself but nqp-j but yeah)
jnthn dogbert17: I never tried that one yet...recommendable?
psch: Aye, that's quite a handful already... :) 23:24
dogbert17 jnthn: indeed, it's very good, picked it up 12 years ago while visiting Islay
babydrop oh, for dog's sake... 23:25
Apparently Wordpress turns gist URLs into some sort of widgets that inline that gist ~_~
jnthn has an unopened bottle of Hammer Head that he'll treat himself to a bit of around Christmas :) 23:26
dogbert17 now that I haven't tried
babydrop Why yes, I do won't 177 lines of code inserted right in the middle of my article... I mean, why else would I have a URL to it!
dogbert17 wonders if Singleton si the programmers whisky of choice :) 23:27
23:27 Ben_Goldberg joined, BenGoldberg left, Ben_Goldberg is now known as BenGoldberg
jnthn Surely most programmers won't drink it because it's an anti-pattern? :P 23:28
babydrop :)
babydrop doesn't drink
oh, misread 23:29
Never mind me...
jnthn You typed won't instead of want a few moments ago too. You sure you don't drink? ;-) 23:30
dogbert17 perhaps this Zoffix guy would appreciate it
psch honestly, what is with that anyway
babydrop With what?
psch i mean, what, 8 different nicks in a year
?
jnthn I guess anger at wordpress is probably a stronger drug than drink anyway :P
babydrop Just 8?
psch actually i've scaled up from those i remember, which is i think 3 :P 23:31
jnthn I host 6guts at wordpress. I tend to author stuff by writing Markdown, using gist to turn it into HTML, and pasting it into wordpress
AlexDaniel by the way, I see no meaningful pattern in these nicks
psch i'll just assume it's some hiding-a-shady-past kinda deal
jnthn Amazingly, this *still* screw up whitespace occasionally.
(A `foo` bar sometimes comes out as if I'd writen `foo`bar) 23:32
babydrop heh, that's actually what I'm doing too :) 23:33
AlexDaniel: there's no pattern
jnthn Did you figure out why it screws up whitespace occasionally? :)
babydrop Never noticed that :P
jnthn Oddness.
Wonder what I'm doing that you're not :) 23:34
psch senses a confusion caused by lack of recipients in messages
23:35 sena_kun left, lizmat joined
AlexDaniel babydrop: Why not? What about “malepregnancy” ? 23:35
dogbert17 heh, valgrinding the other suspicious test, i.e. /perl6-ugga t/spec/S17-supply/syntax.t resulted in the terse comment: MoarVM panic: Collectable 0xbccdbd8 in fromspace accessed 23:36
jnthn dogbert17: ohh...that's really useful
dogbert17: That's produced by MVM_GC_DEBUG
dogbert17 aha
jnthn dogbert17: If you run it in GDB and
break MVM_panic
r
dogbert17 will do 23:37
jnthn Then it'll breakpoint MVM_panic and run it again
And then you can get a stack trace
Sometimes they point right at the bug
Thanks for doing the grunt work of uncovering/filing these by the way
dogbert17 np, if it is of any help at all I'm glad 23:38
jnthn It is; it's pretty much the first bunch of steps I take when hunting these things.
dogbert17 gdb strangeness: ok 67 - No react guts crash in case that once spat out two done messages either 23:39
moar: src/6model/sc.h:163: MVM_SC_WB_OBJ: Assertion `!(obj->header.flags & MVM_CF_GEN2_LIVE)' failed.
jnthn That's...also very interesting :)
dogbert17 bt soesn't say much, it did not break in MVM_panic this time 23:40
jnthn Can always hit r to try it again :)
dogbert17 got it hang on 23:41
gist.github.com/dogbert17/5b275221...6ac94433a9 23:42
[Coke] is FQN here supposed to mean "fully qualified name"? (because it doesn 23:44
yoleaux 14:45Z <tbrowder> [Coke]: I'll be off the normal land-locked internet grid for Advent editing for a few hours while traveling. I just made some minor changes to the ending words and format. I will be monitoring IRC as well as my iPhone and cell carrier can handle it.
jnthn dogbert17: Hmmm 23:45
dogbert17: Can you also
frame 1
p MVM_dump_backtrace(tc)
dogbert17 jnthn: gist updated 23:47
jnthn o.O 23:49
Please file that in a MoarVM issue :)
dogbert17 will do 23:50
jnthn I'm way too tired to dig into it now
23:50 kyclark joined, kyclark left
dogbert17 no digging, just beer :) 23:50
I'll write it up tomorrow, eh later today, at the latest 23:51
jnthn :) 23:52
Thanks
It's not so much beer that's the problem as just being plain tired :)
dogbert17 maybe time for some sleep 23:54
jnthn Indeed :)
Was just thinking that
'night
dogbert17 night