Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2016-06-09

Perl 6 | Reference Documentation | Rakudo

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

All times shown according to UTC.

Time Nick Message
00:00 timeless are they important enough to be in the glossary, or totally hidden?
00:00 timeless (iffy)
00:01 cpage_ joined #perl6
00:04 AlexDaniel m: my $x, my $y Z:= 1,2; say $x; say $y
00:04 camelia rakudo-moar 60731a: OUTPUT«5===SORRY!5=== Error while compiling /tmp/lk3qhBBjGq␤Cannot zip with := because list assignment operators are too fiddly␤at /tmp/lk3qhBBjGq:1␤------> 3my $x, my $y Z:=7⏏5 1,2; say $x; say $y␤»
00:04 AlexDaniel timeless: ↑ these are sometimes present in the error messages
00:04 AlexDaniel timeless: so yes, are important enough
00:06 [Coke] .
00:06 AlexDaniel ,
00:08 BenGoldberg joined #perl6
00:08 araujo joined #perl6
00:11 araujo joined #perl6
00:18 timeless AlexDaniel: ok, i've sent a pr for glossary bits
00:18 timeless no guarantees, i may have messed everything up, but... it's probably an improvement
00:24 AlexDaniel timeless: in your opinion, what's the current state of S99?
00:24 cpage_ joined #perl6
00:25 AlexDaniel timeless: should we just redirect everything to http://doc.perl6.org/language/glossary ?
00:25 timeless i think once my bits are merged you can
00:26 timeless you probably want to run one more set of eyes side-by-side before you do
00:26 timeless but i think i got all the things that were meaningfully different
00:34 dalek doc: 38982ff | (Josh Soref)++ | doc/Language/glossary.pod:
00:34 dalek doc: add NYI to glossary
00:34 dalek doc: review: https://github.com/perl6/doc/commit/38982ff4d2
00:34 dalek doc: de48add | (Josh Soref)++ | doc/Language/glossary.pod:
00:34 dalek doc: define Enum in glossary
00:34 dalek doc: review: https://github.com/perl6/doc/commit/de48add2ad
00:34 dalek doc: add6048 | (Josh Soref)++ | doc/Language/glossary.pod:
00:34 AlexDaniel timeless: ta-daaa!
00:34 dalek joined #perl6
00:35 AlexDaniel timeless: I've also just opened this: https://github.com/perl6/doc/issues/570
00:39 AlexDaniel timeless: also this https://github.com/perl6/doc/issues/571
00:40 timeless erm, yeah
01:06 kalkin- joined #perl6
01:17 dj_goku timeless: are you working on the whole bunch of broken links issue number 561 https://github.com/perl6/doc/issues/561 ?
01:20 timeless dj_goku: not instantly
01:20 timeless if you want to do something, you can, i might in a bit
01:20 timeless first i need to fix something elsewhere
01:21 dj_goku ahh cool.
01:24 Actualeyes joined #perl6
01:48 FROGGS_ joined #perl6
01:58 araujo_ joined #perl6
02:00 araujo_ joined #perl6
02:08 araujo_ joined #perl6
02:15 araujo joined #perl6
02:23 mr-foobar joined #perl6
02:30 cdg joined #perl6
02:35 noganex_ joined #perl6
02:48 * timeless tries to remember how to use some perl tools
02:50 araujo joined #perl6
02:52 timeless can cpan get stuck in a loop?
02:54 huggable joined #perl6
03:08 BenGoldberg It has recursion detection, so no.
03:08 yoleaux 7 Jun 2016 18:18 EDT <hoelzro> BenGoldberg: that's not a bad idea, but if .resume worked on that exception (which I don't know if it does), it would probably just resume parsing the current chunk of code
03:08 yoleaux 7 Jun 2016 18:18 EDT <hoelzro> BenGoldberg: sadly, the exception handling for that probably isn't smart enough (yet) to ask the caller to feed the parser more input
03:10 BenGoldberg timeless, But, wouldn't that question be better on #perl? :)
03:10 timeless BenGoldberg: i'm only using perl because i'm trying to help perl6 :)
03:11 * BenGoldberg wonders if perl 5 is a dependency for perl 6.
03:15 parabolize rakudobrew is perl5
03:15 parabolize strange perl5 at that
03:20 timeless dj_goku: anyway, i'm now running linkchecker
03:26 araujo joined #perl6
03:35 araujo joined #perl6
03:36 BenGoldberg m: (loop { True }).perl.say
03:36 camelia rakudo-moar 60731a: OUTPUT«(timeout)»
03:36 BenGoldberg m: (loop { True }).WHAT.say
03:36 camelia rakudo-moar 60731a: OUTPUT«(Seq)␤»
03:40 mohae joined #perl6
03:41 dj_goku timeless: cool, does that crawl all the html making sure it works?
03:41 timeless seems to, slowly.
03:42 timeless https://www.irccloud.com/pastebin/bj4O0gdP/
03:43 timeless https://www.irccloud.com/pastebin/f03rwveG/
03:50 dj_goku cool
03:52 timeless dj_goku: so, here's a good one
03:52 timeless doc/Language/operators.pod
03:52 timeless L<#prefix ...> etc. stub operators.
03:52 timeless =head2 infix C«...»
03:52 timeless =head2 listop C«...»
03:52 timeless there's no prefix thing...
03:54 * timeless doesn't know what's correct
04:01 TimToady building rakudo has Perl 5 dependencies
04:01 TimToady running it does not, unless you need Inline::Perl5
04:03 dj_goku timeless: there needs to be two links
04:03 dj_goku timeless: I was looking at the one specifically too.
04:05 dj_goku timeless: you see prefix/postfix for --? it links to a landing page: http://doc.perl6.org/routine/--
04:05 dj_goku it looks like a few of those just need those landing pages.
04:08 timeless https://www.irccloud.com/pastebin/NfFlVyUz/
04:09 timeless https://www.irccloud.com/pastebin/YxLpqlqk/
04:10 timeless i'm not actually sure how to read this report...
04:11 timeless ok, i think this is how to read it https://www.irccloud.com/pastebin/BpNWeEi4/
04:15 timeless oh, wow
04:16 timeless here's my pod checker https://www.irccloud.com/pastebin/FhpPEmTp/
04:17 cpage_ joined #perl6
04:17 timeless hrm, no, my thing is broken
04:17 timeless ok, this is better https://www.irccloud.com/pastebin/3MG7OrQ0/
04:18 timeless ok, that looks correct
04:18 timeless doc/doc/Type/CurrentThreadScheduler.pod is really missing `method cue`
04:20 timeless so, what's the correct form?
04:20 timeless that L<Scheduler#method cue> blocks until the code has finished executing.
04:20 timeless that L<../Scheduler#method cue> blocks until the code has finished executing.
04:20 xinming joined #perl6
04:21 Cabanossi joined #perl6
04:35 timeless dj_goku: so, what's the operator precedence for `prefix ...` ?
04:41 inokenty joined #perl6
04:43 dj_goku timeless: does that matter? with the docs?
04:43 timeless if i understand this thing, it influences where in the file i put it
04:44 dj_goku I haven't looked to far, I am just trying to get up to speed to try and help. :D
04:44 timeless that'd help me
04:44 timeless i can send a pr for some other bits
04:45 dj_goku plus I am trying to figure out weechat so. :)
04:47 d^_^b joined #perl6
04:52 sammers_ joined #perl6
05:26 pnu_ joined #perl6
05:26 jnap_ joined #perl6
05:29 SmokeMachine____ joined #perl6
05:31 Spot__ joined #perl6
05:31 avenj joined #perl6
05:31 Lucas_One_ joined #perl6
05:32 olinkl joined #perl6
05:33 kipd joined #perl6
05:34 BuildTheRobots joined #perl6
05:34 M-matthew joined #perl6
05:35 ggherdov joined #perl6
05:36 notbenh_ joined #perl6
05:36 chansen_ joined #perl6
05:36 rurban joined #perl6
05:37 parisba joined #perl6
05:43 Actualeyes joined #perl6
05:45 ufobat joined #perl6
05:51 rindolf joined #perl6
05:55 ufobat morning :D
05:57 cpage_ joined #perl6
06:01 ssotka joined #perl6
06:02 firstdayonthejob joined #perl6
06:08 parisba joined #perl6
06:08 notbenh_ joined #perl6
06:15 sno joined #perl6
06:29 rurban_ joined #perl6
06:34 masak morning!
06:36 araujo_ joined #perl6
06:39 moritz \o!
06:40 timeless masak: i sent some pr's...
06:43 timeless do any of the bots here let me ask for perldoc?
06:48 timeless m: use Pod::Perldoc; @ARGV="pod"; Pod::Perldoc->run()
06:48 camelia rakudo-moar 60731a: OUTPUT«===SORRY!===␤Could not find Pod::Perldoc at line 1 in:␤    /home/camelia/.perl6␤    /home/camelia/rakudo-m-inst-2/share/perl6/site␤    /home/camelia/rakudo-m-inst-2/share/perl6/vendor␤    /home/camelia/rakudo-m-inst-2/share/perl6␤    CompUnit::…»
06:48 timeless :-(
06:52 timeless http://doc.perl6.org/type/Pod::Item
06:52 timeless Pod::Item inherits from class Pod::Block, which provides the following methods:
06:53 timeless mst: could you change the pod generator to not escape `:`s in links?
06:53 timeless Pod::Block links to http://doc.perl6.org/type/Pod%3A%3ABlock
06:53 timeless which is ugly, and unnecessary
06:54 timeless (well, as long as the beginning is properly written)
06:54 huggable joined #perl6
06:56 moritz timeless: you can change it yourself
06:56 timeless htmlify.p6?
06:56 moritz yes
06:56 moritz though it's quite dangerous to change the escaping
06:57 moritz you should run a link checker before and afterwards to make sure you didn't accidentally break anything
06:57 rurban joined #perl6
06:57 * timeless would have to get a perl6 engine
06:57 * timeless doesn't have any
06:57 timeless can i ask cpan(perl5) for perl6? :)
06:58 moritz use https://github.com/tadzik/rakudobrew
06:58 moritz you can ask a lot :-)
07:01 rurban joined #perl6
07:02 timeless wow, no obvious spelling errors there :o
07:03 zakharyas joined #perl6
07:04 darutoko joined #perl6
07:05 rurban1 joined #perl6
07:06 timeless m: use URI::Escape; uri_escape("Pod::Block")
07:06 camelia rakudo-moar 60731a: OUTPUT«===SORRY!===␤Could not find URI::Escape at line 1 in:␤    /home/camelia/.perl6␤    /home/camelia/rakudo-m-inst-2/share/perl6/site␤    /home/camelia/rakudo-m-inst-2/share/perl6/vendor␤    /home/camelia/rakudo-m-inst-2/share/perl6␤    CompUnit::R…»
07:07 * timeless frowns
07:07 moritz star-m: use URI::Escape; say uri_escape("Pod::Block")
07:07 camelia star-m 2016.01: OUTPUT«Pod%3A%3ABlock␤»
07:07 timeless what's the `star` thing do?
07:07 moritz huh, I guess a more up-to-date star wouldn't hurt
07:07 timeless star-m: use URI::Escape; uri_escape("Pod::Block")
07:07 camelia star-m 2016.01: ( no output )
07:07 moritz timeless: Rakudo Star is a distribution containing the Rakudo compiler, some modules, some docs
07:07 moritz and panda
07:08 _mg_ joined #perl6
07:08 timeless star-m: use URI::Escape; sub href_escape($ref) {return uri_escape($ref).subst('%3A%3A', '::', :g);} say href_escape("Pod::Block")
07:08 camelia star-m 2016.01: OUTPUT«5===SORRY!5=== Error while compiling /tmp/xk8VTaJjjP␤Strange text after block (missing semicolon or comma?)␤at /tmp/xk8VTaJjjP:1␤------> 3escape($ref).subst('%3A%3A', '::', :g);}7⏏5 say href_escape("Pod::Block")␤    expecting any of:␤      …»
07:08 timeless star-m: use URI::Escape; sub href_escape($ref) {return uri_escape($ref).subst('%3A%3A', '::', :g);}; say href_escape("Pod::Block")
07:08 camelia star-m 2016.01: OUTPUT«Pod::Block␤»
07:12 timeless moritz: https://github.com/perl6/doc/pull/573 wanna try it?
07:13 rurban joined #perl6
07:13 * timeless wonders if this output from rakudobrew is ok https://www.irccloud.com/pastebin/ja99RQ5f/
07:16 moritz timeless: no, I don't wanna try it. I want you to try it.
07:16 * timeless is trying to get rakudobrew to brew...
07:17 timeless is the idea of rakudobrew that it should get everything it needs, or does it expect a fully working everything?
07:18 moritz timeless: what's your perl 5 version?
07:18 timeless This is perl 5, version 18, subversion 4 (v5.18.4) built for ppc64le-linux-thread-multi
07:18 moritz timeless: then your perl 5 installation is broken
07:18 moritz timeless: Perl::OSType is a core module starting from perl v5.13.9, so it should be available
07:19 moritz are you using some Redhat (or derivative) that ships a Perl without its core modules?
07:19 timeless Description:    Fedora release 21 (Twenty One)
07:19 timeless and `using` is correct, it's not my box
07:20 timeless Linux gcc2-power8.osuosl.org 3.17.4-301.fc21.ppc64le #1 SMP Mon Dec 1 07:51:01 UTC 2014 ppc64le ppc64le ppc64le GNU/Linux
07:20 * moritz sighs
07:20 moritz Redhat, breaking perl since at least 2003
07:20 * timeless was using redhat in '98
07:21 timeless although, tbh, i don't recall that being something that was broken
07:21 * timeless had more problems w/ nic drivers
07:23 Actualeyes joined #perl6
07:26 * timeless gets a perl6
07:29 * timeless fails to get a panda https://www.irccloud.com/pastebin/iDwElRS4/
07:36 timeless so, i /think/ that's panda trying to use /usr/bin/prove (perl5)
07:39 wamba joined #perl6
07:47 araujo__ joined #perl6
07:50 araujo__ joined #perl6
07:52 RabidGravy joined #perl6
07:52 araujo__ joined #perl6
07:55 rurban joined #perl6
07:59 rurban1 joined #perl6
08:04 dakkar joined #perl6
08:05 Zero_Dogg joined #perl6
08:05 Zero_Dogg joined #perl6
08:10 Actualeyes joined #perl6
08:13 bjz_ joined #perl6
08:33 Hotkeys why doesn't p6 play nice with unicode anymore on windows
08:33 Hotkeys in the repl anyway
08:34 grassass joined #perl6
08:39 g4 joined #perl6
08:55 ocbtec joined #perl6
09:01 masak Hotkeys: could you be more specific?
09:03 araujo joined #perl6
09:09 masak in Perl 6, what's an idiomatic way to ask for all the declared classes/roles provided by a `use`d module?
09:11 moritz you mean besides looking at the docs? :-)
09:11 masak yes, I mean in code
09:11 masak my question is triggered by this comment: https://github.com/masak/007/commit/31b531946cc513476cbb9c8d966a08be35c6767c#commitcomment-17801202
09:11 masak in fact, "by a `use`d module" is not essential
09:12 moritz by the time you use'd a module, it's GLOBAL:: is already merged, I don't know if it's still available somehow
09:12 masak main thing is to have some way from within the code itself to (say) find all the roles and classes with a name like Q::*
09:12 rurban joined #perl6
09:13 moritz m: say X::.keys
09:13 camelia rakudo-moar d292d7: OUTPUT«(Method NoSuchSymbol NotEnoughDimensions OS InvalidType Pairup NYI PseudoPackage EXPORTHOW IO Set Package Declaration Inheritance Redeclaration StubCode UnitScope PoisonedAlias Placeholder IllegalOnFixedDimensionArray Temporal Proc InvalidTypeSmiley Numeri…»
09:13 masak hmm :)
09:13 * masak tries this
09:13 moritz note that it only gives you one level at first
09:13 masak nodnod
09:13 moritz m: say X::Proc::.keys
09:13 camelia rakudo-moar d292d7: OUTPUT«(Async Unsuccessful)␤»
09:13 masak so I need to tree-walk, no biggie
09:13 masak yay, that works
09:14 moritz .oO( let's do the tree-walk again)
09:17 wamba joined #perl6
09:18 ocbtec joined #perl6
09:26 mr-foobar joined #perl6
09:27 rurban1 joined #perl6
09:29 Hotkeys masak I'm in bed now, but basically it deletes Unicode chars or turns them into "?" (not a weird representation, a literal "?")
09:31 rurban joined #perl6
09:31 Hotkeys This is with powershell
09:31 Hotkeys But powershell parses the chars fine outside the p6 repl
09:34 nine masak: note that this only works if the module is nice and sticks to it's namespace. It also doesn't include modules use'd by the module which may or may not be an advantage for you.
09:37 masak all that works to my advantage
09:38 masak the test and the module are all in the same code base
09:38 masak but yeah, good points
09:40 nine masak: a more complete solution would be $*REPO.need(CompUnit::DependencySpecification.new(:short-name('Q'))).handle.globalish-package.WHO.keys
09:40 nine m: $*REPO.need(CompUnit::DependencySpecification.new(:short-name("Test"))).handle.globalish-package.WHO.keys.say
09:40 camelia rakudo-moar d292d7: OUTPUT«(Test)␤»
09:40 masak nine: I don't quite grok that, but looks neat
09:42 nine masak: "use Test;" is pretty much BEGIN GLOBAL.WHO.merge-symbols($*REPO.need(CompUnit::DependencySpecification.new(:short-name("Test"))).handle.globalish-package.WHO);
09:43 araujo joined #perl6
09:44 labster1 joined #perl6
09:54 moritz plus a bit of importing
10:03 gregf_ https://gist.github.com/anonymous/af60e979bfa4df3a33b1ed99e84f6b68 # coded something in perl6 ;)
10:10 rurban joined #perl6
10:15 donaldh joined #perl6
10:27 TEttinger joined #perl6
10:29 cpage_ joined #perl6
10:29 rurban joined #perl6
10:32 rurban1 joined #perl6
10:36 rurban joined #perl6
10:37 Xliff joined #perl6
10:42 bjz joined #perl6
10:43 retupmoca joined #perl6
10:49 iH2O joined #perl6
10:56 kid51 joined #perl6
10:59 yqt joined #perl6
10:59 colomon joined #perl6
11:06 aries_liuxueyang joined #perl6
11:06 sexy-coder-girl joined #perl6
11:09 literal so, it looks like <.ws> can't capture anything, correct? I have <comment> in there that I would like captured in between tokens, how would I do that?
11:09 moritz literal: call <ws>
11:10 moritz the . in <.ws> instructs it not to capture
11:10 literal I see
11:10 literal works now, thanks
11:10 moritz literal: the other option is bypass the traditional capturing mechanism, and track some comments through a different mechnaism
11:11 moritz token comment { # \N* $$ { self.record_comment($/) } }
11:11 zxd joined #perl6
11:13 iH2O left #perl6
11:14 rindolf joined #perl6
11:16 gnull joined #perl6
11:31 mr-foobar joined #perl6
11:35 rurban_ joined #perl6
11:40 mr-foobar joined #perl6
11:47 SHODAN joined #perl6
12:07 kalkin- joined #perl6
12:09 mst timeless: on RHish OSen, you need perl-core to have a complete perl5 install. they did a split like debian did, except being RH, did it the opposite way around to the way debian does it
12:10 SHODAN joined #perl6
12:10 mst timeless: so on RH 'perl' doesn't have half the core modules and you need 'perl-core', on debian 'perl-core' is the stripped down version and 'perl' is an actual perl
12:10 mst timeless: distros \o/
12:10 mst moritz: ^^ eventually you get used to explaining this to people
12:11 tadzik ...'perl-core' is the *super*set of 'perl'
12:11 tadzik do I read that right?
12:12 tadzik #damnitredhat
12:15 masak tadzik: it's the kind of core that's bigger than the thing it's contained in
12:15 masak hashtag #tardis
12:17 moritz sounds like an IoS device
12:17 moritz where IoS stands for Inversion of Space :-)
12:18 tadzik mst: I recall you putting Rakudo::Star on CPAN, do I remember it wrong or is there some reason why it's not there anymore?
12:21 rurban joined #perl6
12:26 mst tadzik: it isn't there anymore because I did it as a proof of concept, and then it turned out because I had pumpking bits in pause and had forgotten to no_index the rakudo sources the release was causing confusion
12:26 tadzik ah, fairy nuff
12:27 mst tadzik: and the *idea* was it would always get updated when star was released, but nobody doing star release management ever tried or ever remembered to remind me
12:27 mst when I realised this had happened four releases in a row and nobody'd noticed, I sort of assumed people didn't care much
12:27 mst also that was parrot era
12:27 mst I should get around to reinstating it now
12:27 tadzik it was? Damn, time flies
12:27 mst since I can go onto #perl6-release and bully people until they get used to reminding me or something
12:30 tadzik we could put that step in the release instructions
12:30 mst and eventually I can probably integrate it into the release process
12:30 tadzik "update wikipedia, poke mst to update Rakudo::Star"
12:30 masak was gonna say
12:31 masak step in the release process or it didn't happen
12:31 masak mst: how do you prefer to be contacted on such a thing?
12:31 mst and nobody added one last time, and it didn't
12:32 mst masak: once I've got the thing re-done (remember I need to do a new set of hackery for Moar) I suspect 'kick mst repeatedly on #perl6-release' will work fine
12:32 masak mst: what if you're not online on IRC during the release process -- does the process block at the last step until you can be contacted?
12:32 masak I mean, Wikipedia tends to be there fairly regularly. less reliably so with human meat sacks like us
12:34 mst I would suggest that it doesn't, because I only intend to do it manually a few times anyway, and then I write a script that does it and give the release managers co-maint on the cpan5 namespace
12:46 masak ok, so async poking of mst
12:46 rurban joined #perl6
12:47 mst but lemme make it work at all again first ;)
12:47 donaldh joined #perl6
12:49 tadzik hmm
12:50 tadzik m: sub MAIN($arg?) { say "OK" }
12:50 camelia rakudo-moar 2095ed: OUTPUT«OK␤»
12:50 tadzik m: sub MAIN($arg? is rw) { say "OK" }
12:50 camelia rakudo-moar 2095ed: OUTPUT«5===SORRY!5=== Error while compiling /tmp/bT55Ys0oFo␤Cannot use 'is rw' on an optional parameter␤at /tmp/bT55Ys0oFo:1␤»
12:50 tadzik *grumble*
12:50 masak I vaguely recall hearing a rationale for that
12:50 masak don't remember if I still grumbled afterwards, though
12:51 tadzik my point being: https://gist.github.com/tadzik/a63294b0cf753af8faf348e812ffe6e3
12:51 masak basically `is rw` means "I should be able to mutate this container"
12:51 masak whereas `?` means "there might not be a thing there at all"
12:51 tadzik I guess that makes it an LTA error message at least
12:52 * jnthn tries to work out how main could get something that matches `rw`
12:52 yqt joined #perl6
12:53 tadzik not that I know exactly what I expected... :)
12:53 jnthn :-)
12:53 jnthn `is copy` makes sense, just means you'd like an assignable $arg
12:54 tadzik right, this is what I was after
12:54 tadzik I mistakenly put 'is rw' in there, and suddenly my invocation failed, I didn't expect 'is rw' to be something that may decide whether the dispatch is valid or ont
12:55 jnthn Your invocation didn't fail, the script didn't even compile?
12:55 sufrostico joined #perl6
12:56 jnthn Or did you not have it as optional?
12:57 jnthn Ah yeah, your original in the gist didn't have the ?
12:57 jnthn But yeah, you can dispatch on rw-ness
12:57 jnthn m: multi m($a is rw) { say 'container' }; multi m($a) { say 'value' }; my $x = 42; m($x); m(5)
12:57 camelia rakudo-moar 2095ed: OUTPUT«container␤value␤»
13:02 MilkmanDan joined #perl6
13:04 stmuk_ is GH slow or is it me?
13:05 domidumont joined #perl6
13:06 lizmat status.github.com seems to indicate response times going up
13:06 rurban1 joined #perl6
13:09 kaare_ joined #perl6
13:10 aries_liuxueyang joined #perl6
13:13 rurban joined #perl6
13:34 timeless mst: so... pretend i'm not root (because i'm not)
13:34 timeless what magic do i need tot get the equivalent of perl-core?
13:34 timeless i probably can just ask cpan to give me a perl5....
13:34 mst timeless: http://p3rl.org/Perl::Build into ~ then
13:35 mst timeless: cpan probably doesn't work properly if you've only got half a perl
13:35 colomon joined #perl6
13:35 timeless mst: basically ignore the fact that i have something vaguely resembling perl?
13:35 timeless once i fixed my env so that it found lib::local, cpan worked well enough for me to install and run the w3 link checker
13:35 mst heh, you mean local::lib
13:36 timeless sure.. whatever
13:36 mst hey, I wrote that module, I'm allowed to care about getting the name right :P
13:36 timeless i probably did it between 1 and 4am today
13:36 timeless and i had a few hours sleep before
13:36 timeless ok, well, thanks, it worked -- once i added the env
13:36 mst well, basically, either you can mess around trying to backfill your way to a full perl install
13:36 mst or you can outsource building yourself a standard one to Perl::Build
13:37 timeless (that part was annoying, it could have warned me that i'd needed to add stuff to my env for it to be useful)
13:37 mst I'm a perl5 toolchain hacker and I wouldn't bother trying option one
13:37 timeless i've built perl5 before, (not recently)
13:37 timeless and i've recently built both py2 and py3
13:37 mst and Perl::Build is a cpan module that provides a perl-build script that automates the process
13:37 timeless i think i want `plenv` first
13:38 mst plenv is some shell stuff wrapped around Perl::Build
13:38 mst you can if you want, I don't bother
13:38 timeless oh
13:38 timeless so it isn't like virtualenv?
13:38 * timeless asks cpan to install Perl::Build
13:38 mst local::lib is more like virtualenv
13:38 timeless :o
13:38 mst or perlbrew + perlbrew's local::lib support
13:39 timeless btw, this is pretty nuts
13:39 mst honestly, I just perl-build into ~/local/
13:39 mst stick a local::lib on top
13:39 mst and move on
13:39 timeless i shouldn't have to use cpan to install perl::build to build perl5 to build perl6 {stuff}
13:39 mst also, http://p3rl.org/local::lib does tell you that you need the env vars, but if memory serves RH comes with a customised initial CPAN config that uses local::lib but doesn't tell you
13:39 timeless although, if it means i don't have to read makefiles, i suppose i should be thankful
13:39 stmuk_ you should be able to build perl6 with any recent distro's system perl
13:40 timeless stmuk_: it so didn't work
13:40 mst stmuk_: it's an RHish system with half a perl install and he doesn't have root
13:40 mst so he can't install the other half via rpms
13:40 timeless i could technically unpack them and futz around w/ their locations
13:40 * timeless has done that in another life too
13:40 mst you could. but just making perl-build do it seems like less hassle
13:40 timeless but at some point my hair will go grey
13:40 stmuk_ you will need perl 5.10 or better
13:41 timeless This is perl 5, version 18, subversion 4 (v5.18.4) built for ppc64le-linux-thread-multi
13:41 timeless Linux gcc2-power8.osuosl.org 3.17.4-301.fc21.ppc64le #1 SMP Mon Dec 1 07:51:01 UTC 2014 ppc64le ppc64le ppc64le GNU/Linux
13:41 timeless Description:    Fedora release 21 (Twenty One)
13:41 timeless stmuk_: i trust mst
13:41 timeless (and the fact that it didn't work...)
13:41 stmuk_ I don't
13:42 stmuk_ what error did you see?
13:42 timeless stmuk_: set up an fc21 vm and create a non root user :)
13:42 timeless https://www.irccloud.com/pastebin/iDwElRS4/
13:42 timeless is what i got when i tried to build panda
13:43 stmuk_ panda doesn't need perl5
13:43 mst wait, that's a perl6 error, I thought you said it was missing a perl5 module
13:43 timeless i got a different error when i tried to build moar which i was able to resolve once i had added + configured local::lib + used cpan to get one more thing
13:43 mst timeless: that's not the error I thought you had, please switching to listening to stmuk_ for a bit
13:43 mst since that's a perl6 error and I'm better at perl5 errors, sorry
13:43 mst stmuk_: doh
13:43 timeless mst: here was the failure for moar: https://www.irccloud.com/pastebin/ja99RQ5f/
13:44 skids joined #perl6
13:44 timeless mst: the perl6 error is coming from `prove` which is a perl5 script...
13:44 mst timeless: yeah, but prove is just running a test file
13:44 mst prove can run perl6 tests as well as perl5
13:44 mst (actually, I have a project where some of the tests are Tcl scripts)
13:44 * timeless shrugs
13:45 mst if you have local::lib going then yeah throwing Perl::OSType into there should be fine
13:45 timeless ok stmuk_ the panda error is for you
13:45 timeless mst: yeah, Perl::OSType was all i needed once local::lib was happy
13:45 mst timeless: sorry for the confusion, the 'RH half-a-perl' thing is so common I pattern matched
13:45 timeless otoh, the moar error at least gave me a hint about how to fix
13:45 timeless the panda error is basically useless
13:45 timeless it doesn't suggest how to get debugging
13:45 stmuk_ timeless: I don't think you are following the instructions of putting rakudobrew in the path you are calling it relative
13:46 stmuk_ neither are you using .rakudobrew
13:46 rurban1 joined #perl6
13:46 stmuk_ I think if you follow the instructions exactly it should work .. I just build on a PI3 and I'm sure Fedora works
13:47 stmuk_ s/build/built
13:47 timeless still fails :)
13:47 stmuk_ start again and follow the exact instructions
13:47 timeless hrm
13:47 * timeless kicks ln -s
13:47 stmuk_ is this on PPC?
13:48 timeless yes
13:48 stmuk_ ah .. does anyone know if that actually works? JIT of course won't work
13:49 timeless there's now a ~/.rakudobrew https://www.irccloud.com/pastebin/3GvsIhoi/
13:49 cpage joined #perl6
13:50 timeless (it's a symlink, but i'd hope that doesn't trip anything up)
13:50 stmuk_ I would still follow the instructions to set the PATH
13:50 timeless I set that path...
13:51 stmuk_ you didn't export it
13:51 cpage_ joined #perl6
13:51 timeless PATH is already exported
13:51 timeless if it wasn't, rakudobrew wouldn't have been found...
13:52 timeless i know it's possible to have a system w/o PATH exported, but it's pretty rare for someone who's actually using a system to manage that
13:55 timeless fwiw, the README.md instructions for rakudobrew are disappointing
13:55 timeless this doesn't mention removing stuff from your .profile https://www.irccloud.com/pastebin/HXqY3J07/
13:55 stmuk_ its intended for developers on perl6 itself
13:56 timeless that's nice
13:56 wamba joined #perl6
13:56 stmuk_ you might be better off with rakudo star if you are less familar with UNIX
13:56 timeless url?
13:56 stmuk_ http://www.perl6.org/downloads/
13:56 bjz_ joined #perl6
13:57 timeless mst: let's temporarily assume that perl5 isn't well configured, what version should i tell perl-build to get for me?
13:57 stmuk_ you might also be better trying on intel before PPC if you haven't built before
13:58 stmuk_ although it would be interesting to hear of PPC builds
13:58 mst timeless: 99% sure that isn't the problem now
13:58 timeless i'm using this box because of the boxes in the gcc farm, it's one of the beefier ones
13:59 timeless cpuinfo|tail https://www.irccloud.com/pastebin/rMUV0ciK/
13:59 timeless stmuk_: could you temporarily assume that i can run and understand basic unix commands and help me troubleshoot the .t failure for panda?
14:00 stmuk_ timeless: Sorry I have to leave shortly to do other things
14:00 timeless is there a way to get rakudobrew to ask prove/friends for debugging/verbose output?
14:00 timeless (should I try zef?)
14:02 cdg joined #perl6
14:02 timeless Done, built zef for moar-nom
14:04 timeless ok, zef worked
14:04 * timeless can finally run `make html` in perl6/doc
14:06 timeless mst: is there a simple `perl web server`? :)
14:07 timeless something i can use to host the perl6doc repo on localhost:someport ?
14:10 iH2O joined #perl6
14:10 perlpilot timeless: HTTP::Server::Simple::Static is fairly simple.  Mojo can be used to serve static files.
14:11 perlpilot etc.
14:11 timeless perlpilot: i just want to serve the entire perl6 generated doc tree so i can run a link checker
14:11 timeless i'd rather avoid fighting apache
14:11 iH2O who wants to fight the apaches?
14:12 timeless yeah, HTTP::Server::Simple::Static looks like it'll do what i need
14:12 timeless although i'll have to figure out how to tell it  my port
14:12 timeless yay! the docs built
14:13 tharkun joined #perl6
14:16 iH2O left #perl6
14:19 ptolemarch joined #perl6
14:21 SHODAN joined #perl6
14:23 timeless of course for `Static` to work, i'd have to teach it about `index.html`
14:23 * timeless looks at mojo
14:24 perlpilot timeless: if you've got Mojolicious and Mojolicious::Plugin::Directory installed, you can use perl -Mojo -E 'a->plugin("Directory", root => "/path/to/files") ->start' daemon --listen http://*:<port>
14:24 * timeless asks cpan for those two things
14:25 timeless it's really `-Mojo`?
14:25 perlpilot yep
14:25 timeless (that means the library is ojo, right?)
14:26 perlpilot It's a cute thing that comes with Mojo
14:26 perlpilot yes.
14:26 cpage__ joined #perl6
14:26 mst perlpilot: that was the one I was thinking of and couldn't remember
14:26 perlpilot Though, it's still not going to do anything special for a file named index.html.  It's just another file to present.
14:27 timeless perlpilot: i just need something smart enough to treat `*/` => `*/index.html`
14:27 timeless and again, a goal being to avoid talking to apache (or tomcat, or...)
14:28 perlpilot mst: aye, it's in the docs for Mojolicious::Plugin::Directory
14:28 * perlpilot meeting &
14:29 timeless so... `::` seems to be working...
14:30 * timeless is tempted to try to improve the w3 link checker
14:30 molaf joined #perl6
14:31 timeless this is odd https://www.irccloud.com/pastebin/ikZRdzaq/
14:32 tony-o timeless: might check out xenu's link sleuth
14:32 tony-o http://home.snafu.de/tilman/xenulink.html
14:33 timeless System requirements: Microsoft Windows 95/98/ME/NT/2000/XP/Vista/7/8/10, WININET.DLL required (included with Internet Explorer).
14:33 timeless this is on a firewalled linux-ppc machine...
14:33 perigrin What about 3.12?
14:33 timeless No, it won't work on Windows 3.11, not even with Win32s.
14:33 tony-o i think he addresses that
14:34 perigrin I'm running DOS 6.5 though ...
14:34 timeless No, I won't make a Java, MacOS, Linux, Knoppix, Ubuntu, Beos, Palm, C64, SAP, AmigaOS, Blackberry, Symbian, iPhone or Android version. Don't even ask!
14:34 tony-o i have os2
14:34 tony-o runs fine for me
14:34 timeless he managed to misspell BeOS and BlackBerry :-(
14:34 timeless ok, so, apparently i could run it w/ wien
14:34 timeless wine
14:34 tony-o is RIM still around?
14:34 timeless of course, i don't think this system has x11
14:34 timeless my former employer? yes
14:34 timeless they're still laying people off
14:35 timeless they changed their name to BlackBerry a few years ago
14:36 pmurias joined #perl6
14:36 tony-o that seems like a poor decision, BB already lost the race by then
14:37 timeless it wasn't their poorest decision
14:38 timeless `hiring` Alicia Keys w/o a very strongly designed contract was near the top of the list of poor decisions (imo)
14:38 tony-o did leont leave the perl6 community?
14:39 timeless (a brand embassador w/o a very strict penalty for being caught using [tweeting from]  a competitive product)
14:40 dogbert17 o/ #perl6
14:40 timotimo .seen leont
14:40 yoleaux I saw leont 24 Apr 2016 14:45Z in #perl6: * leont has lived with --std=c++11 for a while now, and wouldn't ever want to look back
14:40 timotimo leont often stays away from the irc for a few months at a time, as he's got plenty other stuff to worry about (that's my guess anyway)
14:41 dogbert17 have just stumbled upon a regression, one of my projext euler scripts suddenly segfaults
14:41 timotimo uh oh
14:41 timotimo segfaults are good, though. usually comparatively easy to debug
14:41 timotimo can you get a backtrace with perl6-gdb-m?
14:41 timotimo does it have any threading in it?
14:42 dogbert17 timotimo: https://gist.github.com/dogbert17/d7822b6db5daa88a372c835727bf0729
14:43 timotimo OK, and just for completeness sake, can you add perl6 --version to it?
14:44 dogbert17 gimme a sec :)
14:45 dogbert17 timotimo: done gist updated
14:47 perlpilot does bisectbot handle gists?
14:47 timotimo the perl6 stuff on my desktop is a bit out of date, so i'll build newest first
14:48 perlpilot Also, does bisectbot handle long-running programs gracefully?
14:48 tadzik so, the general pattern for "I need time more precisely than time()" is "use nqp; nqp::time_n()"? I don't see Time::HiRes or anything in the ecosystem
14:49 jnthn m: say now
14:49 camelia rakudo-moar 2095ed: OUTPUT«Instant:1465483787.822199␤»
14:51 tadzik ...oh
14:51 timotimo oh hey tadzik :)
14:51 tadzik oh hey timo :)
14:52 tadzik for some reason I thought 'now' gives integer values
14:52 rindolf joined #perl6
14:52 tadzik 9¯\_(ツ)_/¯
14:54 * dogbert17 perl6-valgrind-m is slooooow
14:54 huggable joined #perl6
14:55 tony-o timeless: thanks -
14:55 nemo dogbert17: s/perl6-//;s/-m//
14:55 timotimo oh yes, it is
14:55 tony-o tadzik: i wasn't able to get that working when i made `bench` - i'd be interested to know if it works now.  it was specifically erroring out on win
14:55 dogbert17 perhaps I should have run perl6-gdb-m instead
14:56 nemo dogbert17: I've run valgrind so many times on hedgewars matches at like a few fps to catch our fails.  I'd usually launch it then check back half an hour later ☺
14:56 timotimo i'm reproducing the crash and it's interesting
14:56 dogbert17 nemo: interesting
14:56 dogbert17 timotimo: cool
14:56 * timeless so doesn't understand the magic of doc.perl6.org
14:57 tadzik tony-o: hm, no idea, I don't have windows around to try it on
14:57 tadzik seems to work fine on linux though
14:58 tony-o err oh, my problem was getting sys/cpu/real or whatever
14:58 tony-o i'll try it on win
14:59 tadzik ah
14:59 dogbert17 timotimo: I know for a fact that it, at least, worked on This is Rakudo version 2016.03-40-g81558b8 built on MoarVM version 2016.03-46-g50c7f6a
14:59 timotimo the problem appears when we're On-Stack-Replacing (or at least trying to) the count-only method of some class
15:00 [Coke] timeless: I'm catching up. but perl6doc has "make run"
15:00 timotimo since it's about spesh and optimization and such, it can be very finnicky
15:00 tony-o i wrote bench in 2014 so things have had a bit of time to develop since then :)
15:00 timeless [Coke]: oh boy, that sounds like what i want (doh) -- thanks
15:01 [Coke] that requires your p5 to have Mojolicious::Lite
15:01 dogbert17 timotimo: I eun these programs from time to time in order to track the improvements made by jnthn, lizmat et al
15:01 timeless [Coke]: well, hrm,...
15:01 timeless run https://www.irccloud.com/pastebin/iU4NIUly/
15:01 dogbert17 s/eun/run/
15:01 timotimo that's good :)
15:01 timeless no output from curl?? https://www.irccloud.com/pastebin/MlXpjVM7/
15:02 [Coke] we'd all be happy if it used the perl6 mini server, but it wasn't crucial for testing to self host there.
15:02 timeless [Coke]: i'm kinda expecting output from curl...
15:02 dogbert17 timotimo: on a positive note, this latest run was by far the quickes :)
15:02 timotimo the quickest to crash :D
15:02 [Coke] timeless: does it show in chrome or other gui browser?
15:02 * colomon built a small p6 Inline::Python script yesterday, because that seemed easier than learning enough Python to set up a command line interface for what he wanted to do.
15:03 timotimo i like using "curl -v" to figure out what's going on with http requests
15:03 timeless [Coke]: the machine isn't very reachable
15:03 timotimo it shows you outgoing and incoming headers, too
15:03 timeless -v https://www.irccloud.com/pastebin/CjqVf7rh/
15:03 dogbert17 when I ran it in March it took 5m50s on my systen
15:03 timeless i'm trying to decide who to blame here
15:03 timeless the server clearly says it's serving a 302
15:03 timotimo huh, so it's really just giving you an empty document for index.html
15:03 timeless oh
15:03 * timeless wonders if it needs to follow redirects?
15:04 timeless ok, w/ curl -L things work better
15:04 timotimo ooooh
15:05 dogbert17 that sounds ominous :)
15:05 tony-o [Coke]: i've been playing with HTTP::Easy for perl6 - it looks pretty easy, as the name suggests
15:06 [Coke] tony-o: nice.
15:06 tony-o if it's just serving static docs, it looks like it could handle that
15:06 timeless um
15:06 timeless [Thu Jun  9 15:05:38 2016] [debug] GET "/robots.txt.html.html.html.html.html.html.html"
15:06 timeless perhaps that isn't the best outcome from a web server?
15:06 timeless "If i don't have an X, I should suggest an X.html, w/o first checking to see if it exists!"
15:07 [Coke] not just static docs.
15:10 khw joined #perl6
15:11 ssotka joined #perl6
15:13 tadzik uh-oh
15:13 tadzik any idea if Inline::Perl5 is thread-safe? :S
15:14 timotimo only in the same way perl5 itself is thread-safe :P
15:14 timeless [Coke]: would you take this? https://www.irccloud.com/pastebin/WHgOl1g4/
15:14 timeless that should make my robots problem go away...
15:14 timotimo i think you can have a single Inline::Perl5 thingie per thread, but it's essentially the same as having multiple processes
15:14 timotimo i.e. you have to rely on perl6 to marshal the objects between the interpreters
15:15 tadzik I'm getting happy segfaults
15:15 timeless hrm, it doesn't
15:15 mohae joined #perl6
15:16 tadzik "Internal error: Unwound entire stack and missed handler" well that's new :o
15:16 timotimo nah, it's very old
15:16 timotimo that's when you throw an exception inside a nativecall callback and it reaches the C code
15:16 tadzik hm
15:17 timeless [Coke]: ok, this works https://www.irccloud.com/pastebin/ac3E1Adg/
15:17 timotimo if you can't get it to not show up by placing CATCH everywhere, it can be the signature bind for the callback itself doesn't work
15:18 arnsholt Oh, that one, yeah. *sadface*
15:18 maybekoo2 joined #perl6
15:18 timotimo i think we should be able to register a handler for that case in the HLL that prints out the exception with a little hint about callbacks
15:18 timotimo because we should be able to figure out if we're in a sub-interpreter situation
15:19 _mg_ joined #perl6
15:20 timeless watch "grep -B3 'To do' ../checklink.report3.txt |egrep -v 'github|Code:|Lines?:|^--|robots exclusion'"
15:20 tadzik timotimo: apparently it also means "don't 'return' in a start { } block
15:20 tadzik ":)
15:20 timotimo oh
15:20 timotimo that should give a different error
15:20 [Coke] timeless: I'd look at a PR with that. Can't say right now, though.
15:20 timotimo but returning is about to become a whole bunch better :)
15:20 timeless [Coke]: ok, i'll send one in a bit
15:21 timotimo but yeah, a start block doesn't give you a sub, so you'll be returning from the routine that has the start block in it
15:21 tadzik right
15:21 timotimo you can start anon sub() { }, i believe
15:21 timotimo m: await start sub foo() { say "hi" };
15:21 camelia rakudo-moar 2095ed: ( no output )
15:21 timotimo huh.
15:21 timotimo m: start sub foo() { say "hi" }; sleep 1
15:21 camelia rakudo-moar 2095ed: ( no output )
15:22 timotimo heh.
15:22 timotimo that just makes the sub be the return value of the promise
15:22 tadzik anyway, https://gist.github.com/tadzik/add0e94079da1b1645bb8fa78151f25d segfaults my code, basically
15:22 timotimo yeah, that's very expected
15:22 timotimo you need to create an Inline::Perl5 per thread
15:23 timotimo that also means you can't use :from<Perl5>, as you have to address each of the I::P5 objects one-by-one
15:23 tadzik oh bummer
15:24 timeless [Coke]:
15:24 timeless Language/setbagmix.pod:=head3 infix (.)
15:24 timeless Language/setbagmix.pod:Equivalent to L«(.)», at codepoint U+228D (MULTISET MULTIPLICATION).
15:24 timotimo why? it's easy to do and it'll work for your case
15:24 timeless are those two supposed to be related?
15:24 * timeless thinks so
15:24 timotimo yeah, they should
15:25 timeless timotimo: do i change it to  L«infix (.)» or L«(.)|infix (.)»
15:25 tadzik timotimo: yeah, it's doable. My actual code is more complex than this golfed loop though :)
15:25 timeless or something else
15:25 timotimo i'd say have it show (.) but link to infix (.)
15:25 timotimo tadzik: of course
15:26 grondilu joined #perl6
15:26 tadzik I have aplenty of worker objects that just use WWW::Mechanize like it's the most ordinary thing ever
15:26 tadzik so I just need to de-sugarize them
15:26 timeless timotimo: ok, do you know offhand if i need a `#` here?
15:26 tadzik no tragedy :)
15:26 timotimo i do not :(
15:26 * timeless so doesn't understand the pod link syntax
15:26 timotimo i don't even know if you need to spell out the file name, too
15:26 timeless ok, i'll try and see what the build system says
15:31 timeless [Coke]/timotimo: any idea why some links are to `.html` and some aren't? https://www.irccloud.com/pastebin/qNOuWx2S/
15:32 timeless Language/objects.pod:$*IN.L<nl-in> = "\r\n";
15:32 timeless timotimo: i presume that should be C<...>?
15:32 dogbert17 timotimo: valgrind came through but I suspect that you already have all relevant info: https://gist.github.com/dogbert17/bedd7295290dd9490a521970aad85e64
15:32 stmuk_ timeless: maybe mentioned in scrolback already but there are redirects done at the apache level
15:33 timotimo it could be a link to the nl-in method
15:33 timeless timotimo: what's that?
15:34 timotimo well, it's an attribute
15:34 timotimo or rather, the method is the accessor for that
15:40 timeless hrm, ok, i clearly asked an ambiguous question
15:40 timeless "as far as a user is concerned, where would they meaningfully find an explanation of that in doc.perl6.org ?"
15:41 timotimo probably on the IO::Handle class page
15:43 timeless fwiw, L«(.)|infix (.)» seemed to do the right thing.
15:43 timotimo cool!
15:43 azawawi joined #perl6
15:43 azawawi .
15:44 timeless IO::Handle is pretty empty :(
15:44 azawawi Xliff: ping
15:45 stmuk_ which repo has the apache redirects for doc.perl6.org? anyone remember?
15:45 timotimo hmmm
15:45 azawawi Xliff: re How much of GTK is covered in GTK::Simple? ... No idea but we're getting there (full list: https://developer.gnome.org/gtk3/stable/ch03.html)
15:45 moritz stmuk_: iirc there's just a .htacess file somewhere
15:46 moritz hm, doesn't seem to be in the repo :(
15:47 moritz root@www:/var/www/doc.perl6.org# cat .htaccess
15:47 moritz Redirect 301 /language/5to6 http://doc.perl6.org/language/5to6-nutshell
15:47 timeless is that the entire file?
15:47 moritz yessir
15:47 stmuk_ ah I was probably thinking of the main site
15:47 timeless well, um... is it reasonable to ask for it to be added to the doc repo?
15:48 moritz timeless: it totally is
15:48 stmuk_ www.perl6.org has more redirects I think?
15:48 timeless at the very least `make run` should be aware of the lines there
15:48 dalek doc/parallel-htmlify-2: 2dd16d1 | moritz++ | html/.htaccess:
15:48 dalek doc/parallel-htmlify-2: Add .htaccess file for the website
15:48 dalek doc/parallel-htmlify-2: review: https://github.com/perl6/doc/commit/2dd16d196b
15:49 moritz stmuk_: there is irc.perl6.org for eamplw which redirects to http://irclog.perlgeek.de/perl6/today
15:49 moritz should become https, btw :-)
15:49 parabolize timeless: `p6doc -f nl-in` didn't find anything for me `grep -d recures 'method nl-in' ~/.rakudobrew/' had 4 hits
15:50 kurahaupo joined #perl6
15:50 timeless fwiw, here's how I'm watching the link checker https://www.irccloud.com/pastebin/ag9sbL2t/checklink-watcher.sh
15:50 stmuk_ BTW I used a custom google search box on pl6net.org which gets hits for "nl-in"
15:50 timeless parabolize: ok,... what should i do w/ that?
15:50 stmuk_ it indexes most of the obvious sites
15:51 timeless www.pl6net.org’s server DNS address could not be found.
15:51 domidumont joined #perl6
15:51 stmuk_ its pl6anet.org
15:51 stmuk_ typo
15:51 timotimo for some reason i really like that spelling
15:51 timotimo even though it's so strange
15:52 timeless stmuk_: so, i'm still leaning on changing it to C<...>
15:52 timeless none of the pl6anet.org things stand out to me (non-expert hat) as obvious methods
15:53 timeless parabolize: `recures`?
15:53 timeless recurse?
15:53 parabolize yeah, sorry
15:54 parabolize I think its only method definition is in ~/.rakudobrew/moar-nom/src/core/IO/Handle.pm
15:59 Xliff azawawi, neat! Thanks for the info!
15:59 timeless hrm,
15:59 azawawi Xliff: np
15:59 timeless ok,.... so... i guess technically $*IN.nl-in *is* talking about IO::Handle.nl-in
15:59 timeless but even so, i don't think that's what the user is looking for when reading the text that offended the link checker
15:59 timeless so... i'm now 80% confident in changing L<...> to C<...>
15:59 parabolize you talking about the example in https://doc.perl6.org/language/objects#Using_Objects?
15:59 stmuk_ m: say IO::Handle.^methods
15:59 camelia rakudo-moar 2095ed: OUTPUT«(FALLBACK gist new)␤»
15:59 * parabolize was indeed confused by it
15:59 stmuk_ uh
15:59 dogbert17 m: say IO::Handle.^methods(:local)
15:59 camelia rakudo-moar 2095ed: OUTPUT«()␤»
15:59 [Coke] moritz: html/.htaccess already existed.
15:59 dogbert17 hmm
15:59 timotimo i think IO::Handle may just be a role or superclass
15:59 timotimo not sure
15:59 timeless parabolize: yes
16:00 rurban joined #perl6
16:00 moritz [Coke]: oh, I was on the wrong branch
16:00 stmuk_ say IO::Handle.^methods gives me results in my REPL
16:00 timeless (sorry, i had to look it up)
16:00 stmuk_ not from the bot
16:00 moritz dogbert17: IO:: classes are stubbed out in the restricted setting
16:00 stmuk_ s: say IO::Handle.^methods
16:00 stmuk_ star: say IO::Handle.^methods
16:00 camelia star-m 2016.01: OUTPUT«(FALLBACK gist new)␤»
16:00 moritz m: CORE::<IO::Handle>.^methods(:local)
16:00 camelia rakudo-moar 2095ed: ( no output )
16:00 dogbert17 moritz: that explains it :)
16:00 moritz m: say CORE::<IO::Handle>.^methods(:local)
16:00 camelia rakudo-moar 2095ed: OUTPUT«(new Numeric Str sink AT-POS AT-KEY BIND-POS BIND-KEY ASSIGN-POS ASSIGN-KEY STORE push append unshift prepend FALLBACK ords chrs iterator gist)␤»
16:01 moritz ... though as a security measure not very effective :/
16:01 dogbert17 :)
16:01 stmuk_ ah
16:01 dogbert17 cheating - (open nl-in close eof get getc comb words lines read readchars Supply seek tell write opened t lock unlock print put print-nl slurp-rest chmod IO path flush encoding DESTROY e d f s l r w x modified accessed changed watch native-descriptor umask split say Str gist perl chomp nl-out)
16:01 timeless parabolize: so change L<...> to C<...> there?
16:02 timeless presumably one could filter the list for /nl/
16:03 parabolize I don't know. I haven't learned PODs yet.
16:03 stmuk_ m: say IO::Handle.^mro
16:03 camelia rakudo-moar 2095ed: OUTPUT«IO::Handle is disallowed in restricted setting␤  in sub restricted at src/RESTRICTED.setting line 1␤  in method gist at src/RESTRICTED.setting line 33␤  in block <unit> at /tmp/Th9UeNOT2c line 1␤␤»
16:03 timotimo hah, oh crap
16:04 timotimo forgot about the restricted setting stuff
16:04 stmuk_ maybe .^methods should also report an explicit error under RESTRICTED
16:04 timotimo yeah, we could have a RestrictedHOW for that :P
16:04 timotimo actually, we can define a method ^methods
16:07 timeless is NaN a type?
16:08 timotimo it's an instance of Num, i believe
16:08 timotimo m: say NaN.DEFINITE
16:08 camelia rakudo-moar 2095ed: OUTPUT«True␤»
16:08 timotimo m: say NaN.defined
16:08 camelia rakudo-moar 2095ed: OUTPUT«True␤»
16:08 timotimo oh, it's even defined
16:09 stmuk_ not a number *is* a number!
16:09 timeless these things think it's a type https://www.irccloud.com/pastebin/oF5csPz6/
16:09 * azawawi is currently thinking of https://en.wikipedia.org/wiki/Naan  # 45 min left till breaking fast :)
16:09 timeless azawawi: sounds good
16:09 timeless sunset?
16:10 timotimo wellllll, that's just the file name/path, so ... dunno!
16:11 timeless well, you have Type/{Complex,Hash,Int,Numeric,Str}
16:11 timeless sounds like NaN could live in Numeric.pod?
16:11 timeless hrm, no :/
16:12 timotimo maybe because NaN behaves so strangely compared to actual numbers?
16:13 timeless I just need a place to put it to make the link checker (and human readers) vaguely happy
16:14 timeless if someone gets religious they can move it later
16:14 timeless i'm leaning toward Numeric
16:14 rurban_ joined #perl6
16:15 timeless actually
16:15 * timotimo AFK
16:15 timeless where does NaN fit on the pretty graph in https://docs.perl6.org/type/Numeric ?
16:15 azawawi any plans for a Perl 6 book?
16:15 azawawi printed that is
16:15 [Coke] azawawi: yes.
16:16 [Coke] I believe work is in progress, no estimated completion date at the moment.
16:16 azawawi any samples? links?
16:16 timeless m: NaN.unival
16:16 camelia rakudo-moar 2095ed: OUTPUT«Cannot call unival(Num); none of these signatures match:␤    (Str:D $str)␤    (Int:D $code)␤  in block <unit> at /tmp/YKEgASt2WO line 1␤␤»
16:16 timeless ok, so NaN is not a Num? :)
16:17 [Coke] azawawi: nope
16:17 stmuk_ m: NaN.WHAT
16:17 camelia rakudo-moar 2095ed: ( no output )
16:17 [Coke] it's not an open source book.
16:17 azawawi [Coke]: cool thx... who's working on it?
16:17 [Coke] And my info on it is months old, I'm afraid
16:18 timeless m: 5.WHAT
16:18 camelia rakudo-moar 2095ed: ( no output )
16:18 [Coke] I'm sure they'll speak up if they want that info out there. :)
16:18 stmuk_ I think the obvious author popped in a few weeks ago and mentioned it was at least started
16:18 * timeless is totally lost
16:18 stmuk_ but didn't give details
16:19 azawawi [Coke]: yup
16:19 timeless so.... what is NaN? :-/
16:19 perlpilot Not a Number
16:19 perlpilot as usual
16:19 stmuk_ (Num)
16:19 stmuk_ is a number and not one
16:20 timeless stmuk_: so, dump NaN in Num.pod?
16:20 perlpilot stmuk_: It's a "number" that represents notness. :)
16:20 timeless but, if NaN is a Num, then why did NaN.unival fail?
16:20 stmuk_ timeless: I don't know .. TBH it probably doesn't need documentation since it's well known outside the Perl 6 community anyway
16:21 timeless stmuk_: `%` is known outside the C community, but it definitely needs documentation for C ...
16:21 timeless (what is -3 % -2  [C] ?)
16:21 perlpilot timeless: depends on the implementation of C  ;)
16:21 timeless perlpilot is correct.
16:22 timeless (another fun one: `what is/are the allowed prototype[s] for ``main``? [C]`
16:23 perlpilot timeless: you really could go on all day with this stuff for C.  But, please don't :)
16:24 timeless fair enough. but i hope the point is understood, just because a concept exists elsewhere doesn't mean you shouldn't define it for your language
16:24 timeless especially when it's flaky
16:24 timeless and `%`, `NaN` and `main()` all count as flaky
16:24 * timeless dumps it into Num.pod
16:25 timeless odd
16:25 timeless ok, i don't get it
16:25 perlpilot The only problem with % is that people conflate "modulus" and "remainder" (and unfortunately some of those people are compiler writers)
16:26 * timeless chuckles
16:26 timeless yep
16:26 rurban1 joined #perl6
16:30 tharkun joined #perl6
16:34 TimToady timeless: you seem to have a strange idea of what .unival is supposed to do
16:34 timeless TimToady: i have absolutely no idea what it should do :)
16:35 TimToady m: say "⅓".unival
16:35 camelia rakudo-moar 2095ed: OUTPUT«0.333333␤»
16:35 timeless but the error says cannot call unival(Num) which to me implies that unival(Num) exists
16:35 timeless it might be the case that it means "unival() does not accept Num"
16:35 TimToady no, it implies that it *doesn't* exist
16:36 timeless ok. well... that's my problem, the error confused me
16:36 TimToady m: say unival(now)
16:36 camelia rakudo-moar 2095ed: OUTPUT«Cannot call unival(Instant); none of these signatures match:␤    (Str:D $str)␤    (Int:D $code)␤  in block <unit> at /tmp/ULJQ74FbVM line 1␤␤»
16:36 timeless I was hoping to use the call to determine the type of NaN, and apparently it would have worked if i understood the error :/
16:37 TimToady m: say NaN.WHAT
16:37 camelia rakudo-moar 2095ed: OUTPUT«(Num)␤»
16:37 TimToady okay, I think I see what was confusing
16:38 rurban joined #perl6
16:38 TimToady when it says "Cannot call unival(Num)" it's trying to indicate generically what the types of the call were, not that unival(Num) actually exists
16:38 setty1 joined #perl6
16:38 timeless i think replacing `call` with something else might help, maybe
16:39 timeless tentatively `apply`
16:39 timeless (acknowledging that `apply` may be overused/otherwise defined)
16:39 timeless possibly `resolve`
16:40 TimToady 'do anything with' :)
16:40 timeless sure?
16:40 TimToady 'figure out' :)
16:40 timotimo "frumble"
16:41 timeless anyway, if you're taking this UX as input to considering changing the Output, I appreciate it :)
16:41 * timeless goes back to not understanding POD
16:41 TimToady :D
16:42 timeless All I want to do is add `NaN` to Num.pod, now that I've established that's where it belongs
16:42 timeless but...
16:42 timeless this did not work at all https://www.irccloud.com/pastebin/GTxrv0ub/
16:43 timeless now, NaN is an Instance of Num, and not a subclass, right?
16:43 timeless m: say NaN.new
16:43 camelia rakudo-moar 2095ed: OUTPUT«0␤»
16:43 timeless m: say NaN === NaN
16:43 camelia rakudo-moar 2095ed: OUTPUT«True␤»
16:43 timeless m: say NaN === NaN.new
16:43 camelia rakudo-moar 2095ed: OUTPUT«False␤»
16:44 TimToady yes, it's just a special value
16:45 TimToady NaN.new is the same as 42e0.new
16:45 TimToady new generally ignores the actual value and just uses the type
16:46 timeless ok, I think:
16:46 timeless =head1 Constants
16:46 timeless =head2 X<C<NaN>|Nan>
16:46 timeless C<NaN> is a C<Num>.
16:46 timeless maybe?
16:51 TimToady maybe something like NaN is a special Num value that normally results from supplying nonsensical inputs to afloating-point operation.
16:53 timeless the fancy search box on http://doc.perl6.org/ doesn't work well for no-results when your screen is narrow (half width)
16:53 timeless can't see search box https://usercontent.irccloud-cdn.com/file/aB3Lq76u/
16:54 parabolize well, its there but to the right for some reason
16:55 timeless it's behind the red box
16:55 timeless (below)
16:55 timeless (z-order)
16:56 * timeless frowns
16:58 parabolize firefox isn't doing that to me but the error message is bound to the green div.menu thing
16:58 * timeless is using Chrome
16:58 timeless did you narrow your window to approximate my window's width?
16:59 timeless ok, so, i don't understand how Type/Num.pod works
16:59 parabolize so if you make it ridiculously small firefox will cover the search bar
16:59 parabolize it works fine at the width you had it
17:00 parabolize but if a phone did what firefox did I wouldn't be happy
17:00 * timeless launches ff for android
17:01 timeless it does
17:01 timeless want a picture?
17:02 * timeless waits for irccloud to load in chrome for android
17:02 * timeless should just use imgur.com ...
17:03 parabolize firefox fo android isn't really firefox
17:03 parabolize *for
17:03 timeless m.imgur.com/RX49zLj
17:05 timeless ok... so, back to `Num.pod`
17:06 timeless 3 lines above `Type graph`are contributed by the file itself
17:06 timeless everything else seems to be magically inserted
17:07 yqt joined #perl6
17:07 timotimo .u square
17:07 yoleaux U+005B LEFT SQUARE BRACKET [Ps] ([)
17:07 yoleaux U+005D RIGHT SQUARE BRACKET [Pe] (])
17:07 yoleaux U+033B COMBINING SQUARE BELOW [Mn] (◌̻)
17:07 timotimo :\
17:13 timeless could i claim` enum NaN is Num` ?
17:14 timeless ignoring that fact that it isn't actually an enum, would that get me close enough to being in the type-graph that something useful would happen?
17:18 chris2 joined #perl6
17:19 _mg_ joined #perl6
17:19 Xliff FROGGS: Ping
17:19 timeless parabolize: https://github.com/perl6/doc/issues/574
17:20 TimToady timeless++ for suggesting https://github.com/rakudo/rakudo/commit/b6902e5192
17:20 timeless parabolize: and fwiw, the next version of firefox for desktop will finally have e11s which will make them much closer
17:20 timeless TimToady++ for writing a patch!
17:21 TimToady funny thing is, I tried it with just "resolve" and it failed the spec tests, which require the word "call" to be in there somewhere :)
17:21 timeless really, much appreciated. that's what i love about collaborative systems
17:21 timeless lol
17:22 TimToady so really a bit of an overspecific roast buglet
17:22 TimToady but I worked around it
17:23 cdg joined #perl6
17:24 timeless so... I can't figure out how to get Num.pod to list my own content
17:24 timeless the best I can do is get NaN to appear in the graph as a child of Num
17:24 timeless and then just write a NaN.pod
17:24 * TimToady has not played with that at all
17:24 timeless I'm going to go w/ that approach, and if someone objects, they can rewrite it
17:28 Xliff Can you create multi methods that respect the aka trait?
17:30 Xliff m: class A { multi method unreal_name($a, $b) is aka<better_name> { self.unreal_name($a => $b); } multi method unreal_name(Pair $kv) is aka<better_name> { dd $kv } }; A.new.better_name(1, 2);
17:30 camelia rakudo-moar 2095ed: OUTPUT«5===SORRY!5=== Error while compiling /tmp/5XogxaaAmH␤Can't use unknown trait 'is aka' in a method declaration.␤at /tmp/5XogxaaAmH:1␤    expecting any of:␤        rw raw hidden-from-backtrace hidden-from-USAGE␤        pure default DEPRECATED i…»
17:32 beatdown joined #perl6
17:33 ribasushi joined #perl6
17:34 patrickz joined #perl6
17:34 timeless m: NaN.roots
17:34 camelia rakudo-moar b6902e: OUTPUT«Too few positionals passed; expected 2 arguments but got 1␤  in block <unit> at /tmp/ZqQLmad5Gd line 1␤␤»
17:34 timeless m: NaN.roots(1)
17:34 camelia rakudo-moar b6902e: ( no output )
17:35 muraiki what's the perl6 version of perl5 @INC?
17:35 timeless m: NaN.rindex("This is NaN or something")
17:35 camelia rakudo-moar b6902e: ( no output )
17:35 timeless m: NaN.rindex("This is NaN or something", 0)
17:35 camelia rakudo-moar b6902e: ( no output )
17:35 muraiki I couldn't find info on it in 5to6-nutshell
17:36 muraiki it only listed the "use inc" pragma
17:36 * timeless saw something which hinted at not needing to do it
17:36 * timeless didn't ask "why"
17:37 timeless NaN.abs
17:37 timeless m: NaN.abs
17:37 camelia rakudo-moar b6902e: ( no output )
17:37 timeless m: (-2).abs
17:37 camelia rakudo-moar b6902e: ( no output )
17:37 timeless m: say NaN.abs
17:37 camelia rakudo-moar b6902e: OUTPUT«NaN␤»
17:37 timeless m: say NaN.rindex("This is NaN or something", 0)
17:37 camelia rakudo-moar b6902e: OUTPUT«Nil␤»
17:38 timeless m: say NaN.roots(1)
17:38 camelia rakudo-moar b6902e: OUTPUT«NaN+0i␤»
17:38 raiph joined #perl6
17:38 timeless m: say NaN.conj
17:38 camelia rakudo-moar b6902e: OUTPUT«NaN␤»
17:40 timeless m: say NaN.sign
17:40 camelia rakudo-moar b6902e: OUTPUT«NaN␤»
17:40 timeless ok, i like that... since NUMERIC.sign says the possible values are 1, -1, and 0 ...
17:44 timeless m: say 'a'.unival
17:44 camelia rakudo-moar b6902e: OUTPUT«NaN␤»
17:52 timeless is there a replacement for http://www.perlito.org/perlito/perlito5to6.html
17:52 cdg joined #perl6
17:53 timeless http://fglock.github.io/Perlito/perlito/perlito5.html i guess
17:57 * Woodi was just hit by enlightenment - OSes are just compilers! ;)
17:57 Woodi "The first step is to build a new and host-independent toolchain (compiler, assembler, linker, libraries, and a few useful utilities)..."   http://www.linuxfromscratch.org/lfs/view/development/chapter05/introduction.html
17:58 dalek doc: df64b5c | (Zoffix Znet)++ | html/css/style.css:
17:58 dalek doc: Do not cover search field with "no results" on narrow screens
17:58 dalek doc:
17:58 dalek doc: Closes #574
17:58 dalek doc: review: https://github.com/perl6/doc/commit/df64b5c7a0
18:02 [Coke] IWBNI there was an indicator on the docs.perl6.org site about which commit it was built from.
18:02 [Coke] so I might know when, for example, df64b5c was running live.
18:04 timeless joined #perl6
18:04 muraiki left #perl6
18:05 ribasushi joined #perl6
18:06 sno joined #perl6
18:07 Score_Under joined #perl6
18:12 labster joined #perl6
18:24 _mg__ joined #perl6
18:24 llfourn is Hash.keys determinisitic in rakudo/moar?
18:25 moritz m: my %h = ('a' ... 'z') X=> 1; say %h.keys
18:25 camelia rakudo-moar b6902e: OUTPUT«(x p k h g z a y v s q e d m f c w o n u t b j i r l)␤»
18:25 moritz m: my %h = ('a' ... 'z') X=> 1; say %h.keys
18:25 camelia rakudo-moar b6902e: OUTPUT«(x p k h g z a y v s q e d m f c w o n u t b j i r l)␤»
18:25 moritz m: my %h = ('a' ... 'z') X=> 1; say %h.keys
18:25 camelia rakudo-moar b6902e: OUTPUT«(x p k h g z a y v s q e d m f c w o n u t b j i r l)␤»
18:25 moritz seems so, but you can't rely on that
18:25 buharin joined #perl6
18:26 llfourn hmmm yeah, my program has recently started to fail non-deterministically
18:26 * timeless heads out
18:27 [Coke] you using any async/concurrent code?
18:27 llfourn nope
18:27 Zero_Dogg joined #perl6
18:28 jdv79 i had a program stall for 7m the other day but i'm not sure it had anything to do with rakudo
18:28 jdv79 it uses IO::Socket::Async
18:30 llfourn I don't have any IO Asyn or not :S
18:31 llfourn well I have O
18:31 mr-foobar joined #perl6
18:31 llfourn and I can see the difference in the O before it fails. Things happen in a different order.
18:32 llfourn but looking closer the order isn't even dervived from a hash.
18:34 llfourn ahhh yes it is
18:34 llfourn but it's not Hash.keys it's Hash.values which *isn't* non-deterministic
18:35 llfourn m: my %h = "a".."z" Z=> (^100).pick xx *; say %h.values
18:35 camelia rakudo-moar b6902e: OUTPUT«(83 19 30 52 62 1 90 97 74 82 4 62 16 11 53 57 5 80 40 71 67 56 48 19 90 66)␤»
18:35 llfourn m: my %h = "a".."z" Z=> (^100).pick xx *; say %h.values
18:35 camelia rakudo-moar b6902e: OUTPUT«(4 22 33 31 42 89 36 31 48 27 47 75 90 94 28 64 76 4 3 79 31 43 1 53 50 33)␤»
18:35 llfourn which *isn't* deterministic rather
18:37 nowan joined #perl6
18:39 llfourn oh wait I'm an idiot... pick is introducing the randomness
18:39 llfourn m: my %h = "a".."z" Z=> (^100); say %h.values
18:39 camelia rakudo-moar b6902e: OUTPUT«(23 15 10 7 6 25 0 24 21 18 16 4 3 12 5 2 22 14 13 20 19 1 9 8 17 11)␤»
18:39 llfourn m: my %h = "a".."z" Z=> (^100); say %h.values
18:39 camelia rakudo-moar b6902e: OUTPUT«(23 15 10 7 6 25 0 24 21 18 16 4 3 12 5 2 22 14 13 20 19 1 9 8 17 11)␤»
18:39 * llfourn needs to take a break
18:41 dalek doc: 147a284 | moritz++ | / (2 files):
18:41 dalek doc: htmlify: include commit hash in the footer.
18:41 dalek doc: review: https://github.com/perl6/doc/commit/147a284c54
18:41 moritz [Coke]: ^^ there ya go
18:42 girafe joined #perl6
18:44 perlpilot moritz++  (I was just about to do that :)
18:45 cdg joined #perl6
18:48 moritz perlpilot: I was faster :-)
18:52 k-man joined #perl6
18:53 ribasushi joined #perl6
18:53 hcit joined #perl6
18:54 huggable joined #perl6
18:57 ZoffixW joined #perl6
18:57 ZoffixW Does docs.perl6.org actually run on a Mojolicious app? Wasn't there some sort of mowyw thing happening? https://github.com/perl6/doc#building-the-html-documentation
19:03 ZoffixW I guess a better question is what's the purpose of this file: https://github.com/perl6/doc/blob/master/app.pl
19:04 ZoffixW It can be replaced in entirety with a couple of Apache redirects, but if it is in fact in use, I think we have larger possibilities. Like fixing this bug (spurious 404s): https://github.com/perl6/doc/issues/155
19:07 dalek doc: 569472f | (Zoffix Znet)++ | README.md:
19:07 dalek doc: Linkify usefull things
19:07 dalek doc: review: https://github.com/perl6/doc/commit/569472fad2
19:07 dalek doc: d92e30d | (Zoffix Znet)++ | README.md:
19:07 dalek doc: Add missing space
19:07 dalek doc: review: https://github.com/perl6/doc/commit/d92e30d4b3
19:08 [Coke] ZoffixW: it's for running the app locally
19:08 ZoffixW Ah. Thanks.
19:11 yqt joined #perl6
19:11 ocbtec joined #perl6
19:12 dalek doc: cc8577b | (Zoffix Znet)++ | README.md:
19:12 dalek doc: Fix incorrect markdown link syntax
19:12 dalek doc: review: https://github.com/perl6/doc/commit/cc8577b5a7
19:13 dalek doc: 3dccf05 | (Josh Soref)++ | doc/ (6 files):
19:13 dalek doc: clean up various doc links
19:13 dalek doc: review: https://github.com/perl6/doc/commit/3dccf056da
19:13 dalek doc: f993419 | (Aleks-Daniel Jakimenko-Aleksejev)++ | doc/ (6 files):
19:13 dalek doc: Merge pull request #572 from jsoref/docs
19:13 dalek doc:
19:13 dalek doc: clean up various doc links
19:13 moritz ZoffixW: mowyw is used for www.perl6.org
19:13 dalek doc: review: https://github.com/perl6/doc/commit/f99341943c
19:13 ptolemarch joined #perl6
19:13 ZoffixW Ah, thanks. I thought I was going insane for a second :)
19:14 moritz we all do, no worries :-)
19:15 dalek doc: 7f15517 | (Zoffix Znet)++ | README.md:
19:15 dalek doc: Fix broken link
19:15 dalek doc: review: https://github.com/perl6/doc/commit/7f15517d25
19:15 jjido joined #perl6
19:16 FROGGS joined #perl6
19:18 properly joined #perl6
19:18 itaipu joined #perl6
19:20 FROGGS o7
19:20 ZoffixW \o
19:22 AlexDaniel joined #perl6
19:22 AlexDaniel timeless: hi
19:23 AlexDaniel timeless: I have merged/closed some of your pull request
19:23 spider-mario joined #perl6
19:23 AlexDaniel somebody had to do it anyway :)
19:23 AlexDaniel huggable: dunno
19:23 huggable AlexDaniel, ¯\_(ツ)_/¯
19:23 AlexDaniel timeless: if I did something wrong please bite me
19:23 AlexDaniel requests*
19:27 perlpilot joined #perl6
19:36 cognominal joined #perl6
19:40 rurban joined #perl6
19:42 dogbert17 AlexDaniel: can your bisectbot do anything woth this gist? https://gist.github.com/dogbert17​/d7822b6db5daa88a372c835727bf0729
19:43 AlexDaniel dogbert17: hmm…
19:44 dogbert17 AlexDaniel: I'm 100% certain that it worked with - 2016.03-40-g81558b8 built on MoarVM version 2016.03-46-g50c7f6a
19:47 AlexDaniel dogbert17: well, first of all I'm trying to replicate that here
19:47 AlexDaniel dogbert17: no luck so far
19:48 dogbert17 AlexDaniel: timotimo succeeded
19:48 AlexDaniel oh really
19:48 AlexDaniel is there any way to speed it up?
19:49 AlexDaniel e.g. start from some high number, not 1
19:49 dogbert17 AlexDaniel: on my machine it crashed after 55 sec, perhaps it is possible to gold it down a bit
19:50 AlexDaniel dogbert17: and if it doesn't crash, then what happens?
19:50 dogbert17 s/gold/golf
19:50 AlexDaniel dogbert17: there's no time limit on bisectbot, so…
19:51 dogbert17 AlexDaniel: the program is my attempt to solve https://projecteuler.net/problem=12
19:53 AlexDaniel dogbert17: finished here just fine on 2016.05
19:53 timotimo i succeeded what?
19:55 dogbert17 timotimo: you reproduced the SEGV
19:56 AlexDaniel bisect: my $numFactors = 500;my$sum=0;for (1...*) -> $term {$sum+=$term;my$factors=(1..floor(sqrt($sum))).grep(-> $x { $sum % $x == 0} ).elems * 2;$factors-- if floor(sqrt($sum)) == sqrt($sum);if $factors > $numFactors {say "$sum";last;}}
19:56 AlexDaniel that's not going to work :)
19:56 AlexDaniel my bet is that it is going to ping out
19:56 jjido joined #perl6
19:57 dogbert17 AlexDaniel: trying to make it crash faster
19:58 AlexDaniel dogbert17: if you find no way to make it crash faster then I'll just bisect it manually with existing builds and that will be it
19:59 AlexDaniel yeah, that was expected
20:00 timotimo oh, yeah, sorry, i didn't follow it up muhc more yet
20:00 cdg joined #perl6
20:00 dogbert17 timotimo: np, was just asking AlexDaniel if the bisectbot could provide more info
20:00 bisectable joined #perl6
20:00 AlexDaniel bisectable: poor creature
20:01 zakharyas joined #perl6
20:01 dogbert17 AlexDaniel: interesting that it didn't crash for you
20:02 timotimo as said, it's optimization and OSR related
20:02 timotimo that makes it finnicky
20:03 dogbert17 timotimo: perhaps it's a recent change
20:04 mohae joined #perl6
20:05 _mg_ joined #perl6
20:06 AlexDaniel people have no problems with womble huh :)
20:06 AlexDaniel no issue then
20:08 AlexDaniel dogbert17: bisectable runs on another machine, but I cannot reproduce it there too
20:08 AlexDaniel dogbert17: 6bb23a5 is fine
20:09 dogbert17 AlexDaniel: how old is that?
20:09 AlexDaniel dogbert17: pretty old, 7 days ago
20:10 dogbert17 I can now get it to crash in 31 secs instead of 55
20:10 dogbert17 I'm running on an old Core I5 660
20:11 AlexDaniel dogbert17: that's an improvement for sure. But you'll probably have to bisect it yourself, hmm
20:19 AlexDaniel dogbert17: oh, ok, I have something
20:19 dogbert17 AlexDaniel: what do you have?
20:19 AlexDaniel dogbert17: a segmentation fault!
20:19 dogbert17 yay
20:20 AlexDaniel dogbert17: either it flaps or it was actually “added” in the last 7 days
20:22 dogbert17 AlexDaniel: very interesting
20:23 timotimo the crash is call stack or invocation related
20:23 timotimo jnthn just recently changed stuff about that
20:29 dogbert17 ok, I have it down to 3.5 sec now
20:29 Praise- joined #perl6
20:29 Praise- joined #perl6
20:29 AlexDaniel dogbert17: oh really? Just slam it into bisectbot
20:30 dogbert17 AlexDaniel: is it bisectbot: program
20:30 AlexDaniel dogbert17: bisect: code
20:31 dogbert17 bisect: my $sum = 10082800;for (4601...4900) -> $term {   $sum += $term;   my $sq = floor(sqrt($sum));   my $factors = (1..$sq).grep(-> $x { $sum % $x == 0} ).elems * 2; }
20:32 bisectable dogbert17: on both starting points the exit code is 0 and the output is identical as well
20:32 mohae joined #perl6
20:32 dogbert17 will this work?
20:36 AlexDaniel dogbert17: well, my manual git bisect pointed me to this https://github.com/rakudo/rakudo/commit/65e1b54c9d751b5d85ce650369391a80baf8113a
20:36 AlexDaniel but uh…
20:36 lizmat I broke stuff ?
20:36 AlexDaniel lizmat: wait, too early
20:37 lizmat well, looking at who has done commits in the past days, there's a good chance it's my fault  :-)
20:37 AlexDaniel bisect: good=6bb23a my $sum = 10082800;for (4601...4900) -> $term {   $sum += $term;   my $sq = floor(sqrt($sum));   my $factors = (1..$sq).grep(-> $x { $sum % $x == 0} ).elems * 2; }
20:37 dogbert17 m: my $sum = 10082800;for (4601...4900) -> $term {   $sum += $term;   my $sq = floor(sqrt($sum));   my $factors = (1..$sq).grep(-> $x { $sum % $x == 0} ).elems * 2; }
20:37 bisectable AlexDaniel: on both starting points the exit code is 0 and the output is identical as well
20:37 camelia rakudo-moar b6902e: OUTPUT«(signal SEGV)»
20:38 AlexDaniel star: my $sum = 10082800;for (4601...4900) -> $term {   $sum += $term;   my $sq = floor(sqrt($sum));   my $factors = (1..$sq).grep(-> $x { $sum % $x == 0} ).elems * 2; }
20:38 camelia star-m 2016.01: ( no output )
20:38 dogbert17 star is old
20:38 AlexDaniel bisect: good=6bb23a my $sum = 10082800;for (4601...4900) -> $term {   $sum += $term;   my $sq = floor(sqrt($sum));   my $factors = (1..$sq).grep(-> $x { $sum % $x == 0} ).elems * 2; }; say ‘hi’
20:38 sherwool joined #perl6
20:38 bisectable AlexDaniel: exit code is 0 on both starting points, bisecting by using the output
20:38 bisectable AlexDaniel: (2016-06-09) https://github.com/rakudo/rakudo/commit/65e1b54
20:38 AlexDaniel lizmat: ↑
20:39 dogbert17 oops, you got something?
20:39 AlexDaniel lizmat: I think that… yeah…
20:39 AlexDaniel dogbert17: same commit that I've found manually
20:39 dogbert17 AlexDaniel: cool
20:39 AlexDaniel bisect: my $sum = 10082800;for (4601...4900) -> $term {   $sum += $term;   my $sq = floor(sqrt($sum));   my $factors = (1..$sq).grep(-> $x { $sum % $x == 0} ).elems * 2; }; say ‘hi’
20:39 lizmat dogbert17 AlexDaniel and what did I break exactly ?
20:40 bisectable AlexDaniel: exit code is 0 on both starting points, bisecting by using the output
20:40 bisectable AlexDaniel: (2016-06-09) https://github.com/rakudo/rakudo/commit/65e1b54
20:40 lizmat because the spectests have been clean all the time for all the commits I've done
20:40 AlexDaniel lizmat: the code above
20:40 AlexDaniel lizmat: try to run it, it segfaults
20:40 AlexDaniel lizmat: I'm not saying that it's your fault, but it happened after your commit :)
20:41 lizmat segfault confirmed here
20:41 AlexDaniel lizmat: it is also confirmed by timotimo, camelia and bisectbot
20:43 AlexDaniel yep, does not segfault on 42f07cd, segfaults on 65e1b54
20:43 timotimo well, as i said, it's about optimization. in this case, the dynamic optimization we have in moar
20:43 lizmat I wonder whether the use of Nil in a nqp::until could be the trigger
20:44 lizmat oops, looking at wrong method  :-)
20:44 timotimo there's a loop in some iterator method somewhere that gets on-stack-replaced and when looking at the method it's in finds out it doesn't have a callsite, i.e. a null pointer at that place
20:44 timotimo and that shouldn't ever have a null pointer there, jnthn pointed out to me
20:44 lizmat ok, if I change the int $i in count-only to an Int, it doesn't segfault
20:45 * AlexDaniel is more worried about the fact that segfaults are not catched by bisectbot unless you have some output…
20:46 dogbert17 AlexDaniel: this 'good' command, does it enable you to start from a known good commit thereby speeding up the bisect prpcess?
20:46 AlexDaniel dogbert17: yes. It defaults to 2015.12
20:46 AlexDaniel dogbert17: bad defaults to HEAD
20:47 AlexDaniel bisect: good=HEAD~150 my $sum = 10082800;for (4601...4900) -> $term {   $sum += $term;   my $sq = floor(sqrt($sum));   my $factors = (1..$sq).grep(-> $x { $sum % $x == 0} ).elems * 2; }; say ‘hi’
20:47 lizmat timotimo: should I revert the change in count-only, or keep it ?
20:47 bisectable AlexDaniel: exit code is 0 on both starting points, bisecting by using the output
20:47 bisectable AlexDaniel: (2016-06-09) https://github.com/rakudo/rakudo/commit/65e1b54
20:47 timotimo oof
20:48 timotimo but yeah, count-only is the method in which it b0rks
20:48 timotimo there's an underlying problem that wants to be fixed at the moar level
20:49 AlexDaniel can somebody submit a bug report about this?
20:49 dogbert17 AlexDaniel: i believe that perlpilot wondered if the bisectbot can handle gists
20:49 timotimo yeah
20:49 timotimo well, someone did ask that
20:49 AlexDaniel well, then it's probably better to integrate it into camelia
20:50 AlexDaniel because we are starting to see some duplicated functionality :)
20:50 AlexDaniel people voted against that very early, claiming that these bots should probably be separate
20:50 AlexDaniel but if you want gists now… then I don't know :)
20:51 dogbert17 AlexDaniel++ for bisectbot
20:52 rurban1 joined #perl6
20:53 regreg joined #perl6
20:53 regreg is there an imap library for perl6?
20:54 lizmat regreg: no idea, check modules.perl6.org  :-)
20:54 [Coke] https://github.com/retupmoca/P6-Net-IMAP
20:54 [Coke] ... which you can find by doing what lizmat said.
20:54 regreg modules.perl6.org?
20:54 regreg okkkk
20:54 AlexDaniel yes, modules.perl6.org :)
20:54 regreg P6-Net-IMAP
20:54 regreg bingo!
20:55 lizmat timotimo dogbert17 regarding the segfault, it looks like replacing the nqp::eqaddr() back with infix:<=:=> makes the segfault go away
20:55 regreg basically i need to download about 115k email headers and delete the emails from the server
20:55 regreg (a datamining project)
20:56 jjido joined #perl6
20:56 dogbert17 lizmat++
20:58 Praise joined #perl6
20:58 Praise joined #perl6
20:58 lizmat do we have a ticket number for the segfault ?
20:59 dogbert17 lizmat: not from me I'm afraid all my attempts (2) to report something to RT has been dismal failures, don't know why though
21:00 lizmat have you tried sending an email to rakudobug@perl.org ?
21:01 dogbert17 lizmat: that is exactly what I have done
21:01 lizmat hmmmm...
21:01 lizmat lemme try then
21:01 regreg how do i use a library from github?
21:01 dogbert17 lizmat: thank you
21:01 regreg i download and unpack it into a directory, then?
21:02 regreg sorry i'm a noob
21:02 dogbert17 regreg: do you have a program called panda installed on your machine?
21:03 sufrostico regreg: https://github.com/tadzik/panda/
21:03 regreg dogbert17,  a sec i'll do an apt-file search panda
21:03 AlexDaniel no panda in debian yet, as far as I remember
21:03 regreg ouch
21:04 AlexDaniel regreg: by the way, what version of perl6 are you using?
21:04 regreg This is perl6 version 2015.11 built on MoarVM version 2015.11
21:04 regreg should i download a fresh copy instead?
21:04 AlexDaniel regreg: yea
21:04 lizmat regreg: that would be advisable
21:05 regreg ok
21:05 dogbert17 regreg: http://perl6.org/downloads/
21:05 AlexDaniel regreg: sometimes we have people here trying to use 2014.07 or something… that's completely wrong :)
21:05 AlexDaniel regreg: 2015.11 is not too bad but it's still pre-v6.c
21:05 regreg should i first remove the rakudo package from debian?
21:05 AlexDaniel yea
21:06 AlexDaniel regreg: by the way, 2016.04 is in debian testing
21:06 lizmat https://rt.perl.org/Ticket/Display.html?id=128357   # we haz a number
21:06 AlexDaniel lizmat++
21:06 regreg this one: http://rakudo.org/downloads/star/rakudo-star-2016.04.tar.gz?
21:06 dogbert17 lizmat++
21:07 dogbert17 regreg: should work quite nicely I think
21:08 * regreg compiling
21:09 AlexDaniel regreg: another option is to use rakudobrew
21:12 * regreg running tests
21:12 AlexDaniel .tell perlpilot right now it does not handle gists and the way it handles long-running programs is non-ideal. But you can send your stuff to me and I'll run it in the terminal
21:12 yoleaux AlexDaniel: I'll pass your message to perlpilot.
21:13 lizmat AlexDaniel timotimo dogbert17 https://github.com/rakudo/rakudo/commit/41736009bf fixes the segfault (for now)
21:14 dogbert17 lizmat++ again :)
21:15 lizmat yw  :-)   I hope someone will be able to trace the underlying issue that causes the segfault
21:15 lizmat dogbert17: but you should be able to continue now
21:15 dogbert17 lizmat: more project Euler then :)
21:15 lizmat :-)
21:16 dogbert17 I have solved a bunch of them and I sometimes rerun them in order to see performance improvements
21:17 AlexDaniel dogbert17: er, do you realize that this commit was made just 13 hours ago?
21:18 AlexDaniel like… how often do you rerun them? :)
21:18 dogbert17 AlexDaniel: actually I'm ill atm, fighting an ear infection and I had nothing better to do :)
21:19 AlexDaniel dogbert17: I'm ill too :(
21:20 dogbert17 In fact I'll rerun the previously segfaulting code, in march it took 5m50s on my machine, it should be a bit faster now
21:20 * lizmat hopes so
21:20 dogbert17 lets see
21:21 firstdayonthejob joined #perl6
21:21 itaipu joined #perl6
21:25 dogbert17 m: my $sum = 10082800;for (4601...4900) -> $term {   $sum += $term;   my $sq = floor(sqrt($sum));   my $factors = (1..$sq).grep(-> $x { $sum % $x == 0} ).elems * 2; }
21:25 camelia rakudo-moar 8cfa6c: OUTPUT«(signal SEGV)»
21:25 dogbert17 so when will camelia be updated?
21:25 lizmat no idea
21:25 lizmat moritz might know
21:26 AlexDaniel dogbert17: patience, my friend, patience
21:26 dogbert17 I made a rakudobrew bild moar and got This is Rakudo version 2016.05-81-g8cfa6c7 built on MoarVM version 2016.05-17-g6075599, hmm
21:27 * dogbert17 tries again
21:27 AlexDaniel dogbert17: yea, you need g4173600, not g8cfa6c7
21:27 dogbert17 must have been too quick
21:29 regreg AlexDaniel, testing while doing make install is normal? it takes forever to install, much longer than compilation time
21:33 rurban joined #perl6
21:33 AlexDaniel regreg: ¯\_(ツ)_/¯
21:33 AlexDaniel regreg: I've never noticed
21:40 dogbert17 lizmat: everything works and execution time is down to 4m14s from 5m50s in March
21:40 regreg Rakudo Star has been built and installed successfully.
21:40 regreg AlexDaniel,  finally :)
21:41 lizmat m: say 254 / 350
21:41 camelia rakudo-moar 8887a7: OUTPUT«0.725714␤»
21:41 lizmat not bad  :-)
21:41 dogbert17 indeed :-)
21:42 dogbert17 regreg: now you should have the mosule installer panda
21:42 dogbert17 s/mosule/module/
21:42 regreg dogbert17, how do i test if i have it enabled?
21:42 regreg there is no panda binary in rakudo/bin
21:43 dogbert17 regreg: what happens if you type panda on the commandline?
21:43 regreg command not found
21:43 dogbert17 that's strange
21:44 regreg perl6 works ok
21:46 regreg ah
21:46 regreg i've found panda
21:46 dogbert17 regreg: cool
21:46 regreg it's in share/site/bin
21:46 regreg more exacly ~/opt/rakudo-star-2016.04/share/perl6/site/bin
21:47 dogbert17 regreg: panda install <the-module-you-want>
21:47 regreg ok, doing it..
21:48 regreg *resolve stage failed for P6-Net-IMAP: Project P6-Net-IMAP not found in the ecosystem. Maybe try "panda update" first?)
21:48 regreg not found
21:49 regreg panda install Net::IMAP, this one should probably work
21:49 dogbert17 regreg: probably :)
21:49 regreg because it's Net::IMAP not  P6-Net-IMAP
21:50 dogbert17 I think you can write panda list to see all modules
21:50 regreg great
21:50 lizmat good night, #perl6!
21:50 regreg good night lizmat
21:50 dogbert17 good night lizmat
21:52 regreg http://pastebin.com/Khw9xVVV
21:52 regreg i get a failed test
21:53 regreg ==> Testing Email::MIME
21:53 regreg ===SORRY!===
21:53 rurban1 joined #perl6
21:53 regreg dogbert17, could you please if you are kind to check out that error message in above pastecode?
21:58 dogbert17 regreg: hmm, seems to work for me, full disk?
21:58 regreg 28% in /
21:58 regreg 28% /home
21:59 dogbert17 regreg: I was wrong it failed for me as well :(
22:00 regreg actually there seems to be some progress if i run the command panda install again
22:02 dogbert17 regreg: sounds strange but promising
22:03 TEttinger joined #perl6
22:03 rurban joined #perl6
22:04 regreg i think that the dependecies have been installed, currently testing  ==> Testing Net::IMAP
22:04 cpage_ joined #perl6
22:04 regreg failed
22:05 regreg http://pastebin.com/bTb9Ws9T
22:06 regreg here's the error
22:06 regreg is there any way i can mess with the code before compiling and installing it? i might be able to correct minor bugs
22:06 regreg (assuming they are not too perl6 specific since i'm a perl noob)
22:06 regreg ?
22:07 rurban joined #perl6
22:08 dogbert17 regreg: try   panda --notests install  Net::IMAP
22:08 regreg sounds risky :) trying it..
22:09 regreg Successfully installed Net::IMAP
22:09 dogbert17 regreg: risky indeed :)
22:10 regreg does default ubuntu emacs distribution have support for perl6?
22:11 regreg or what is the recommended editor under xfce?
22:14 dogbert17 atom is what I use atm
22:15 dogbert17 I believe that there is support for vim as well
22:16 regreg i don't like vim :)
22:16 dogbert17 :)
22:16 rurban joined #perl6
22:16 Sgeo_ joined #perl6
22:54 huggable joined #perl6
22:58 araujo joined #perl6
23:00 bjz joined #perl6
23:02 araujo joined #perl6
23:03 jack_rabbit joined #perl6
23:16 raiph joined #perl6
23:16 kid51 joined #perl6
23:18 kurahaupo joined #perl6
23:18 sufrostico Vim Rocks!!!
23:31 * grondilu suddenly wonders if some people have Vim tattoos.  Googles it.
23:32 geekosaur this strikes me as something best not searched for :p
23:32 timotimo i can do the vim gang sign
23:32 geekosaur (rule 34 in 3, 2, ...)
23:32 timotimo bedtime!
23:33 grondilu well, apparently there's none.
23:33 timotimo start one, then :)
23:34 grondilu neh, I'm really not into tattoos but I've read a bit about this hobby lately on reddit.
23:37 grondilu not related at all : I've mentioned it the other day, but here I'll go again : how could I have the product of three objects been implemented by a procedure?  like $a*$b*$c would call my-product($a, $b, $c) instead of my-product($a, my-product($b, $c))?
23:37 timotimo you'd have to set the operator be list assoc
23:37 grondilu hum... tell me more.
23:37 timotimo not sure if/how you can do that in your code
23:38 timotimo i think you can only have one kind of assoc per operator, so with * you're potentially out of luck already
23:38 * timotimo goes to bed actually
23:39 grondilu m: sub infix:<é>(*@a) { [+] @a }; say 5 é 4 é 2
23:39 camelia rakudo-moar 87e91c: OUTPUT«11␤»
23:40 grondilu m: multi infix:<*>($a, $b, $c) { Inf }; say 5 * 2 * 3
23:40 camelia rakudo-moar 87e91c: OUTPUT«30␤»
23:51 arnsholt grondilu: You'll probably have to define your own operator and make it list associative
23:52 arnsholt infix:<*> is declared as left associative, so you're out of luck there
23:52 arnsholt It *might* be possible to hack the assoc of the existing operator, but that smells like a bad idea
23:56 BenGoldberg joined #perl6

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

Perl 6 | Reference Documentation | Rakudo