»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'p6: say 3;' or /msg camelia p6: ... | irclog: irc.perl6.org or colabti.org/irclogger/irclogger_log/perl6 | UTF-8 is our friend! 🦋
Set by Zoffix on 25 July 2018.
00:28 mowcat left 00:33 MilkmanDan joined 01:04 dmiserak joined 01:21 adu joined 01:23 dmiserak left 01:28 molaf_ left 01:36 zakharyas joined 01:41 molaf_ joined 01:50 kurahaupo left 02:20 hythm joined
Geth doc: uzluisf++ created pull request #2851:
Add minor rewording		 02:21
02:40 MilkmanDan left 02:41 MilkmanDan joined 02:54 cpup left 03:27 epony joined 03:28 epony left 03:29 epony joined 03:44 veesh left 03:46 veesh joined 04:22 hythm left 04:39 curan joined 04:56 Rayler joined
Geth doc: c193bada57 | (Luis F. Uceta)++ | doc/Language/modules.pod6
Add minor rewording
 05:20
doc: a98de4714c | (Juan Julián Merelo Guervós)++ (committed using GitHub Web editor) | doc/Language/modules.pod6
Merge pull request #2851 from uzluisf/master

Add minor rewording
synopsebot Link: doc.perl6.org/language/modules
05:24 jmerelo joined 05:25 sauvin joined
jmerelo I think I'm gonna publish the Perl 6 user survey today if no one is against it 05:25
In fact, here you go: docs.google.com/forms/d/e/1FAIpQLS...g/viewform 05:26
05:26 cpup joined 05:33 Rayler left
jmerelo releasable6: status 05:40
releasable6 jmerelo, Next release in ≈7 days and ≈13 hours. 2 blockers. 221 out of 616 commits logged (⚠ 1 warnings)
jmerelo, Details: gist.github.com/923c0c4b91a095fe82...d6f8f36635
Geth doc: 091171cffb | (JJ Merelo)++ | 2 files
Updates dockerfiles
 05:45
05:49 AlexDaniel left
jmerelo Reminder: last year survey were published in perl6.github.io/p6survey/ 05:52
06:15 jmerelo left 06:25 abraxxa joined, abraxxa left 06:26 abraxxa joined 06:31 abraxxa left, abraxxa joined, cgfbee left 06:48 patrickb joined 07:01 cgfbee joined 07:07 wyclif joined 07:10 mowcat joined 07:27 Sgeo_ joined 07:30 Sgeo left 07:35 epony left
cpan-p6 New module released to CPAN! ProcStats (0.0.1) by 03LEMBARK 07:36
07:39 dakkar joined 07:47 adu left 07:48 rindolf joined 07:49 aborazmeh joined, aborazmeh left, aborazmeh joined 07:51 dolmen joined 07:52 robertle joined 07:57 [particle]1 joined 07:58 [particle] left 08:01 [particle]1 left 08:05 scimon joined 08:10 [particle] joined 08:14 dolmen left 08:15 dolmen joined 08:18 aborazmeh left 08:36 [particle]1 joined 08:37 [particle] left
daxim need the equivalent of perl -MDevel::Peek -E'say Devel::Peek->VERSION' →→→ perl6 -MInline::Perl5 -e'say Inline::Perl5 ...' 08:45
09:00 [particle] joined 09:02 ravenousmoose left, [particle]1 left 09:03 ravenousmoose joined 09:11 patrickb left 09:18 [particle]1 joined 09:19 [particle] left 09:24 sena_kun joined 09:41 Rayler joined
tobs daxim: perl6 -e 'use Devel::Peek:from<Perl5>; say Devel::Peek.VERSION' 09:54
daxim No such method 'VERSION' for invocant of type 'Inline::Perl5' 09:55
tobs huh, it just says the same as your `perl` line over here. And Devel::Peek.^name is Devel::Peek, not Inline::Perl5. I have Inline::Perl5 v0.35 fwiw. 09:58
09:59 [Sno] left
lizmat $ perl6 -e 'use Devel::Peek:from<Perl5>; say Devel::Peek.VERSION' 10:00
1.27
tobs exactly
daxim how do you know that Inline::Perl5 is v0.35? 10:01
lizmat hmmm... good question... there is a 0.39 out BTW 10:02
tobs I `zef info`d it. But there surely exists a way to introspect it... 10:05
daxim docs.perl6.org/language/typesystem...authorship perl6 -MInline::Perl5 -e'say Inline::Perl5.^ver' →→→ (Mu) 10:09
tobs That's the same here. I suppose the Inline::Perl5 class itself isn't versioned, only the module is. 10:10
This is what I googled together (and works over here): 10:11
m: say CompUnit::DependencySpecification.new(:short-name<Inline::Perl5>).&{$*REPO.resolve: $_}.distribution.meta<ver>
evalable6 Nil
tobs I hope that just means it's not installed on evalable :/ 10:12
daxim that says v0.38 for me 10:13
10:14 wyclif left 10:15 Black_Ribbon left
daxim that's too complicated for this perler's taste. mind the motto "easy thing easy" 10:15
10:17 wyclif joined
tobs hmm, I upgraded to v0.38 specifically and the Devel::Peek.VERSION still works. 10:18
10:19 ravenousmoose left
timotimo how does it report the error when you put something you don't have before the :from<Perl5>? 10:20
tobs Beware that I barely know anything about module authoring and such. It might in fact be easier without me knowing about it. (But I have to say, without understanding the whole complex, it looks overengineered to me at times.)
10:22 robertle left
tobs timotimo: in my case, I get a compile-time error from perl5 about not finding the module, wrapped inside a SORRY from perl6. 10:23
(or who were you asking?)
timotimo yeah, i was just wondering if maybe Inline::Perl5 wasn't finding Devel::Peek on daxim's machine for some reason 10:24
Rayler hiya
10:31 robertle joined 10:32 molaf_ left 10:33 zakharyas left 10:35 patrickb joined
tadzik submitted a PerlCon talk o/ perlcon.eu/talk/101 10:44
daxim let me guess the biggest point: auto-accessors for every field by default? 10:47
10:48 dolmen left 10:51 noisegul joined
timotimo have a camel and it it too? 10:52
that's not how you spell eat :) :)
tadzik daxim: close. Mostly about the blur that `has` introduces between the internal data and (constructor's) public interface 10:53
lizmat tadzik: interesting... only 20 mins ?
tadzik lizmat: my experience suggests that I'll be saying "um, any questions?" around a 10 minute mark ;)
it feels like it could be longer, but I remind myself that I don't think I ever actually filled my time slot in practice 10:54
timotimo tadzeno's paradox
daxim perhaps make a slide that contrasts moose "has" with es6/php7/java8 and what happens there due to the lack of "has" 10:55
tadzik timotimo: indeed :) Thanks, fixed
daxim: hm, anything in particular you have in mind?
other than the obnoxious verbosity :P 10:56
lizmat
.oO( where are the times that you could simply use a hash as an object :-)
daxim I mean the blur you were jsut alluding to - it does not happen in the other languages, right?
tadzik daxim: no, I don't think it does. I don't think I've seen this kind of "declarative constructor" anywhere else 10:57
it's understandable why it found its place in Perl – because argument handling sucks here 10:58
Perl 6 improves it a lot, but I feel like in this particular area it doesn't really capitalize on it 10:59
we still declare public attributes to benefit from the free new() and then save our asses with TWEAK and alike
11:02 Rayler left 11:04 [Sno] joined 11:05 curan left 11:07 jmerelo joined 11:08 molaf_ joined 11:10 [particle] joined 11:11 powerbit left, [particle]1 left 11:14 dolmen joined, [particle] left, cpan-p6 left, [particle] joined
holyghost I revived Soundn::Tina in p6-Sound-Tina on github, last time I checked it compiled, this way I can work on something different sometimes 11:15
11:16 cpan-p6 joined, cpan-p6 left, cpan-p6 joined
holyghost I have to debug it but it should eventually parse several (also custom for sound hackers) WAV file formats then classify patterns in the channels/samples 11:16
I need a 32 bit WAV file for it 11:17
11:17 sena_kun left
holyghost then debug 11:17
If I find ioctl in perl6, it might even play the WAV as I go on 11:19
I didn't choose mp3 TBH as it doesn't have a Dolby functionaility AFAIK 11:20
The above WAV files are very cool with such an amplifier :-) 11:21
11:21 robertle left
holyghost I need to check the 32 bit version first though, there's no dolby code in it yet 11:21
11:32 |Sno| joined, [Sno] left
Geth doc: b867248d64 | (JJ Merelo)++ | util/Dockerfile
Minor formatting
 11:33
11:37 ravenousmoose joined 11:38 demanuel left
jmerelo reminds for those who have not done it yet to fill the survey: docs.google.com/forms/d/e/1FAIpQLS...g/viewform 11:44
tobs jmerelo: for how long do you want to keep it open? I think it would be beneficial to wait for a weekly to pass, for instance. 11:45
(I already filled it, anyway)
jmerelo tobs: at least a couple of weeks. 11:46
tobs alright!
jmerelo tobs: I'd like to have at least like 250 answers or so.
tobs: thanks for filling it in.
if you subscribe to any mailing list, please spread the word. 11:47
11:59 noisegul_ joined 12:02 noisegul left 12:03 dolmen_ joined 12:04 noisegul joined, noisegul_ left, dolmen left
noisegul Will fill it later as well 12:04
12:13 dolmen_ left 12:40 dolmen_ joined, wyclif left 12:45 dolmen_ left 13:03 dolmen_ joined 13:05 |Sno| left 13:06 mowcat left, dolmen_ left 13:07 [Sno] joined 13:09 jmerelo left 13:11 dolmen joined 13:12 dolmen left 13:20 dolmen joined 13:21 atweiden-air joined 13:27 dolmen left 13:33 pamplemousse joined 13:34 dolmen joined 13:37 dolmen left 13:38 johnjohn101 joined 13:46 dolmen joined 13:47 wyclif_ joined, dolmen left 13:52 demanuel joined 13:55 dolmen joined, dolmen left 14:07 AlexDaniel joined 14:10 wyclif_ left, wyclif_ joined 14:12 dolmen joined 14:13 atweiden-air left 14:14 dolmen left 14:30 wyclif_ left 14:31 dolmen joined 14:35 dolmen left 14:39 dolmen joined 14:52 dolmen left 14:55 khisanth_ left
cpan-p6 New module released to CPAN! QM (0.0.2) by 03UZLUISF 14:58
15:06 dolmen joined 15:08 khisanth_ joined 15:09 nadim_ joined 15:17 sena_kun joined 15:18 woolfy left 15:25 veesh left 15:26 veesh joined 15:29 jmerelo joined 15:34 patrickb left 15:45 nadim_ left 15:52 woolfy joined
jmerelo 44 answers already to the Perl 6 survey: docs.google.com/forms/d/e/1FAIpQLS...g/viewform Call for participation! 15:55
Geth doc: lukasvalle++ created pull request #2853:
Examples for 6.d checklist		 15:59
sena_kun oh, jmerelo++ 16:01
16:02 mowcat joined
jmerelo sena_kun: if you think it needs something else to change, just let me know. Or maybe we can do another survey before Riga... 16:04
[Sno] jmerelo: the survey ignores one of the most use-cases, perl covered until now: admin glue, system helper (the dirty work, operations, ...) 16:05
16:09 goon_ left 16:10 holyghost left
timotimo aaw, my compress:zstd module has an extra null byte after decompression 16:13
will be fun to find that off-by-one :)
jmerelo [Sno]: There's one person that has added "system integration". Also, "General scripting" is an option, and in fact the most popular. 16:15
[Sno]: so, probably not the killer app for Perl 6 right now...
16:16 nadim_ joined 16:17 molaf_ left
lucs jmerelo: For what it's worth, have at the end "What do you want to tell us?" or something. 16:18
If that question was there, I would tell you exactly that, I mean, uh, never mind 16:19
;)
jmerelo lucs: like a suggestions box or something? 16:23
lucs Right.
jmerelo OK
lucs Just saying.
[Sno] jmerelo: that's what I mean - Perl 6 ignores it (and so that target, what made Perl successful)
jmerelo lucs: done 16:25
lucs jmerelo: Cool, thanks.
16:25 dakkar left
lucs jmerelo: Do you know if it's possible to view the results without submitting anything? (I want to look at them from another machine) 16:26
jmerelo [Sno]: different times, different languages... I would also say that what made Perl succesful were CGIs, but anyway, a target audience for a language is not exactly something you bake into it.
lucs: I can share the answers with you, but doing it with everyone else might bias the rest of the answers... 16:27
[Sno] jmerelo: that's probably as wrong as it can be - when it's a different language, it shouldn't be named Perl 16:28
lucs jmerelo: The idea is that my current browser, from which I submitted answers, is so old that I wonder if the results displayed are weird just because of unhandled CSS or something -- the other machine is more up-to-date.
jmerelo [Sno]: I don't really follow you here. Can you please clarify? 16:29
lucs jmerelo: For example, the "most wanted feature" display is LTA here. 16:31
[Sno] jmerelo: every Perl 6 survey etc. I see, it always tries to cover "high sophisticated" topics
jmerelo: I don't know how to explain it better in chat - it's just "missing" the swiss-army-chainsaw features 16:32
16:32 scimon left
Woodi hi #perl6 :) 16:32
[Sno] hi Woodi 16:33
Woodi imo, moust wanted feature is already solved: [Ss]pee*d
timotimo not solved yet 16:34
lucs timotimo: You don't say ;) 16:35
Woodi but it's more complicated as ~2 years ago? big slow down comes from "corectness" patches and 05 slowdown from timotimo++ bytecode hardening ?
jmerelo [Sno]: how many surveys are there? This is just the third. Anyway, let me tie that with the speed problem. You're probably not going to use it in that quality until it gets a bit faster
timotimo my bytecode hardening made things slower? :( :(
Woodi and corectness is high on priot
Geth doc: c86c661517 | Coke++ | util/update-and-test
Before running tests, insure new deps are installed
Woodi *priority...
[Sno] jmerelo: I mean also any other "language" survey 16:36
jmerelo: speed is in most admin cases not that problem -it's from my point of view fast enough for those tasks 16:37
Woodi timotimo: it is just guess
that's why I put ? there
timotimo phew
[Sno] jmerelo: and please - take it as a feedback where the Perl 6 community is a bit blind at the moment 16:38
jmerelo [Sno]: I don't know then. The point is if that people can add that if they want to the survey. Also, we can't fix perception of the language with a survey.
16:38 dolmen left
[Sno] jmerelo: I'm not sure that you understand what I mean 16:39
jmerelo [Sno]: I'm not sure either. That's why I asked for clarification.
timotimo i think sno is looking at our survey "what do you use the Camelia Mountainbike for?" and is seeing only "cutting onions", "painting pictures", "sending email" and is wondering where "riding up and down mountains" and "riding on streets" went 16:40
[Sno] I try again: Perl 6 can satisfy most of the needs a sysadmin has - but it's not promoted and it leaves that field for Perl 5 and Python 16:41
Get it as a working area which needs to be served
16:41 johnjohn101 left
jmerelo [Sno]: that's what I got. And I told you that that was covered by "general scripting" and that so far only one person has answered something closer to that. Also, the objetive of the survey is not promotion. 16:42
[Sno] Why was Perl 5 with CGI successful? Because the people who work with the computers knew Perl already
Woodi [Sno]: just get *that* thing done :)
16:43 [Sno] left
jmerelo [Sno]: well, they had cut their teeth with cgi-lib.pl... which was the only thing available at the time. 16:43
Woodi jmerelo: the only dynamic web pages admins allowed to be installed :) C CGI was too hard to review probably 16:44
jmerelo But again, this is just a survey. I have no idea about marketing, really. So maybe you are right about focusing on something else. But one can do what one can do. 16:45
timotimo i don't recall there being any phase where community members were asked to review the questions and answers for the survey?
jmerelo timotimo: there kind of was last month...
timotimo i wonder if i missed it or just forgot about it 16:46
Woodi hey, surveys are not science, relax...
jmerelo timotimo: but maybe I should have done it again. sena_kun and me are changing things on the fly; it might hamper statistical accuracy, but, well...
timotimo: if you want us to change somethign just let us know. Or we can give you permission. 16:47
timotimo too late to change stuff :) 16:49
tobs jmerelo: didn't you mention you wanted to do another one before Riga, or did I misunderstand? Maybe we can channel suggestions into that one.
lucs jmerelo: Thanks for doing this. 16:51
jmerelo tobs: please do that. 16:52
so one suggestion (by [Sno]), if I got it correctly, is to add "system administration/application glue" to the "What you utilize perl6 for", correct? 16:53
Woodi jmerelo: maybe 2 of them separate ?
jmerelo Woodi: say again? 16:54
Woodi like eg. - system administration; - application glue 16:55
jmerelo Woodi: OK
Let me do this. I'm going to create a copy of this survey and add directly all suggestions to the copy. That way, they're not lost.
tobs jmerelo++ 16:56
before I even found the survey repo (I'm sure there was one at some point…)
jmerelo tobs: the previous one is at github.com/JJ/p6survey 16:57
I'll post the results and nice charts there too when this one is finished.
tobs ah, so the survey itself is not in an open format and can't be issue'd or pr'd 16:58
Geth doc: fbf05c38bd | Coke++ | xt/words.pws
learn new word
tobs (would maybe have been overkill anyway, but that would have been open all year round)
jmerelo tobs: it's in GDocs. We'll share the results there too.
tobs: but I usually export it to CSV and post it at the repo 16:59
Woodi: done
Woodi jmerelo++
17:01 Black_Ribbon joined
Geth doc: 09b84e62be | lukasvalle++ | doc/Language/operators.pod6
Add example for "R meta operator used with list-associative operators"
 17:13
doc: 988a9df134 | lukasvalle++ | doc/Language/phasers.pod6
Extend examples for "CX::Done can be caught inside CONTROL phaser"
synopsebot Link: doc.perl6.org/language/operators
Link: doc.perl6.org/language/phasers
doc: 8a17d5d3b9 | (Juan Julián Merelo Guervós)++ (committed using GitHub Web editor) | 2 files
Merge pull request #2853 from lukasvalle/master

Examples for 6.d checklist
17:25 kotrcka joined 17:32 MasterDuke left 17:33 pamplemousse_ joined 17:35 pamplemousse left, Rayler joined 17:48 mowcat left
nadim_ evening! 17:59
tobs indeed!
nadim_ bug, I believe, with role from required module, golfed down to death and a something very surprising that makes it work. nopaste.linux-dev.org/?1217418 18:00
shellscript that generated the code and tests it 18:01
18:01 bobby joined 18:05 goon_ joined 18:09 john_parr left
vrurg Do we have any docs on valid variable names? 18:11
tobs vrurg: check "identifiers" 18:12
vrurg m: 'my $v:foo = 42; say $v:foo' 18:14
evalable6 WARNINGS for /tmp/fCDqGgufXy:
Useless use of constant string "my $v:foo = 42; say $v:foo" in sink context (line 1)
vrurg m: 'my $v:foo<42> = 42; say $v:foo<42>'
evalable6 WARNINGS for /tmp/yNzgbJ6GSc:
Useless use of constant string "my $v:foo<42> = 42; say $v:foo<42>" in sink context (line 1)
vrurg Strange...
m: my $v:foo<b> = 42; say $v:foo<b> 18:15
evalable6 42
vrurg The doc is incomplete then. :)
18:15 kybr left
vrurg m: my $v:foo<42> = 42; say $v:foo<42> 18:15
evalable6 42
tobs that falls under extended identifier, no? 18:16
vrurg Looks like... Need to refresh that one in mind. 18:17
vrurg spent too much in nqp/mop for the last two month.
18:18 lucasb_ joined
jmerelo vrurg: if you don't add the <> it's an adverb; it can't be used in that context. 18:23
vrurg: the doc is always incomplete... but tobs is right, that's extended identifier syntax
vrurg m: my $v:foo = 42; say $foo 18:24
evalable6 (exit code 1) 04===SORRY!04=== Error while compiling /tmp/f_NHgdJWOj
Variable '$foo' is not declared
at /tmp/f_NHgdJWOj:1
------> 03my $v:foo = 42; say 08⏏04$foo
tobs might be a good opportunity to link R#2948
synopsebot R#2948 [open]: github.com/rakudo/rakudo/issues/2948 Extended identifiers without keys raise "You can't adverb" error
vrurg m: my $v:foo = 42; say $v:foo
evalable6 42
vrurg jmerelo: you mean this form?
vrurg already got his mistake.
jmerelo vrurg: hum. Yes. I thougt that wouldn't work. 18:25
vrurg jmerelo: moreover, its documented. :)
tobs as per docs, it should work. You can omit key or value in the extended part.
jmerelo vrurg: or rather, if it wouldn't have worked, I would have had an explanation. It does work, well, it makes sense.
vrurg: :-) 18:26
18:26 Rayler left
vrurg This is how experts act. They can explain any outcome of a flipping coin, but never predict... ;) 18:26
18:26 sauvin left
jmerelo vrurg: well, experts with a bad memory do at least... 18:26
vrurg jmerelo: now you're talking about me... 18:27
tobs nadim_: well, I can reproduce that and don't understand it. Maybe file an issue for rakudo lest it gets swept away in here. 18:29
nadim_ I don't understand it either, the "solution" is still making me laugh. Computing is fun! 18:31
vrurg nadim_: I don't know how the regex influences it, but the problem is that required_role is being exported as a class (i.e. it has ClassHOW meta). 18:38
nadim_ OK, how does one get it exported as a role? 18:41
vrurg nadim_: Exactly as you've do it. It is a bug. 18:42
I can promise you nothing, but if you file a ticket I may try looking into it. I have squashed two import bugs already, would be the third one. ;) 18:44
sena_kun vrurg++ # great work on bug fixing 18:45
timotimo ci.appveyor.com/project/timo/gtk-s.../artifacts
get your gtk3, libxml2, zstd, libxml2 dlls for windows 64 here 18:46
nadim_ vrurg: I am filling it already
vrurg sena_kun: just dequeing collected PRs since Feb... 18:47
It seems to be a regression. 2018.12 doesn't have the problem. 18:50
18:50 kotrcka left 18:51 kybr joined
vrurg wish to have bisectable locally... 18:51
timotimo you can have it! 18:52
it'll download whateverable builds on demand, though, so i twon't be as fast
vrurg timotimo: It's fine as long as I can have it in background. 18:53
nadim_ bug report github.com/rakudo/rakudo/issues/2983
18:56 kaare_ left, kaare_ joined
nadim_ in any case, this definitively proves that it's wrong to say "if you have 99 propblems and a regexp then you have 100 problems" 19:00
vrurg timotimo: whateverable/xbin/Bisectable.p6?
timotimo i'm not exactly sure
but i've been sent a link in the past, a few times even
gist.github.com/AlexDaniel/88b38af...3df31dd431 - is this the same code?
vrurg timotimo: no, it's different. Thatnks for the link! 19:02
timotimo cool 19:03
19:06 jmerelo left
tobs Suppose I have a Match object without positional captures. I really want to add some to the same Match object. Does anyone have an idea how to do that, even if the answer starts with "use nqp"? 19:06
timotimo you can't .clone and use the named arguments for that purpose? 19:07
tobs I don't think so. The Match is part of a Grammar.parse, so if I make a different object, I will have a hard time fiddling the new one into the Match tree. 19:08
timotimo not necessarily; the match you're handling is usually just an instance of your very own grammar 19:09
tobs I don't understand. The match will be from a subrule and it will be referenced as a (named) capture in the calling rule. So replacing it will involve changing the parent, I think. 19:15
To be specific, I'm trying to rip out the EXPR and related rules from HLL:Grammar into a module. And after a few days' pondering, I think this stunt with just adding positional captures to the operator matches is the most economic solution: github.com/perl6/nqp/blob/master/s...#L542-L544 19:16
vrurg timotimo: any idea how to get bisect work quickly? Complains on 'no build for <release>' 19:17
tobs oh, but I'm wide open to suggestions there as well
timotimo i don't :(
vrurg Hoplessly: ping AlexDaniel? 19:18
lizmat weekly: ajs.github.io/tools/2019/06/12/perl...ences.html
notable6 lizmat, Noted!
tobs vrurg: URLs like this work: whateverable.6lang.org/6565810f5c3...f340732345 19:24
it gives you a .tar.zst of the named commit. But it's probably not nice/viable to download all of them over HTTP.
vrurg tobs: you mean that in line run-snippet(ensure-build(‘2018.10’), $code-file); where it actually fails I can use a commit URL? 19:26
tobs vrurg: ah, no, I don't think so. But bisectable collects its local cache of prebuilt rakudos from the above URL and probably unpacks them somewhere to use them. I only glanced at the code. 19:31
vrurg I thought it autofetches what is needed. Even if I download it – where it's to be unpacked then? 19:33
tobs vrurg: I'm not further into the sources as you are probably :) The paths seem to be controlled by a config file which is also mentioned in that bisect.p6 script and fetching and storing builds is in Whateverable::Builds 19:49
(sorry, was distracted)
AlexDaniel ?
yoleaux 12 Jun 2019 13:13Z <lizmat> AlexDaniel: I think I've documented all of my commits in the ChangeLog wiki
AlexDaniel vrurg: hello? :) 19:50
you're trying to do what exactly? Bisect locally? 19:51
vrurg AlexDaniel: I'm trying to get bisect.p6 working locally to locate a regression. It complains on `no build`
Sorry, trying to handle three tasks in parallel. ;) 19:52
AlexDaniel vrurg: this, right? gist.github.com/AlexDaniel/88b38af...3df31dd431
vrurg: did you put config-default.json into the same directory?
vrurg AlexDaniel: yes, I did. The token is in place. 19:53
AlexDaniel you don't need the token, I think
vrurg It clones into data/moarvm and then fails.
AlexDaniel vrurg: can I see the gist of what you see? 19:54
19:54 kaare_ left
AlexDaniel vrurg: are zstd and lrzip installed on your system? 19:54
19:54 kaare_ joined
vrurg gist.github.com/vrurg/3ea8177056e2...16e25a97f3 19:54
Both are available. Though if shared libs are needed – its macOS. I would probably need to symlink them. 19:55
Symlinking turns to be useless. 19:58
19:59 sivoais joined 20:01 Geth left, Geth joined
AlexDaniel vrurg: ah, it's macOS 20:01
I don't think you can use rakudo compiled for linux on macOS, right? 20:02
vrurg Kinda... ;) As simple as that.. No problem, a Mint VM is for my service anytime. Gonna give it a try. 20:03
Thanks!
20:09 pamplemousse_ left 20:10 |oLa| left 20:17 powerbit joined 20:24 AlexDaniel left, antoniogamiz joined 20:31 AlexDaniel joined, AlexDaniel left, AlexDaniel joined
AlexDaniel .tell jmerelo sounds like a fun squashathon ajs.github.io/tools/2019/06/12/perl...ences.html 20:31
yoleaux AlexDaniel: I'll pass your message to jmerelo.
20:43 AlexDaniel left 20:45 AlexDaniel joined 20:46 noisegul left
xinming_ !status 20:47
releasable6: status
timotimo status: slightly drunk, happy that code he's been working on finally works
releasable6 xinming_, Next release in ≈6 days and ≈22 hours. 2 blockers. 221 out of 616 commits logged (⚠ 1 warnings)
xinming_, Details: gist.github.com/8e1ce920a1985b0774...0097c63388
20:48 AlexDaniel left 20:49 AlexDaniel joined 20:55 farcas82regreg left 21:15 antoniogamiz left 21:34 molaf joined 21:41 sena_kun left 21:43 pamplemousse_ joined
AlexDaniel vrurg: any reason you can't throw it into the bot? 21:45
21:53 dolmen joined, Thron joined 21:54 Thron left 21:58 pamplemousse_ left 21:59 Xliff joined
Xliff \o 22:00
22:00 pamplemousse_ joined
Xliff What does it mean when Perl6 returns this exception: "You cannot create an instance of this type (<MyType>)" 22:00
timotimo it's probably a type backed by the Uninstantiable REPR
the only feature that repr has is to give that error when you try to create an instance of it 22:01
22:01 Sharparam left, molaf left
Xliff OK. Problem is, this happens when I am under Test. If I do perl6 -Ilib -e 'use MyClass; MyClass.new' I do NOT get this error. 22:01
22:02 Sharparam joined
Xliff Ah. So right now, .HOW is returning PackageHOW. I would guess that makes it uninstantiable, yes? 22:02
Ah! Misspelled class name. LOL! 22:03
vrurg AlexDaniel: sorry, I was afk. It's a multi-file import-related issue. 22:04
22:04 abraxxa left
AlexDaniel vrurg: so? Make a gist with multiple files and it will work 22:04
vrurg Hm, I've missed that in the documentation. Sure, it will work. 22:06
AlexDaniel vrurg: try it, there are some minor tricks that are required, so while I'm here… :) 22:09
vrurg bisectable6: bad=2019.03.01 good=2018.12 gist.github.com/vrurg/2700db3d8124...-import-p6 22:11
bisectable6 vrurg, Cannot find revision “2019.03.01” (did you mean “2019.03.1”?)
vrurg bisectable6: bad=2019.03 good=2018.12 gist.github.com/vrurg/2700db3d8124...-import-p6 22:12
bisectable6 vrurg, It looks like a URL, but mime type is ‘text/html; charset=utf-8’ while I was expecting something with ‘text/plain’ or ‘perl’ in it. I can only understand raw links, sorry.
AlexDaniel bisectable6: bad=2019.03 good=2018.12 gist.github.com/vrurg/2700db3d8124...2aa44a855f
bisectable6 AlexDaniel, Using file “to_require.pm” as a main file, other files are placed in “sandbox/”
AlexDaniel, On both starting points (old=2018.12 new=2019.03) the exit code is 0 and the output is identical as well
AlexDaniel, Output on both points: «»
AlexDaniel ok it used the wrong file 22:13
vrurg bisectable6: bad=2019.03 good=2018.10 gist.githubusercontent.com/vrurg/2...-import.p6
bisectable6 vrurg, Successfully fetched the code from the provided URL
vrurg, Bisecting by output (old=2018.10 new=2019.03) because on both starting points the exit code is 1
AlexDaniel bisectable6: bad=2019.03 good=2018.12 gist.github.com/AlexDaniel/91bd4f1...7c60ee4bbd
bisectable6 vrurg, bisect log: gist.github.com/baff39eae3e2398544...a655c395b2
vrurg, (2018-10-27) github.com/rakudo/rakudo/commit/84...458acc2e9a
AlexDaniel, Using file “role-import.p6” as a main file, other files are placed in “sandbox/”
AlexDaniel, Bisecting by exit code (old=2018.12 new=2019.03). Old exit code: 0
AlexDaniel, bisect log: gist.github.com/9af23f161333175c97...b8c9af6f9a
AlexDaniel, There are 2 candidates for the first “new” revision. See the log for more details
AlexDaniel vrurg: does that look right? 22:14
vrurg No, it didn't load the module.
22:14 pamplemousse_ left
vrurg Perhaps it tried with .p6 only, didn't fetch .pm 22:14
AlexDaniel vrurg: so the file is in sandbox/to_require.pm 22:15
what's the deal with $*PROGRAM.parent.absolute ?
that's probably in a completely different place, can we just say `require ("sandbox/to_require.pm")` ? 22:16
vrurg It's how the bug was reported. I'll try locally without it thing.
22:17 pamplemousse_ joined
vrurg require ("./to_require.pm") doesn't find the file. 22:18
bisectable6: bad=2019.03 good=2018.12 gist.github.com/vrurg/2700db3d8124...2aa44a855f
bisectable6 vrurg, Using file “role-import.pl” as a main file, other files are placed in “sandbox/” 22:19
vrurg, Bisecting by output (old=2018.12 new=2019.03) because on both starting points the exit code is 1
vrurg, bisect log: gist.github.com/5e186f35d1c8171414...50d19a185a
vrurg, (2017-07-30) github.com/rakudo/rakudo/commit/2d...2da0e5bdae
AlexDaniel c: 2d007bf43^,2d007bf43 gist.github.com/vrurg/2700db3d8124...2aa44a855f 22:20
committable6 AlexDaniel, Using file “to_require.pm” as a main file, other files are placed in “sandbox/”
AlexDaniel, ¦2d007bf43^,2d007bf: «»
vrurg bisectable6: bad=2019.03 good=2018.12 gist.github.com/vrurg/2700db3d8124...2aa44a855f
AlexDaniel vrurg: please add # MAIN comment into the executable file
22:21 rindolf left
AlexDaniel or rename your .pm file as .pm6 22:21
that's why: github.com/perl6/whateverable/blob...#L110-L114 22:22
vrurg I did both.
AlexDaniel bisectable6: bad=2019.03 good=2018.12 gist.github.com/vrurg/2700db3d8124...2aa44a855f
bisectable6 AlexDaniel, Using file “role-import.pl” as a main file, other files are placed in “sandbox/”
AlexDaniel, Bisecting by output (old=2018.12 new=2019.03) because on both starting points the exit code is 1
AlexDaniel, bisect log: gist.github.com/655f757b3b1ec45886...b4dedec61c 22:23
AlexDaniel, (2017-07-30) github.com/rakudo/rakudo/commit/2d...2da0e5bdae
AlexDaniel c: 2d007bf43^,2d007bf43 gist.github.com/vrurg/2700db3d8124...2aa44a855f
committable6 AlexDaniel, Using file “role-import.pl” as a main file, other files are placed in “sandbox/”
AlexDaniel, gist.github.com/1cc63356d0fe120d00...0c9676e70c
vrurg Doesn't it add sandbox to includes? 22:24
AlexDaniel vrurg: ah, perhaps add `use lib ‘sandbox’;`
vrurg bisectable6: bad=2019.03 good=2018.12 gist.github.com/vrurg/2700db3d8124...2aa44a855f
bisectable6 vrurg, Using file “role-import.pl” as a main file, other files are placed in “sandbox/” 22:25
vrurg, Bisecting by exit signal (old=2018.12 new=2019.03). Old exit signal: 0 (None)
vrurg YES! ;)
AlexDaniel ok yeah, that looks promising 22:26
bisectable6 vrurg, bisect log: gist.github.com/c3585346ab2366ae78...15f9c8a4c2
vrurg, (2019-03-07) github.com/rakudo/rakudo/commit/47...d6adea5d84
AlexDaniel what the
sooooo, ok, I guess it doesn't happen every time? 22:27
c: 47e34445f^,47e34445f^,47e34445f^,47e34445f,47e34445f,47e34445f gist.github.com/vrurg/2700db3d8124...2aa44a855f
committable6 AlexDaniel, Using file “role-import.pl” as a main file, other files are placed in “sandbox/”
AlexDaniel, ¦47e34445f^,47e34445f^,47e34445f^,47e3444,47e3444,47e3444: «Cannot mix in non-composable type Any into object of type Int␤ in block <unit> at sandbox/role-import.pl line 6␤␤ «exit code = 1»» 22:28
AlexDaniel vrurg: what are we looking for exactly?
vrurg on 2019.03+ it is invariably reproducible. A role gets punned upon import and mixing in with does dies because of uncomposable. 22:29
I think it is punned or how else it's meta becomes ClassHOW. 22:30
Let me change it so the output would be changing, not the exit code. 22:31
22:31 dolmen left
vrurg bisectable6: bad=2019.03 good=2018.12 gist.github.com/vrurg/2700db3d8124...2aa44a855f 22:31
bisectable6 vrurg, Using file “role-import.pl” as a main file, other files are placed in “sandbox/” 22:32
vrurg, Bisecting by output (old=2018.12 new=2019.03) because on both starting points the exit code is 0
vrurg, bisect log: gist.github.com/6865413a3e1bdd8c6d...1e3eba8cbb
vrurg, (2017-07-30) github.com/rakudo/rakudo/commit/2d...2da0e5bdae
AlexDaniel c: 2d007bf431^,2d007bf431 gist.github.com/vrurg/2700db3d8124...2aa44a855f
committable6 AlexDaniel, Using file “role-import.pl” as a main file, other files are placed in “sandbox/”
AlexDaniel, ¦2d007bf431^,2d007bf: «Perl6::Metamodel::ParametricRoleHOW␤» 22:33
AlexDaniel that's what I don't get?
22:33 dolmen joined
vrurg "Could not open sandbox/role-import.pl. Failed to stat file: no such file or directory" in bisect log. Is it normal? 22:34
ParametricRoleHOW is the right answer.
AlexDaniel well, it affects the result for sure, but I'm not sure why that happens
vrurg But it bisected to a commit dating back to 2017 whereas things worked in 2018. That doesn't look right. 22:35
AlexDaniel 6c: gist.github.com/vrurg/2700db3d8124...2aa44a855f 22:36
committable6 AlexDaniel, Using file “role-import.pl” as a main file, other files are placed in “sandbox/”
AlexDaniel vrurg: I'm guessing that it is something that was merged much later 22:37
I don't remember the history of that commit though
committable6 AlexDaniel, gist.github.com/82ee7e55e6d39d35e7...d28130c16d
AlexDaniel let's try our luck :)
bisectable6: old=2018.10 gist.github.com/vrurg/2700db3d8124...2aa44a855f 22:38
bisectable6 AlexDaniel, Using file “role-import.pl” as a main file, other files are placed in “sandbox/”
AlexDaniel, Bisecting by output (old=2018.10 new=6565810) because on both starting points the exit code is 0
AlexDaniel, bisect log: gist.github.com/78877344cdd3794888...896dae07df 22:39
AlexDaniel, There are 2 candidates for the first “new” revision. See the log for more details
vrurg Good, something to look into! 22:41
AlexDaniel that's still not right because it hit some weird stuff on its way…
vrurg AlexDaniel: right. Both candidates are not related. 22:45
22:46 Cabanossi left
vrurg Ok, I though that'd be something to work with on my flight to Pittsburg. No luck. 22:46
AlexDaniel: anyway, thanks a lot! It was a good opportunity to learn bisectable tricks. 22:47
22:53 Cabanossi joined 22:56 pamplemousse_ left
AlexDaniel vrurg: see this: 23:00
c: 5bbaabb9c6^^^^^^,5bbaabb9c6^^^^^,5bbaabb9c6^^^^,5bbaabb9c6^^^,5bbaabb9c6^^,5bbaabb9c6^,5bbaabb9c6 gist.github.com/vrurg/2700db3d8124...2aa44a855f
committable6 AlexDaniel, Using file “role-import.pl” as a main file, other files are placed in “sandbox/”
AlexDaniel, gist.github.com/fc766d48964f62a232...00f30dd46c 23:01
AlexDaniel vrurg: so one of these 5 commits
vrurg AlexDaniel: indeed. 23:02
23:04 dolmen left
vrurg AlexDaniel: Thanks again! Now it should be easier to locate the problem. 23:06
AlexDaniel you're welcome :) 23:20

Powered by Cro and the Raku® Programming Language.

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

Thank you!