Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2016-08-26

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 canopus joined #perl6
00:02 mempko joined #perl6
00:04 mcmillhj joined #perl6
00:16 BenGoldberg joined #perl6
00:17 mr-foobar joined #perl6
00:24 pierre_ joined #perl6
00:35 darutoko joined #perl6
00:38 bioduds do you guys know about koans?
00:38 bioduds any for perl6?
00:40 MilkmanDan joined #perl6
00:49 profan does anyone have some examples of how to use perl6's macro system anywhere?
00:49 profan i've seen a few mentions of quasi and macro but really no more examples beyond the most basic, would be appreciated :)
00:51 cognominal joined #perl6
00:53 geekosaur at the moment there isn't a lot to say, because it was tagged experimental due to problems with it
00:54 geekosaur anything more interesting than the basic examples tend to throw weird errors
00:58 skids joined #perl6
00:58 arw [Coke], timotimo: unstuck yes, i've got it doing what i wanted. https://pbot.rmdir.de/sKwAyzCdctRMrsIzvBTIfQ
00:58 arw [Coke], timotimo: thanks alot for all your patience and help, wouldn't have managed without.
00:59 profan geekosaur: ah, shame :P, at least i know then
00:59 arw .o( and yes, that code needs cleaning up, but now is bedtime )
01:06 Zoffix joined #perl6
01:07 Zoffix bioduds, yeah, I know about them, but am not aware of any Perl 6 ones. You should definitely make some!
01:07 MasterDuke profan: i think 007 ( https://github.com/masak/007 ) is considered the farthest anyone has gone with macros
01:14 Zoffix bioduds, Dave Doyle was also suggesting we make some. So you should :) https://twitter.com/meraxes
01:16 bioduds definitely
01:17 bioduds I bet if we join we can put up one. I'm still learning P6
01:18 Zoffix \o/
01:18 Zoffix bioduds, that's the best time to write them. You can still see them for how good they are from the perspective of someone new to P6
01:19 Zoffix And if you need space to host them, I'm sure we can find one.
01:20 mcmillhj joined #perl6
01:25 bioduds well
01:25 bioduds I will then
01:46 MilkmanDan joined #perl6
01:54 Actualeyes joined #perl6
01:57 ShimmerFairy joined #perl6
02:03 pierre_ joined #perl6
02:08 pierre_ joined #perl6
02:24 mspo joined #perl6
02:26 kyclark_ joined #perl6
02:30 cyphase joined #perl6
02:32 wamba joined #perl6
02:39 nowan joined #perl6
02:46 mcmillhj joined #perl6
02:47 noganex joined #perl6
02:50 rodarmor joined #perl6
02:51 dalek whateverable: 3510065 | MasterDuke17++ | / (2 files):
02:51 dalek whateverable: Convert benchable to use the Perl 6 SVG::Plot
02:51 dalek whateverable: review: https://github.com/perl6/wha​teverable/commit/3510065082
02:55 NEveD joined #perl6
02:57 revdiablo joined #perl6
02:57 edenc joined #perl6
03:00 prussian joined #perl6
03:00 adrusi joined #perl6
03:02 prussian joined #perl6
03:04 DrForr joined #perl6
03:06 cibs joined #perl6
03:06 geraud joined #perl6
03:13 bob777 joined #perl6
03:14 adu joined #perl6
03:26 labster joined #perl6
03:33 kyclark_ joined #perl6
03:39 adu joined #perl6
03:53 mcmillhj joined #perl6
03:55 obfusk_ joined #perl6
04:10 cognominal joined #perl6
04:12 Cabanossi joined #perl6
04:19 awwaiid bioduds: koans, ala ruby-koans, would be fantastic. Is on my TODO list, but that is a lazy-infinite generator so unlikely to happen
04:19 Xliff I have a module A::B with submodules A::B::C..E
04:20 Xliff I would like to be able to load symbols from A::B::C, et all from A::B via 'use A::B :c'
04:20 Xliff Can I do that without dealing with the export mechanism?
04:37 Xliff I'm getting "Malformed UTF-8" when attempting to open a file. What do I need to do to fix that?
05:00 pierre_ joined #perl6
05:05 awwaiid j: say CORE::.keys
05:05 camelia rakudo-jvm cd19db: OUTPUT«ContextRef representation does not implement elems␤  in block <unit> at <tmp> line 1␤␤»
05:08 Xliff Ooh! Rakudo-JVM!
05:09 Xliff awwaiid, do you know if the user can extract the extra values form a use statement ala 'use Module "i want this string"'
05:19 pierre_ joined #perl6
05:22 awwaiid Xliff: I think you can write an EXPORT sub that will be passed those values. https://docs.perl6.org/language/modul​es#Exporting_and_Selective_Importing has some examples.
05:23 prussian joined #perl6
05:25 lizmat joined #perl6
05:27 lizmat joined #perl6
05:33 Xliff awwaiid++ # Thanks!
05:37 Actualeyes joined #perl6
05:38 pierre_ joined #perl6
05:38 leont joined #perl6
05:39 lizmat j: dd CORE::.iterator
05:39 camelia rakudo-jvm cd19db: OUTPUT«ContextRef representation does not implement elems␤  in block <unit> at <tmp> line 1␤␤»
05:39 lizmat :-(
05:44 prussian joined #perl6
05:49 awwaiid lizmat: since you merged that PR, should I close the RT, or do we wait to close RT for a release?
05:50 lizmat what was the RT again ?
05:50 awwaiid https://rt.perl.org/Ticket/Display.html?id=1290​20&amp;results=14e870f9b2998d2106b164a32f2d32a2
05:50 awwaiid #129020: [JVM] REPL does not work anymore: 'ContextRef representation does not implement elems'
05:50 synopsebot6 Link:  https://rt.perl.org/rt3//Publ​ic/Bug/Display.html?id=129020
05:50 lizmat no, please keep open
05:50 awwaiid ooo fancy
05:51 awwaiid ok -- then close upon release?
05:51 lizmat awwaiid: well, it should be closed when completions work again on JVM maybe
05:51 awwaiid ah. ya, there is that
05:51 lizmat or maybe a new ticket should be made about ContextRef not having .elems suppoart
05:52 lizmat I understand it was a small change in MoarVM
05:52 lizmat I assume it's going to be a similar small change on the JVM
05:52 awwaiid I opened RT#129092 re ContextRef jvm
05:52 synopsebot6 Link:  https://rt.perl.org/rt3//Publ​ic/Bug/Display.html?id=129092
05:52 lizmat (but I could be o so wrong  :-)
05:54 awwaiid lizmat: for ^C in the REPL, I read up on your implementation and have a better (though not totally complete) understanding of why the .tap on the signal(SIGINT) happens in another thread, and thus why we likely need to embrace concurrency
05:55 awwaiid lizmat: I tried to build some toy examples of merging STDIN and a signal into a single supply as another approach, but haven't gotten it to work so far
05:55 lizmat well, we need to anyway somehow  :-)
05:56 pierre_ joined #perl6
05:56 lizmat fwiw, I'm working on it as well now
05:56 lizmat I put some part of the original patch just in (centralizing the reset logic)
05:56 awwaiid https://gist.github.com/awwaiid/​7f1c6819ec00a0cc8f8bdb3de14e2ab3 is my toy
05:57 awwaiid ah nice. I touched some code that might conflict in https://github.com/rakudo/rakudo/pull/847, but now see that my PR is already merged
05:59 awwaiid ah, I see your commit
06:00 awwaiid lizmat: did you type the output of 'sub reset(--> Nil) {...}' as an optimization?
06:00 lizmat yeah, and as an indicator it's not supposed to return anything
06:04 awwaiid So the big thing I want is for people to be able to call 'REPL.here' at any random spot in their code and have it drop down into the REPL hooked into that lexical scope. Probably integrated in with some debugger stuff. I've been getting closer. I'm fascinated by the subtle differences between EVAL and compiler.eval implementations.
06:04 nadim joined #perl6
06:05 awwaiid I use binding.pry in ruby-land all the time, and find it essential. I have LREP::here that works for p6, but would be better in the core
06:06 lizmat Strange consistency starts :-)
06:06 nine masak: TimToady++'s giving a keynote about your blog :)
06:06 lizmat http://act.yapc.eu/ye2016/talk/6881
06:07 nwc10 joined #perl6
06:07 nwc10 LIVE TimToady  https://youtu.be/A3ZSt5QKEAg
06:08 awwaiid ooo
06:09 nwc10 with a fluffy camelia in front
06:09 nwc10 (product placement)
06:10 awwaiid https://thelackthereof.org/Perl6_Colons has a list of lots of colons :)
06:11 awwaiid Is that a 4x4 grid of televisions?
06:12 prussian joined #perl6
06:14 pierre_ joined #perl6
06:14 prussian joined #perl6
06:22 lizmat awwaiid: yes
06:25 awwaiid fancy
06:25 awwaiid I want a pile of camelia stickers
06:31 leont joined #perl6
06:33 autarch joined #perl6
06:38 domidumont joined #perl6
06:43 domidumont joined #perl6
06:46 rkazak joined #perl6
07:11 webstrand joined #perl6
07:17 firstdayonthejob joined #perl6
07:22 mcmillhj joined #perl6
07:25 bjz joined #perl6
07:30 cognominal joined #perl6
07:31 rindolf joined #perl6
07:34 RabidGravy joined #perl6
07:40 killbill joined #perl6
07:41 masak 'o morning, #perl6
07:41 Xliff RabidGravy: https://github.com/Xliff/p6-color-names
07:41 masak nine: ...about my blog? /o\
07:42 Xliff Now to figure out the best way to abstract everything in Color::Names::* behind Color::Names. Any ideas?
07:42 ab6tract joined #perl6
07:42 ab6tract https://github.com/ab5tract/Terminal-Print
07:42 ab6tract nine ^^
07:43 masak "it turns out the person we thought was the stakeholder wasn't the stakeholder" (from livestream) -- been there, done that ;)
07:44 Xliff masak: Blog?
07:44 masak Xliff: http://irclog.perlgeek.de/p​erl6/2016-08-26#i_13093189
07:45 killbill Why I don't see last line? $    perl6 -ne 'LAST .say' file.txt
07:45 yoleaux 7 Aug 2016 23:39Z <TimToady> killbill: Try something like this: say '𝐀𝐁𝐂'.NFKD.Str === 'ABC'
07:45 nine masak: Title was Strange Consistency and he started off by showing your blog
07:45 masak oh, ok :)
07:45 masak ...I should really redecorate my blog... :P
07:46 bjz joined #perl6
07:47 masak killbill: because bug, I'd say
07:47 masak killbill: mind if I rakudobug it?
07:48 masak nine: by the way, nowadays when I know about CT's "natural transformations", I think that's what I was after when I talked about "strangely consistent" things
07:48 masak at least most of the time
07:50 g4 joined #perl6
07:50 g4 joined #perl6
07:53 leont joined #perl6
07:59 * masak submits killbill's LAST bug
08:01 lizmat joined #perl6
08:03 leont joined #perl6
08:07 DrForr_ joined #perl6
08:09 Actualeyes joined #perl6
08:11 imcsk8_ joined #perl6
08:12 matt_ joined #perl6
08:20 zakharyas joined #perl6
08:23 ab6tract jnthn: looking into the bug i mentioned yesterday with nine++
08:23 ab6tract seems to be related to async hash construction, but like a good heisenbug it hides when golfed
08:24 mvorg joined #perl6
08:24 nine jnthn: crashes with consistent backtrace in a free() in extract_key so most probably actually MVM_string_flatten. We're creating hashes with static keys so maybe the key strings are flattened by multiple threads at the same time.
08:28 ab6tract notably, it *appears* to stop crashing when wrapping with a Lock protect call
08:40 rindolf joined #perl6
08:40 jnthn nine: Yeah, I'm aware of that one...it won't go away until the hash re-work I've got planned is done, which I guess will happen after I'm done with the array one to fix a similar set of crashes upon mis-use
08:40 nine jnthn: ok. I'm curious what a fix would look like?
08:40 nadim joined #perl6
08:41 jnthn nine: Implement hashing that doesn't need the flattening first :)
08:41 nine Sounds good :)
08:42 jnthn It was a hack enforced by using ut_hash, which seemed like a better idea than a home-grown impl, but by this point we've customized the thing so much anyway... :)
08:43 ab6tract lucky for my case (not having perl6 explode while showing my fancy pants matrix-y screensaver thingy), the Lock "fixes" it
08:44 jnthn Phew. :)
08:47 nine Probably another workaround would have been to just create a similar hash (using the same constant keys) before starting all the threads, wouldn't it?
08:49 jnthn Aye
08:56 zakharyas1 joined #perl6
08:59 ab6tract seems to work :D
09:03 chienjo joined #perl6
09:09 bjz_ joined #perl6
09:13 ab6tract .u u+2620
09:13 yoleaux U+2620 SKULL AND CROSSBONES [So] (☠)
09:13 ab6tract .u u+2621
09:13 yoleaux U+2621 CAUTION SIGN [So] (☡)
09:13 ab6tract .u u+2622
09:13 yoleaux U+2622 RADIOACTIVE SIGN [So] (☢)
09:13 ab6tract .u u+2623
09:13 nwc10 left #perl6
09:13 yoleaux U+2623 BIOHAZARD SIGN [So] (☣)
09:14 labster joined #perl6
09:14 ab6tract is there a quick way to get that .u functionality from the repl?
09:14 ab6tract i guess we don't ship the unicode corpus with rakudo
09:18 brrt joined #perl6
09:19 jnthn m: say uniname(0x2623)
09:19 camelia rakudo-moar 508192: OUTPUT«BIOHAZARD SIGN␤»
09:21 El_Che mst talks is perltrelated
09:21 El_Che https://www.youtube.com/watch?v=A3ZSt5QKEAg
09:22 El_Che perl6 related I meant
09:24 Xliff Is there a way to capture the options used after use statement? Ala "use Module :a :b 'c'"; Yes. The c is positional.
09:28 Actualeyes left #perl6
09:39 leont joined #perl6
09:45 konobi it's a normal method call on the Module, no?
09:47 mcmillhj joined #perl6
09:59 mcmillhj joined #perl6
10:08 frithnanth joined #perl6
10:10 frithnanth m: for 1, 2, 3 -> $item { my $x; FIRST { $x = 42 }; say $item, $x }
10:10 camelia rakudo-moar 508192: OUTPUT«142␤2(Any)␤3(Any)␤»
10:11 mcmillhj joined #perl6
10:11 frithnanth Is this the expected behavior?
10:11 jnthn Yes.
10:12 jnthn Maybe you wanted `state $x`?
10:13 frithnanth @jnthn thanks
10:15 frithnanth On the same topic
10:15 frithnanth m: for 1, 2, 3 -> $item { my $x = FIRST { 42 }; say $item, $x }
10:15 camelia rakudo-moar 508192: OUTPUT«1(Any)␤2(Any)␤3(Any)␤»
10:15 frithnanth I thought that FIRST could return a value
10:16 frithnanth m: for 1, 2, 3 -> $item { state $x = FIRST { 42 }; say $item, $x }
10:16 camelia rakudo-moar 508192: OUTPUT«1(Any)␤2(Any)␤3(Any)␤»
10:16 jnthn No, it's not one of the phasers that do that set.
10:17 jnthn Probably in some future Perl 6 version we'll allow it :)
10:17 jnthn *do that yet
10:18 jnthn Note that since state variables are only initialized on the first iteration, then `state $x = 42;` would also work fine
10:18 frithnanth I see. Here: https://docs.perl6.org/language/phasers there's this note "Phasers marked with a * have a run-time value, and if evaluated earlier than their surrounding expression, they simply save their result for use in the expression later when the rest of the expression is evaluated" and FIST was one of those
10:18 jnthn Ah
10:18 frithnanth s/FIST/FIRST/
10:18 frithnanth Probably the doc is ahead of the compiler :-)
10:19 jnthn Feel free to file a doc bug on that; the docs probably nabbed that list from the language design docs (which are speculative in nature) and didn't check the reality
10:19 jnthn Well, the docs are meant to tell users what they can do today. :)
10:24 lizmat joined #perl6
10:27 nadim joined #perl6
10:37 leont joined #perl6
10:43 mcmillhj joined #perl6
10:49 melezhik good day!
10:49 melezhik anybody knows a doc on perl6 case operator?
10:50 melezhik could not find a refs
10:50 cognominal joined #perl6
10:52 lizmat given/when ?
10:53 melezhik yes
10:59 Xliff https://docs.perl6.org/language/control#given
10:59 zakharyas joined #perl6
11:01 melezhik Xliff: thanks!
11:01 Xliff yw =)
11:01 * Xliff finally grokked given/take just this week.
11:02 Xliff Question: I have a set of modules A::B, and A::B::*. I want to abstract all of the modules A::B::* behind A::B in a selective manner. What is the best way to do this?
11:04 Xliff I was hoping I could "use A::B :a :b;" to have A::B load A::B::A and A::B::B but I haven't found a good way to do that. sub EXPORT() doesn't seem to pick it up and I don't think UNIT::EXPORT has been implemented, yet.
11:05 labster joined #perl6
11:09 richi235 joined #perl6
11:15 mcmillhj joined #perl6
11:30 mcmillhj joined #perl6
11:30 mst nine: hey, any chance of tweaking your install for -Perl5 to have 'make homeinstall' and 'make siteinstall' to be explicit, then make 'make install' default to site if it can, then first writable if not
11:35 El_Che why is SawyerX making perl 5 faster? It will make Perl6 look bad ;)
11:35 leont joined #perl6
11:37 jnthn To keep us on our toes :)
11:43 krunen joined #perl6
11:46 tadzik #damnitsawyer
11:51 arnsholt OpenSSL is kind of odd
11:52 arnsholt Error code are 32-bit ints, of which 12 bits are a reason code, 12 bits a function code and 8 bits a library identifier
11:52 mst that's probably the nicest thing I've ever heard anybody say about OpenSSL
11:52 arnsholt I'm trying to be diplomatic
11:52 arnsholt The docs are also an endless source of quotable material
11:53 arnsholt Like this one: "It can also occur of action is need to continue the operation for non-blocking BIOs."
11:53 arnsholt It can also occur of action is need!
11:55 jnthn If it's BIO, it must be good for you...
11:56 Xliff Is there a Perl6 equivalent to @INC?
11:56 matt_ joined #perl6
11:56 mst Xliff: $*REPO is a linked list of CompUnit::Repository objects
11:57 jnthn m: say $*REPO.repo-chain # I think this traverses it for you
11:57 camelia rakudo-moar 508192: OUTPUT«(inst#/home/camelia/.perl6 inst#/home/camelia/rakudo-​m-inst-2/share/perl6/site inst#/home/camelia/rakudo-m​-inst-2/share/perl6/vendor inst#/home/camelia/rakudo-m-inst-2/share/perl6 CompUnit::Repository::AbsolutePath.new(next-repo => CompUnit::Repository::NQ…»
11:58 mst Xliff: I think the question is -
11:58 mst step back. explain what you're trying to achieve.
12:00 mcmillhj joined #perl6
12:01 Xliff mst: I have. No body has answered.
12:01 Xliff I will repeat it again.
12:01 Xliff <Xliff> Question: I have a set of modules A::B, and A::B::*. I want to abstract all of the modules A::B::* behind A::B in a selective manner. What is the best way to do this?
12:01 Xliff <Xliff> I was hoping I could "use A::B :a :b;" to have A::B load A::B::A and A::B::B but I haven't found a good way to do that. sub EXPORT() doesn't seem to pick it up and I don't think UNIT::EXPORT has been implemented, yet.
12:01 Xliff Thanks, jnthn++. If I don't use it now, it will be helpful later... for something. =)
12:03 mst "I want to abstract all of the modules behind A::B::*" doesn't actually parse
12:03 canopus joined #perl6
12:03 mst you're currently explaining what you're failing to implement
12:03 mst please take a further step back and explain the goal
12:03 arw left #perl6
12:04 Xliff Er. That was the goal.
12:04 mst why?
12:04 mst what is the point?
12:05 Xliff I have objects in A::B::A, A::B::B and A::B::C.
12:05 Xliff I don't want the user to have to touch those.
12:05 nine mst: how about DESTREPO=stie make install?
12:05 mst nine: would be fine by me
12:05 Xliff A::B will have the interface and the user can specify what decendents they want loaded.
12:06 Xliff I mean in the end, I guess I could just leave it with A::B::A et all, but that would be messy.
12:07 Xliff So I want one package to dynamically load other decendants in its namespace.
12:07 Xliff With the user controlling which descendent packages to load.
12:07 mst that seems like a gun looking for a foot
12:07 Xliff Yes, it may seem so. That's why I am wearing my kevlar boots.
12:08 leont joined #perl6
12:08 literal so you want something like "use POE qw(Component::Foo);" does for POE in Perl 5
12:08 Xliff If I can't get it to work, fine. I will try and find another method.
12:08 mst I've implemented this several times in perl5
12:08 Xliff literal: Something like that, yes.
12:08 ejf joined #perl6
12:08 mst and concluded it always ends up as a footgun
12:09 Xliff mst: Warning noted and appreciated. I would still like to know if it is possible in Perl6.
12:10 mst maybe find another module with custom EXPORT() logic and try and crib?
12:10 Xliff mst: Well, that would be brute!
12:10 Xliff I'll put that as Plan Z, for now. ;)
12:10 mst well, if you want to find out if it's possible, finding prior art would be a good way to tell
12:10 Xliff True enough.
12:11 Xliff But iterating through modules.perl6.org hoping to find that needle....well...
12:11 dalek Inline-Perl5: 865e918 | niner++ | configure.pl6:
12:11 dalek Inline-Perl5: Support DESTREPO for manually picking the make install target
12:11 dalek Inline-Perl5: review: https://github.com/niner/Inl​ine-Perl5/commit/865e918884
12:11 nine mst: pushed
12:11 n1lp7r joined #perl6
12:11 Xliff I'm willing to do it, but I've got higher priorities.
12:12 ejf Heya everybody :-) I'm learning Perl6 and trying things out, and got stuck on something I can't really reason to myself about or find what is wrong. It concerns wrapping a sub with mod_trait:<is> throwing an error if I import said sub from a different module, but working just fine if I use the same code in the exact same module. Made a gist out of it: https://gist.github.com/eikef/​7de3095f0034ea17b2c85710df1faf
12:12 ejf 76 -- maybe somebody could point me in the right direction?
12:12 ejf (the error is Cannot invoke this object (REPR: Null; VMNull))
12:13 nine ejf: link's broken
12:13 mst Xliff: https://rt.perl.org/Public/​Bug/Display.html?id=127305 suggests perhaps your problem is using 'unit module' ?
12:14 Xliff Proper link for ejf - https://gist.github.com/eikef/7​de3095f0034ea17b2c85710df1faf76
12:14 ejf https://gist.github.com/eikef/7​de3095f0034ea17b2c85710df1faf76
12:14 ejf sorry, that got split into second line, I apologize
12:14 AlexDaniel melezhik: I'd love to see support for zstd :) (https://github.com/Cyan4973/zstd)
12:16 dalek ecosystem: c8bb5bf | (Alexey Melezhik)++ | META.list:
12:16 dalek ecosystem: Extract various archives using tar/unzip programs
12:16 dalek ecosystem: review: https://github.com/perl6/e​cosystem/commit/c8bb5bf40c
12:16 dalek ecosystem: 710dd14 | azawawi++ | META.list:
12:16 dalek ecosystem: Merge pull request #243 from melezhik/master
12:16 dalek ecosystem:
12:16 dalek ecosystem: Extract various archives using tar/unzip programs
12:16 dalek ecosystem: review: https://github.com/perl6/e​cosystem/commit/710dd14f2d
12:16 Xliff mst: I remember reading that, earlier but couldn't get it to work.
12:16 nine mst: $*REPO.need(CompUnit::DependencySpecification.new(​:short-name<Inline::Perl5>)).distribution.meta.say
12:17 nine mst: the provides and files keys contain what you need
12:17 tbrowder hi, #perl6!
12:17 Xliff I will try to work around "unit package" though. Thanks for the suggestion.
12:17 mst nine: ta
12:17 tbrowder I have a problem with a regex capture (using the latest rakudo from the nom branch).
12:18 tbrowder See my github gist "https://gist.github.com/tbrowder/​d39967e5fb30d43573744af3042812f2".  I may be doing something stupid, but I get this error when it's executed:
12:18 tbrowder https://www.irccloud.com/pastebin/LIMG4q6Q/
12:19 AlexDaniel melezhik: ah, it's something specific for Sparrowdo… nevermind then… :)
12:19 nine ejf: your issue may be a bug in rakudo (those REPR messages usually indicate that) or a DIHWIDT as trying to reason what exactly should be exported is kinda hard.
12:19 nine ejf: sorry that I can't be more helpful there :/
12:20 melezhik AlexDaniel: if Zstd has console clients I will
12:21 tbrowder Note that if I remove the Int type I get another error about not being able to get 'succ' on a match type at the "++$serial" statement.
12:21 melezhik looking at  https://github.com/Cyan4973/zstd it's called zstd ?
12:22 melezhik anyway you probably mean something out of sparrowdo scope )))
12:22 ejf nine: when I also export the mod_trait:<is> the same thing happens, as it does when I have both of those things inside a class that is exported. Ultimately I'm trying to write an error handler that wraps NativeCalls to an external library. Maybe I am going about it the wrong way :)
12:22 ejf nine: should I put this snippet on RT as a bug?
12:23 AlexDaniel I'm just thinking that something is missing in this list ;) http://cyan4973.github.io/zstd/#other-languages
12:23 wamba joined #perl6
12:23 AlexDaniel I see Delphi, I see OCaml… something is not there, not sure what ;)
12:24 ejf and thanks for looking at it. At least there then doesn't seem something very, very obviously wrong :)
12:25 nine ejf: well you probably mean to export the wrapped sub but that would mean that rakudo has to process the traits in the right way and I'm not sure that's guaranteed. Could be that that's just a combination of features that's not really feasible to support.
12:26 nine ejf: could also be just a plain straight forward bug in rakudo :) I'm not familiar enough with that part to make a good guess and am too tired to start debugging right now
12:32 ejf nine: alright, thanks :-) I'll send in a bug report to be sure. If it's unsupportable that's an answer too which could yield a documentation PR :)
12:34 RabidGravy I really ought to write some software
12:39 mcmillhj joined #perl6
12:47 Skarsnik joined #perl6
12:51 iH2O joined #perl6
12:54 iH2O left #perl6
13:01 rindolf RabidGravy: to do what
13:01 rindolf ?
13:02 cdg joined #perl6
13:03 rmmm joined #perl6
13:04 RabidGravy well actually finishing some of these ten modules that I've started in the last few months would be a start :)
13:10 domidumont joined #perl6
13:14 kyclark_ joined #perl6
13:15 smls joined #perl6
13:32 kyclark_ joined #perl6
13:37 pierre_ joined #perl6
13:40 skids joined #perl6
13:42 leont joined #perl6
13:42 DrForr_ joined #perl6
13:42 tbrowder Bye #perl6, I'll check for a response after I get home later today.
13:43 tbrowder left #perl6
13:51 MilkmanDan joined #perl6
13:53 smls .tell tbrowder Change «$serial = $0;» to «$serial = +$0;».
13:53 yoleaux smls: I'll pass your message to tbrowder.
13:53 mcmillhj_ joined #perl6
14:06 MilkmanDan joined #perl6
14:09 espadrine joined #perl6
14:10 cognominal joined #perl6
14:15 zakharyas joined #perl6
14:21 kyclark_ joined #perl6
14:21 jameslenz joined #perl6
14:26 gfldex joined #perl6
14:27 sufrostico joined #perl6
14:30 gfldex m: enum Noms('Greenstuff', 'Walkingstuff', 'Syntetics'); my %eaten{Noms}; %eaten{Greenstuff}++; dd %eaten, %eaten.keyof;
14:30 camelia rakudo-moar 1a03ef: OUTPUT«Hash[Any,Noms] %eaten = (my Any %{Noms} = Noms::Greenstuff => 1)␤Noms␤»
14:34 richi235 joined #perl6
14:41 Xliff ejf: Odd. I've tried to export traits and didn't have much luck there, either.
14:42 gfldex m: enum Noms('Greenstuff', 'Walkingstuff', 'Syntetics'); my %eaten{Noms} but role :: { method keys() { Noms.enums} }; %eaten{Greenstuff}++; dd %eaten, %eaten.keyof, %eaten.keys;
14:42 camelia rakudo-moar 575dda: OUTPUT«Hash[Any,Noms] %eaten = (my Any %{Noms} = Noms::Greenstuff => 1)␤Noms␤(Noms::Greenstuff,).Seq␤»
14:42 gfldex m: enum Noms('Greenstuff', 'Walkingstuff', 'Syntetics'); my %eaten{Noms} but role :: { method keys() { 'oi!' } }; %eaten{Greenstuff}++; dd %eaten, %eaten.keyof, %eaten.keys;
14:42 camelia rakudo-moar 575dda: OUTPUT«Hash[Any,Noms] %eaten = (my Any %{Noms} = Noms::Greenstuff => 1)␤Noms␤(Noms::Greenstuff,).Seq␤»
14:42 gfldex why does it call the wrong .keys ?
14:42 gfldex (wrong from my POV ofc)
14:47 Actualeyes joined #perl6
14:49 movlex joined #perl6
14:50 movlex joined #perl6
14:52 jnthn gfldex: Try "does"?
14:52 gfldex m: enum Noms('Greenstuff', 'Walkingstuff', 'Syntetics'); my %eaten{Noms} does role :: { method keys() { 'oi!' } }; %eaten{Greenstuff}++; dd %eaten, %eaten.keyof, %eaten.keys;
14:52 camelia rakudo-moar 575dda: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Invalid typename 'role'␤at <tmp>:1␤------> 3 'Syntetics'); my %eaten{Noms} does role7⏏5 :: { method keys() { 'oi!' } }; %eaten{␤»
14:52 jnthn heh :)
14:53 jnthn Yeah, that ain't gonna do it
14:53 gfldex m: enum Noms('Greenstuff', 'Walkingstuff', 'Syntetics'); my %eaten{Noms} is Hash does role :: { method keys() { 'oi!' } }; %eaten{Greenstuff}++; dd %eaten, %eaten.keyof, %eaten.keys;
14:53 camelia rakudo-moar 575dda: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Invalid typename 'role'␤at <tmp>:1␤------> 3ics'); my %eaten{Noms} is Hash does role7⏏5 :: { method keys() { 'oi!' } }; %eaten{␤»
14:53 jnthn does expects a typename after it
14:53 jnthn I think (my %eaten{Noms}) does role :: { ... }; would work
14:53 gfldex m: enum Noms('Greenstuff', 'Walkingstuff', 'Syntetics'); (my %eaten{Noms}) does role :: { method keys() { 'oi!' } }; %eaten{Greenstuff}++; dd %eaten, %eaten.keyof, %eaten.keys;
14:53 camelia rakudo-moar 575dda: OUTPUT«Hash[Any,Noms]+{<anon|49843632>} %eaten = (my Any %{Noms} = Noms::Greenstuff => 1)␤Noms␤"oi!"␤»
14:54 gfldex m: enum Noms('Greenstuff', 'Walkingstuff', 'Syntetics'); (my %eaten{Noms}) does role :: { method keys() { Noms.emums } }; %eaten{Greenstuff}++; dd %eaten, %eaten.keyof, %eaten.keys;
14:54 camelia rakudo-moar 575dda: OUTPUT«Method 'emums' not found for invocant of class 'Noms'␤  in method keys at <tmp> line 1␤  in block <unit> at <tmp> line 1␤␤»
14:54 gfldex m: enum Noms('Greenstuff', 'Walkingstuff', 'Syntetics'); (my %eaten{Noms}) does role :: { method keys() { Noms.enums.keys } }; %eaten{Greenstuff}++; dd %eaten, %eaten.keyof, %eaten.keys;
14:54 camelia rakudo-moar 575dda: OUTPUT«Hash[Any,Noms]+{<anon|75746832>} %eaten = (my Any %{Noms} = Noms::Greenstuff => 1)␤Noms␤("Greenstuff", "Walkingstuff", "Syntetics").Seq␤»
14:54 Skarsnik_ joined #perl6
14:55 jnthn bbiab
14:55 gfldex :)
14:56 mcmillhj joined #perl6
14:56 infina joined #perl6
14:57 Roamer` ok, here's a weird one...
14:57 Roamer` Is there a way for me to "mimic" the default values of some other method's parameters?
14:58 Roamer` like, say, I want to add :nl and :enc to Shell::Capture.capture(), which invokes Proc.new(), and pass :nl and :enc to Proc.new
14:58 Roamer` but only if they have been supplied to Shell::Capture.capture()
14:59 Roamer` I don't want to have several invocations of Proc.new, one with :nl, one with :enc, one with both, etc
14:59 kjs_ joined #perl6
14:59 MilkmanDan joined #perl6
14:59 Roamer` and, quite obviously, I don't want to go this way:
14:59 Roamer` m: use v6.c; say Proc.^methods.first(*.name eq 'BUILD').signature.params.first(-> $param { ($param.name // '') eq '$chomp' }).default()()
14:59 camelia rakudo-moar 575dda: OUTPUT«Cannot find method 'Nil' on object of type List␤  in block <unit> at <tmp> line 1␤␤»
14:59 Roamer` hm, this is weird, it works here on 2016.08
14:59 Roamer` (it outputs "True", which is the default value for Proc.new :chomp)
15:01 Roamer` so is there some less hacky (and, apparently, more portable across Rakudo versions) way of saying "only pass :$nl to Proc.new() if it is defined"?
15:01 kurahaupo joined #perl6
15:03 FlyingPanda joined #perl6
15:04 FlyingPanda Roamer`: camelia is restricted, so you don't get the same Proc as in non-restricted version
15:04 Roamer` hmm, and I just thought of the quite obvious run ..., $nl.defined?? :$nl!! () ... but I'm not completely sure that it will work
15:05 FlyingPanda Roamer`: You can ninja them in: Proc.new: |(:$nl if $nl), |(:$enc if $enc)
15:05 Roamer` FlyingPanda, oh, sorry, I didn't realize that... did I just summon you here by doing that?  Sorry!
15:05 FlyingPanda Nah, I was just bored and passing by :)
15:05 Roamer` FlyingPanda, ah, yes, slips would work... thanks!
15:06 RabidGravy is anyone doing anything where they are testing a script that is part of a module distribution?
15:06 Roamer` (and here I am yet again, coming up with a awfully (or wonderfully) convoluted hack instead of thinking of the obvious...)
15:07 FlyingPanda |(:) looks like a ninja if you tilt your head right, which is how I remember the syntax :)
15:08 rkazak joined #perl6
15:08 FlyingPanda RabidGravy: why?
15:09 RabidGravy why what?
15:10 RabidGravy why am I interested in testing scripts?
15:10 rba_ joined #perl6
15:11 FlyingPanda RabidGravy: why are you asking?
15:12 dustinm` joined #perl6
15:12 RabidGravy because I want to know if anyone has done anything where they are testing a script that is part of a module-distribution ;-p
15:12 FlyingPanda ~_~
15:13 FlyingPanda Well, I did lots of script testing, but since they weren't part of a module distribution, I guess I'm not the right fit for your survey.
15:13 RabidGravy because the script that I install as part of JSON::Infer may not have been working properly since February and it isn't tested, and if someone is doing tests of scripts I could save the reinvention of the wheel
15:15 Roamer` RabidGravy, I'm doing something like this in a module that I haven't released yet... it's a large part of the reason why I wrote Shell::Capture
15:16 leont joined #perl6
15:16 RabidGravy I will however resist the urge to make something like Test::Script
15:16 RabidGravy for the time being
15:16 Roamer` RabidGravy, but if you're asking whether there is something a bit more like Perl 5's Test::Command, I don't know of any
15:16 Roamer` (but I *am* using Test::Command in Perl 5 in-house stuff)
15:16 RabidGravy yeah I've used that before
15:17 RabidGravy last big P5 project had tens of scripts which needed to be tested
15:18 RabidGravy this would also suggest that either no-one has installed and used JSON::Infer since february or they have avoided the bug
15:18 kyclark I would love any suggestions to make this more Perl6-ish:
15:18 yoleaux 25 Aug 2016 22:57Z <MasterDuke> kyclark: /:i ^ keys?|values? $ / may not do what you want. the '^' and '$' don't distribute to the different values in an alternation. i suspect you want /:i ^ [key|value]s? $ /
15:18 kyclark https://github.com/kyclark/metagenomics-book/bl​ob/master/hpc/pbs/blast/workers/fasta-split.pl6
15:18 mcmillhj joined #perl6
15:19 RabidGravy I'm going to the former
15:19 infina joined #perl6
15:21 * [Coke] cautiously declares tmux an improvement over screen.
15:22 Roamer` RabidGravy, and now you've gone and made me push my horribly ugly, WIP, not even close to finished yet, thingy to Github just so I can suggest that you take a look at https://github.com/ppentchev/gan​g/blob/master/t/01-test-local.t
15:22 gfldex kyclark: in `my $out-fh = &next-fh();`the & is redundant
15:23 Roamer` of course, run-check() itself is in https://github.com/ppentchev/gang/b​lob/master/t/lib/GANG/Test/Lib.pm6
15:23 kyclark OK, thanks.  One thing I /REALLY/ don't like is the $*SPEC thing.  I would rather import the functions I need from a module.  Is there a way around this?
15:23 setty1 joined #perl6
15:27 [Coke] kyclark: what problem are you trying to solve ther?
15:27 gfldex kyclark: see https://github.com/rakudo/rakudo/bl​ob/02fdcf9a1f7bc34e2d263fce1c89fb30​09ec7e9f/src/core/IO/Path.pm#L234
15:27 [Coke] (by using $*SPEC
15:28 kyclark Just want to use the catfile/catdir functions to put together a path to a resource
15:29 kyclark Also, is there a method to get the "basename" of a file minus its extension like in bash $(basename 'file.fa' '.'fa)?
15:29 kjs_ joined #perl6
15:29 mcmillhj joined #perl6
15:29 Roamer` RabidGravy, and, of course, now you've got me thinking about writing a Perl 6 Test::Command, but it may not happen very soon (or at all) :)
15:29 kyclark Rather than "001-file.fa" I would rather make "file-001.fa" etc
15:30 b2gills my &catfile = $*SPEC.^lookup('catfile')
15:30 gfldex m: "file-001.fa".succ.say
15:30 camelia rakudo-moar 575dda: OUTPUT«file-002.fa␤»
15:32 RabidGravy Roamer, bwahahah - my entire purpose in life is tempting people to write more P6 modules
15:34 tbrowder joined #perl6
15:34 tbrowder left #perl6
15:40 Xliff Are the CompUnit::Repository::* modules documented anywhere?
15:41 gregf_ m: $*SPEC.curdir.say
15:41 camelia rakudo-moar 575dda: OUTPUT«.␤»
15:45 mempko joined #perl6
15:47 mcsnolte joined #perl6
15:50 mcmillhj joined #perl6
15:52 lust joined #perl6
15:58 lustlife joined #perl6
16:03 kaare_ joined #perl6
16:07 gfldex kyclark: could you put an example file for fasta-split on github?
16:08 mcmillhj joined #perl6
16:08 iH2O joined #perl6
16:08 RabidGravy I'm finding myself increasingly annoyed with APIs that use halfCamel
16:08 ugexe the CompUnit::Repository:: namespace is full of interface roles that basically document themselves
16:08 RabidGravy that's just dump
16:08 ugexe er, CompUnit:: namespace
16:11 muraiki joined #perl6
16:13 canopus joined #perl6
16:13 kyclark gfldex, I just pushed a new version that has a link to a sample file you can pull down from FTP
16:14 kyclark https://github.com/kyclark/metagenomics-book/bl​ob/master/hpc/pbs/blast/workers/fasta-split.pl6
16:15 harmil_wk joined #perl6
16:15 [Coke] if someone wants to package this up into a reply on the "bootstrap" thread on perl6-users; removing the p5 dep for building is nice to have; the language implementors have been focusing on providing a great language first; what would we replace the p5 dep with?
16:17 mcmillhj joined #perl6
16:24 iH2O left #perl6
16:26 [particle] good news for audrey and for open source: http://www.chinapost.com.tw/taiwan/national/n​ational-news/2016/08/26/476652/Cabinet-to.htm
16:26 ugexe m: my $basename = "foo.tar.gz"; say $basename.IO.extension # kyclark this may or may not matter to you
16:26 camelia rakudo-moar 575dda: OUTPUT«gz␤»
16:28 kyclark ugexe, actually that reminds me that I need to handle gzipped files.  Would I need to pipe the file through "gunzip" or is there a native library to decompress?
16:28 harmil_wk [particle]: wow, that's kind of amazing.
16:28 mcmillhj joined #perl6
16:29 [particle] indeed it is!
16:31 kjs_ joined #perl6
16:32 [Coke] holy crap it's [particle]. Hi, man.
16:33 [particle] hey there, [Coke]!
16:34 ugexe kyclark: `tar` also works. If its truly only gz (not tar as well) you can open on windows with powershell using something like `powershell "$xxx.NameSpace('output dir').CopyHere($xxx.NameSpace($gz_file).items(), 0x14)"`
16:34 kyclark To find regular files in a directory, I should be able to do this: dir($path, test => *.IO.f) ???
16:35 ugexe no
16:35 ugexe $test = none('.', '..')
16:35 ugexe so by default only where `$CURRENT-PATH ~~ none('.', '..')`
16:37 zacts joined #perl6
16:37 ugexe https://github.com/ugexe/Perl6-Path​Tools/blob/master/lib/PathTools.pm6 # This shows how to filter by file (:f), directory (:d), with or without recursion (:r), and setting :$test to non-default values
16:39 ugexe the thing is, you probably *want* that none(".", "..") in addition to whatever other check you are trying to add
16:39 ugexe not to override it totally
16:41 cyphase joined #perl6
16:41 kyclark ugexe, PathTools looks cool, but I can't install via panda.  I get "Please remove leading 'v' from perl version in PathTools's meta info."
16:42 ugexe ... there isnt a leading v though
16:42 ugexe https://github.com/ugexe/Perl6-P​athTools/blob/master/META6.json
16:42 domidumont joined #perl6
16:43 kyclark In my REPL, "dir('.', test => *.IO.f)" works to find only files (excludes a directory), but in my script (on another machine) it doesn't
16:44 dalek doc: d5a91dc | (Jan-Olof Hendig)++ | doc/Language/phasers.pod6:
16:44 dalek doc: FIRST does not return al value, see http://irclog.perlgeek.de/p​erl6/2016-08-26#i_13094318
16:44 dalek doc: review: https://github.com/perl6/doc/commit/d5a91dc984
16:46 [Coke] m: say Int(Int(3434)) # RT 128326
16:46 camelia rakudo-moar 575dda: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Unable to parse expression in typename; couldn't find final ')' ␤at <tmp>:1␤------> 3say Int(Int(7⏏3434)) # RT 128326␤»
16:47 mempko joined #perl6
16:48 ugexe kyclark: regarding the leading v: you probably need to `panda update` as the leading V was removed in January
16:49 kyclark ugexe: ==> Successfully installed PathTools
16:49 kyclark Thanks!
16:50 ugexe that'll probably fix a bunch of other weird module errors you may have encountered with a year-old META data list heh
16:51 leont joined #perl6
16:53 kyclark Re: reading from a file or a pipe from gunzip, does anybody see a problem with this?
16:53 kyclark http://lpaste.net/180530
16:56 ugexe it should work, but on windows you usually have to avoid passing absolute paths on the command line for many unix-y apps
16:58 ugexe https://github.com/ugexe/zef/blob/mas​ter/lib/Zef/Service/Shell/tar.pm6#L23 # This is how zef handles .tar.gz pipe. Note how it passes a *relative* path on the command line and sets run's :cwd parameter to make it act the same as the absolute path
16:58 ugexe because `-c C:\\x\y/a/b/c.tar` doesn't always get passed correctly
16:59 kyclark Cool, thanks for sharing that.
17:02 ugexe https://github.com/ugexe/Perl6-Distribution--Com​mon/blob/master/lib/Distribution/IO/Proc/Tar.pm6 # Mostly the same thing just as a role, just apply to a class with a `has $.prefix = 'my/file.tar.gz'`
17:03 OnlineCop joined #perl6
17:04 OnlineCop I was referred here. Some of you from #perl probably recognize me. :P
17:04 ugexe what is your referral code
17:05 OnlineCop I'm helping implement some PCRE2 into the regex101 site, and we're wondering if the format has changed from the "Perl 5" regex format to another layout.
17:05 OnlineCop ugexe: :)
17:05 ugexe what do you mean by regex format? the syntax?
17:05 OnlineCop Right.
17:06 ugexe yes it has changed
17:06 OnlineCop I was looking at the "PSIX" regex syntax, and things were given as examples like    /<[+-]>/  instead of   /[+-]/
17:06 OnlineCop I don't know if that means that only "perl 6" regex has changed but "pcre2" format is the same, or if "pcre2" _is_ "perl 6" format.
17:07 OnlineCop ("pcre2" is the same as "pcre", that is)
17:07 ugexe im pretty sure perl6 regex is its own thing
17:08 OnlineCop If we don't need to change much as far as regex syntax goes for PCRE2, and only add in more features (I think there are now UTF8 or UTF16 by default, and possibly variable-length look-behinds or something...), then we'll try to account for that.
17:08 [particle] from http://www.pcre.org/readme.txt: PCRE2 is written in C, and it has its own API.
17:08 OnlineCop I'm still going over the http://www.pcre.org/current​/doc/html/pcre2syntax.html documentation with him to try to find the differences.
17:09 geekosaur that's not perl6 regex
17:09 OnlineCop Okay, so "perl6 regex" and "pcre2" are not the same?
17:09 [particle] sorry, you're looking for syntax, not api. the syntax for pcre2 is close to pcre, which is close to perl5 and not much like perl6 regexes.
17:10 geekosaur pcre2 is pcre with some additions. it's still perl5-oriented
17:10 OnlineCop http://www.pcre.org/current​/doc/html/pcre2pattern.html states "PCRE2 tries to match Perl syntax and semantics as closely as it can." and I wanted to ensure that it's not saying that "PCRE2 tries to match Perl __6__ syntax and..."
17:10 geekosaur perl6 regex is a very different thing
17:11 geekosaur pcre2 is very definitely NOT perl 6 regex
17:11 OnlineCop Okay, awesome. Thanks for helping me understand the differences.
17:11 [particle] from http://www.pcre.org/pcre2.txt: Details of exactly which Perl regular expression features are  and  are not  supported  by  PCRE2  are  given  in  separate  documents. See the pcre2pattern and pcre2compat pages. There is a syntax  summary  in  the pcre2syntax page.
17:11 canopus joined #perl6
17:12 OnlineCop http://www.pcre.org/pcre2.txt says to see the 'pcre2compat' pages, but http://www.pcre.org/pcre2compat.txt doesn't exist. :-/
17:12 rudolfochrist joined #perl6
17:13 empT joined #perl6
17:16 tbrowder joined #perl6
17:25 sufrostico joined #perl6
17:29 canopus joined #perl6
17:29 Xliff Thanks, ugexe++ # CompUnit::
17:30 cognominal joined #perl6
17:39 kjs_ joined #perl6
17:48 iH2O joined #perl6
17:57 adu joined #perl6
18:13 labster joined #perl6
18:18 iH2O left #perl6
18:25 wamba joined #perl6
18:26 Xliff are opendir and readdir supported in P6?
18:30 moritz no, dir() is
18:30 Xliff docs?
18:31 OnlineCop left #perl6
18:31 moritz https://docs.perl6.org/routine/dir
18:31 cdg joined #perl6
18:31 moritz if you really need opendir and readdir, patch https://github.com/cspencer/perl6-posix :-)
18:32 Xliff \o/
18:32 Xliff moritz++
18:32 Xliff Now I vill google "basename perl6"
18:33 moritz or just look into the IO::Path docs
18:33 moritz which contains all the path manipulation methods
18:35 Xliff Funny, I missed that in IO::Path... but maybe I was looking in IO::Handle by mistake.
18:36 Xliff But I now have a list of submodules in a given namespace. Now I need to figure out what to do with 'em.
18:39 zakharyas joined #perl6
18:50 zacts joined #perl6
19:06 ejf_ joined #perl6
19:08 espadrine joined #perl6
19:12 r_i_d joined #perl6
19:13 kjs_ joined #perl6
19:13 r_i_d Can anyone tell me why: my Int $x = Nil; $x ~~ Nil; returns false?
19:18 Xliff m: my $x = Nil; say $x ~~ Nil; my Int $y = Nil; say $y ~~ Nil;
19:18 camelia rakudo-moar 575dda: OUTPUT«False␤False␤»
19:19 Xliff I'm thinking Nil always returns False in such contexts.
19:19 Xliff But I am not quite sure.
19:19 Xliff m: say Nil ~~ Nil
19:19 camelia rakudo-moar 575dda: OUTPUT«True␤»
19:19 Xliff And that adds more confusion to the fire.
19:19 nadim joined #perl6
19:19 Xliff m: my $x = Nil; say $x ~~ Nil; my Int $y = Nil; say $y =:= Nil;
19:19 camelia rakudo-moar 575dda: OUTPUT«False␤False␤»
19:20 Xliff m: my $x = Nil; say $x ~~ Nil; my Int $y = Nil; say $y == Nil;
19:20 camelia rakudo-moar 575dda: OUTPUT«False␤Use of uninitialized value of type Int in numeric context  in block <unit> at <tmp> line 1␤Use of Nil in numeric context  in block <unit> at <tmp> line 1␤True␤»
19:20 r_i_d what was that last one you did? =:= ?
19:20 Xliff Container equivalence
19:20 Xliff m: my $x = Nil; say $x == Nil; my Int $y = Nil;
19:20 camelia rakudo-moar 575dda: OUTPUT«Use of uninitialized value of type Any in numeric context  in block <unit> at <tmp> line 1␤Use of Nil in numeric context  in block <unit> at <tmp> line 1␤True␤»
19:21 r_i_d ok, so like (Int) == (Int) ?
19:21 Xliff Yeah. Comparisons to Nil are screwy. Best to use .defined
19:21 geekosaur m: my $x = Nil; say $x ~~ Nil; my Int $y = Nil; say $y === Nil;
19:21 camelia rakudo-moar 575dda: OUTPUT«False␤False␤»
19:21 r_i_d Oh? that's perfect, I didn't know about .defined.  Thanks!
19:21 Xliff m: my $x = Nil; say !$x.defined; my Int $y = Nil; say !$y.defined;
19:21 camelia rakudo-moar 575dda: OUTPUT«True␤True␤»
19:21 benchable6 joined #perl6
19:21 bisectable6 joined #perl6
19:21 committable6 joined #perl6
19:21 Xliff ^^ That's probably what you are looking for.
19:22 r_i_d Yep! that's exactly what I need.  Thanks again!
19:22 Xliff yw! =)
19:29 canopus joined #perl6
19:37 dalek doc: d2466f4 | gfldex++ | doc/Type/Hash.pod6:
19:37 dalek doc: remove leftover conflict marker
19:37 dalek doc: review: https://github.com/perl6/doc/commit/d2466f4f19
19:37 dalek doc: 85894ad | gfldex++ | doc/Type/Nil.pod6:
19:37 dalek doc: fix typos
19:37 dalek doc: review: https://github.com/perl6/doc/commit/85894adbd0
19:52 tony-o who is retupmoca?
19:52 tony-o and also, why are there so many michiganders in p6
19:56 timotimo to find mst presentation in a live stream, just look for volume peaks :)
19:57 timotimo this wall of 4x4 TVs is ... amazing
19:58 timotimo .seen retupmoca
19:58 yoleaux I saw retupmoca 8 Jun 2016 21:09Z in #perl6: <retupmoca> m: say ('a'..'d').rotor(2 => -1).perl;
19:59 tony-o timotimo: pics?
19:59 Khisanth joined #perl6
19:59 vendethiel joined #perl6
20:00 timotimo https://youtu.be/LfrlqxzeSSE?t=3h26m12s - go back a few minutes to see it in "restart mode"
20:02 tony-o nice
20:02 tony-o i'm guessing that's MST
20:02 timotimo yup
20:05 cdg joined #perl6
20:06 El_Che good travels for those leaving yapce
20:11 timotimo i'm looking forward to getting all the changes mst made into our build system
20:13 timotimo my $$q = "'"'"'";  -  this is the best thing ever
20:16 timotimo wow, 6serv, that is amazing
20:18 kjs_ joined #perl6
20:18 timotimo mst: that presentation was very entertaining and also quite informative
20:18 mst timotimo: also, as of my lightning talk, it's even easier
20:18 mst grab the 6serv.tgz
20:18 mst unpack it
20:18 mst install App::opan
20:18 mst opan cpanm --installdeps .
20:19 tony-o this is super entertaining
20:19 mst presto, running 6serv, if nothing blows up
20:19 timotimo fantastic
20:19 timotimo when can we start grabbing your terrible ad-hoc patches and making them official parts of the build "system"?
20:19 timotimo or would you like us (or yourself) to do a bit more of a "full rewrite" ;)
20:20 mst honestly, the makefiles aren't terrible except for (1) not consistently using the variables (2) YOU NEED MORE THAN ONE SORT OF PREFIX FFS
20:20 timotimo right
20:20 timotimo i've always been more than happy that we only have a single prefix where everything goes
20:20 timotimo but that's admittedly short-sighted and egocentric
20:20 timotimo (.. as fuck)
20:20 mst the configure scripts could probably do with being re-engineered with fire
20:20 timotimo right
20:21 mst what I was going to do was get as far as shipping what I have
20:21 timotimo at some point the configure scripts are also going to be re-written in nqp and we'll perhaps ship a fully bootstrapped moar + nqp to run the configure stuff :P
20:21 mst then test an upgrade to .08 for them
20:21 timotimo does that sound insane enough? good.
20:21 mst then start figuring out how to get patches back into the makefiles
20:21 mst if the configure scripts actually passed everything they knew to the makefile that would help
20:21 mst also if the rakudo one had options at all
20:21 timotimo right
20:23 timotimo after the first months of dealing with perl6 at all, i've stopped using --gen-anything from the configure.pl anyway and just have the three check-outs provide the same thing to --prefix; i'm astounded that sysadmin you mentioned would flip table immediately instead of trying that, but i expect it wasn't actually documented in any unsurprising location at all
20:23 mst well, no, once they'd recovered from "it shat all over my filesystem during configure time" they did go back and do that
20:23 mst it just took them a while to stop being angry at having had to roll back their filesystem because of the massive violation of expectations
20:24 timotimo oh, shit, they used a system-wide --prefix?
20:24 mst (remember that those install directly into prefix, no chance for DESTDIR)
20:24 mst they were trying to setup freebsd ports
20:24 mst so it shat over /usr/local without warning
20:24 timotimo until recently i didn't even have a clue DESTDIR would be anything different from --prefix, tbh
20:24 mst DESTDIR is required for packaging
20:25 mst basically the --gen-* options need documenting as THIS WILL DO SOMETHING COMPLETELY INSANE
20:25 nbg joined #perl6
20:25 mst rather than being recommended without explanation
20:25 timotimo ah, yes, definitely
20:25 mst so, yeah, zfs snapshot rollback followed by taking a week to calm down
20:25 mst I don't feel this is unreasonable
20:26 tony-o Method 'input-line-separator' not found for invocant of class 'IO::Socket::INET'
20:26 mst I'd tend to rm -rf something that did that to me on the grounds it clearly isn't production code or anything close
20:26 mst "I'll try it again in a year, maybe they'll be less stupid" is a completely reasonable response to it to my mind
20:26 timotimo how do you feel about when --gen-nqp is used, the program will output a blinking, colorful warning that things will immediately be installed to live under the given --prefix and you have *countdown from 15* seconds to ctrl-c before your system gets augmented
20:27 mst that would certainly be a start
20:27 timotimo though as far as i know we only "pollute" the bin folder, everything else goes under nqp/ or moarvm/ or rakudo/ or perl6/ somewhere
20:27 timotimo so not a terrifying clean-up job
20:27 mst installing things, at all, without warning, during configure time
20:27 mst is ... agh
20:28 mst yes. it's not hard to clean up. but it's still *horrifying*
20:28 timotimo quite
20:29 mst "sure, I took a shit in your bathtub, but it was a solid shit so it'll be easy to carry to the toilet"
20:29 timotimo i'm adding the ", and install" part to the configure.pl helps now
20:29 timotimo and then to readmes
20:30 mst also, if we could have nqp --show-config actually say what *its* prefix is as well as moar's
20:31 mst that would simply stuff a lot
20:31 mst I should probably try and jot down exactly what I think should be changed
20:31 timotimo that'd be fantastic
20:31 mst I mean, in terms of external behaviour
20:31 [Coke] mst: thank you for being bothered enough by this to fix it. :)
20:31 mst there was no freaking way I was putting my name to a cpan dist that used the current behaviour :P
20:31 [Coke] hoping this helps other packagers as well.
20:32 timotimo https://gist.github.com/anonymous​/c0bec85694d684af98c38f4350c49e1e - is that verbiage okay?
20:32 mst no, because it still doesn't make it clear that it will install it direct to --prefix before Configure.pl completes
20:32 mst and *that* is the violation of clue
20:33 masak mst++ # not giving up on us
20:33 mst "Downloads, builds and installs NQP before completing configuration and writing Makefile"
20:33 mst might be a little clearer
20:33 timotimo oh
20:33 timotimo is putting "Immediately" in the front acceptable, too?
20:34 mst I would want 'before writing Makefile' in there as well
20:34 timotimo OK
20:34 mst this is sufficiently insane from a how-configure-make-make-insta​ll-works-in-grown-up-software POV
20:34 mst that it needs calling out really hard
20:34 timotimo i've put "before writing the Makefile" after the three options
20:35 mst that should help
20:35 mst I don't want people to be horrified at the immaturity before they get as far as actually getting rakudo installed
20:35 timotimo and i put "- at Configure time -" into the longer blob
20:36 mst we're in early days enough here that immaturity is ok, but we need a chance to hook them with fun before they realise just how immature ;)
20:36 SmokeMachine____ joined #perl6
20:37 tbrowder m: my $s = '.  1.  '; say $s
20:37 camelia rakudo-moar 575dda: OUTPUT«.  1.  ␤»
20:37 yoleaux 13:53Z <smls> tbrowder: Change «$serial = $0;» to «$serial = +$0;».
20:37 tbrowder m: say $s
20:37 camelia rakudo-moar 575dda: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Variable '$s' is not declared␤at <tmp>:1␤------> 3say 7⏏5$s␤»
20:37 mst timotimo: I mean, ideally, we'd figure out how to be able to build against an uninstalled version
20:38 mst if the --gen-* options did the configure for each thing at configure time
20:38 mst then built them in order at 'make time'
20:38 mst then 'make install' installed all three
20:38 mst they'd be totally sane
20:38 mst also, I could've had one cpan dist instead of three
20:40 mst but that probably requires more linker etc. fu than I honestly have
20:41 timotimo pushed another patch to the README.md
20:41 timotimo now i'll be AFK for a bit, and open to a bit more build system or documentation hackery later in the day, if need be
20:41 mithaldu_ joined #perl6
20:43 Lucas_One joined #perl6
20:43 nadim_ joined #perl6
20:43 tbrowder does anyone have any idea about what's wrong with the regex capture I example showed earlier?
20:44 mcmillhj joined #perl6
20:44 El_Che mst: I saw some of your talk through the streaming. "The story of my life in 50^W30minutes"
20:44 tony-o tbrowder: did smls' message not do anything for you?
20:44 tony-o 20:42 <+yoleaux> 13:53Z <smls> tbrowder: Change «$serial = $0;» to «$serial = +$0;».
20:44 mst El_Che: yeah, pretty much
20:45 El_Che mst: were you able to make up for the disappeared time? (I didn't see the end, had to leave to a meeting)
20:45 tbrowder sorry I didn't see that--thanks
20:46 mst El_Che: I finished at 29:59 apparently
20:46 El_Che impressive
20:46 mst then I had tech problems and had to compress the perl6 one which was *designed* for 30min, did that in ~22
20:46 BuildTheRobots joined #perl6
20:46 mst which still meant finishing 10min late
20:47 tony-o where are you from mst?
20:47 El_Che mst: I saw most of the perl6 talk. The configure installing stuff made me laugh out loud
20:47 El_Che a colleague came in to my office to check why :)
20:47 mst tony-o: lancashire, grew up near wigan, now in lancaster
20:49 cognominal joined #perl6
20:50 El_Che mst: takahashi slide method and faster presenting because of time contraints. I hope there weren't any epileptics in the room :)
20:51 mst :D
20:52 mst oh, yeah, and I built an Alien-Rakudo-Inline-Perl5 after the talk with the aid of a couple tweaks to its installer from nine++
20:52 kipd joined #perl6
20:52 ajr_ joined #perl6
20:53 avar mst: is that rakudo as a perl5 module?
20:53 mst avar: as three.
20:53 mst avar: http://trout.me.uk/6dists/
20:54 ajr_ How do I identify the OS under which P6 is running? I can't find it in the docs.
20:54 mst to try them out, run:
20:54 mst cpanm -n App::opan; curl -L http://trout.me.uk/6dists/
20:54 mst bah
20:54 mst cpanm -n App::opan; curl -L http://trout.me.uk/6dists/6serv.tgz | tar xzf -; cd 6serv; opan cpanm --installdeps .
20:54 tony-o m: $*OS.say;
20:54 camelia rakudo-moar bc187b: OUTPUT«Dynamic variable $*OS not found␤  in block <unit> at <tmp> line 1␤␤Actually thrown at:␤  in block <unit> at <tmp> line 1␤␤»
20:55 tony-o apparently i've been away too long
20:55 edehont joined #perl6
20:56 tony-o m: $*DISTRO.say;
20:56 camelia rakudo-moar bc187b: OUTPUT«opensuse (13.2.Harlequin)␤»
20:56 tony-o m: $*KERNEL.say;
20:56 camelia rakudo-moar bc187b: OUTPUT«linux (3.16.6.2.default)␤»
20:56 kjs_ joined #perl6
20:57 avar mst: Cool, literal did the same thing @ work pretty much, but it's just called Six::Rakudo and is one dist that contains the rakudo-star release
20:57 avar and inline-perl6 then depends on it
20:57 mcmillhj joined #perl6
20:58 mst avar: yeah, that really doesn't count. mine installs to three independent prefixes and doesn't touch install dirs during configure time
20:58 mst for in-house use, what you did is totally reasonable
20:58 mst but I was unwilling to amateur hour it :)
20:58 skids joined #perl6
20:59 olinkl joined #perl6
21:00 avar Have you discovered that you need this monkeypatch on rakudo:
21:00 avar sed -i 's/"install"/"--nodeps", "install"/' tools/build/module-install.pl
21:00 jnthn mst++ # implementing installing to separate prefixes :)
21:00 avar there's a pull request for it, otherwise panda needs to go to the network to resolve deps
21:02 mst avar: I did not, because I ignored star, because its configure is even more of a nightmare and the current release manager keeps abusing other devs on twitter and won't speak to me on IRC since I called him out for attacking other perl5 devs because he's politically right wing and treats anybody who doesn't vote conservative like they're stupid.
21:02 mst I have better things to do with my life than deal with that
21:02 tony-o lol. what?
21:02 mst also, nine gave me a manaul insaller for Inline-Perl5 so I don't need that
21:03 avar Okey I don't know how star is put together, I assumed from the 2016.07 bit that this was basically a split-up version of the same thing
21:03 mst the underlying releases have the same numbering
21:04 mst star is basically a copy-pasta of rakudo's configure.pl from aages back with some stuff added to install extra modules, and then about 1/3 of the subsequent fixes to rakudo merged down
21:04 mst I did not feel like dealing with that
21:04 avar For your MY::postamble sections: s/cp/$(CP)/g; s/perl/$(PERL)/g etc.
21:04 mst yeah, it's very much proof of concept quality
21:05 avar Anyway, looks really good, would me great to have this on the CPAN
21:05 mst that's why I didn't even upload 'em as devrels yet
21:05 avar we hacked up or own because nothing like this was there, but would totally use it if it was there
21:05 mst didn't want smokers to deal with the obvious amateur hour bits in my attempts at de-amateur-houring the stuff I'm wrapping ;)
21:06 avar Does it have any modules like DBIish, or is that a rakudo-star only thing?
21:06 mst ... my test run on $other_box just died because rakudo used 1Gb RAM building
21:06 El_Che from my docker infra use case I always pick rakudo and add modules needed by the app.
21:06 mst that's star only
21:06 notbenh_ joined #perl6
21:06 avar We build ours from star and it comes with a couple of dozen modules
21:06 mst star is meant as an example of how to include modules, from which you derive your own distribution
21:06 avar Yeah I was thinking that it would make sense to build those all as Alien-Rakudo-Module-$NAME or something, but we didn't get around to it, it's just one huge bundle
21:06 mst people packaging it was never what it was for, and every time people do that it makes bunny cry
21:07 timotimo mst: how terrible is it that moar's Configure.pl will hit the 'net to update its 3rdparty dependency submodules?
21:07 AngeloMichael joined #perl6
21:07 mst perl5's Configure needs to ssh into a system of the target arch for cross compilation
21:07 mst that sort of thing is filed under "gross but not actively evil"
21:07 jnthn timotimo: Does it actually do that for the release tarballs?
21:08 mst avar: that would be one way to do it. my plan is to build a hybrid cpan client designed for perl5/perl6 mixed apps
21:08 avar mst: Well if there's something you're actually meant to package in the perl6 universe I'd be happy to use it. No disrespect to anyone here, but it seems *really* basic from a packaging point of view.
21:08 mst but I'm attempting to do it Right so as with all things involing perl6 and 'doing it right' (like, say, perl6), don't hold your breath
21:08 avar I.e. it's all "run this ad-hoc script and it'll give you a module", not "here's something you could easily make into a distro package" like perl5
21:09 mst correct
21:09 mst I believe that what you're meant to package is the underlying actual releases, like I did
21:09 mst but remember I'm still mostly a tourist here
21:09 El_Che mst: if you're looking on things to fix on the perl6 ecosystem, are you taking requests?
21:09 mst I'm trying to de-amateur-hour things for my own purposes, but I'm starting from the bottom up
21:10 mst I'm probably going to throw a chair and fix the freaking CUR format to not use freaking sha1s for the filenames
21:10 avar Yeah same, that didn't seem so hard, it's just some C code talking to other C code, but I haven't found how to get stand-alone perl6 modules in some .tar.gz format or whatever that I can unpack, build, and package up as an rpm or whatever
21:10 mst jnthn: btw, remember I said the sha1s would horrify unix people, and had to explain why?
21:11 avar The same unix people who are using git for all their source control ?:)
21:11 mst jnthn: I met a sysadmin at YAPC::EU who'd tried to get permission to use perl6 and discovering the sha1s was the point at which their boss went 'this breaks locate, this is not production ready, you're banned from using it'
21:11 mst avar: my git *checkout* has lib/Foo.pm in it
21:11 tony-o the sha1 file names for precomp?
21:11 mst avar: my perl6 site directory has a sha1 file instead :(
21:11 timotimo jnthn: i don't think it does. i haven't looked if it touches the dependency git repos if you use --has-lib* in configure.pl; that might be worth a look-see. it's probably already fine, though
21:12 mst tony-o: for CUR
21:12 tony-o yea
21:12 tony-o should've voted for the extra file extension
21:12 tony-o rather than the sha1
21:12 timotimo i was phantasizing about a bunch of docker-using scripts/thingies that'll go through a bunch of different use cases for our configure scripts and tries to verify we handled things properly
21:12 mst I was thinking about much the same thing, except with less CADT
21:13 timotimo i.e.: regular user, installs under /home; packager, has a DESTDIR and fakeroot; embedded device developer, has to cross-compile for two different architectures; ...
21:13 mst yes
21:13 timotimo what is CADT?
21:13 mohae_ joined #perl6
21:14 mst timotimo: https://www.jwz.org/doc/cadt.html
21:14 mcmillhj joined #perl6
21:14 timotimo ugh :)
21:14 timotimo i remember that text
21:14 timotimo your "i was thinking about much the same thing" was in response to my "phantasizing" bit?
21:15 mst yes
21:15 firstdayonthejob joined #perl6
21:15 mst just docker is currently in the CADT category for me
21:15 jnthn mst: Did things already get changed so the SHA-1 is in the path, but the filenames are in a directory under that? But, sigh, you can't please everyone.
21:15 mst jnthn: no
21:15 mst jnthn: well, it's enraging me as well
21:15 tony-o the extension thing was on the table before sha1 came about
21:15 ugexe need a internal puny encoder or some such to keep original file names
21:15 timotimo i was picking out docker in particular because it's more or less a solution to "i want to have a scratch filesystem with some distro installed and be able to just throw out the complete thing with the touch of one button when i'm done with it"
21:16 jnthn mst: Good. That means it'll get solved. :)
21:16 jnthn (Not by me, though. Phew. :))
21:16 timotimo heh :)
21:16 * jnthn is *very* glad that other folks are now taking this stuff on, and only wish they'd done so about 6 months sooner than they did
21:17 timotimo who knew the solution would be to just throw some cash in mst's face :)
21:17 mst the cash was a while back
21:17 mst my problem was mostly lack of motivation
21:17 tony-o it's difficult to want to when PRs are rejected and that code magically ends up in the repo anyway
21:17 timotimo i'd just have to eat only ramen out of plastic bag-like containers for a month and then any bug that annoys me could get fixed by An Actual Professional!
21:17 mst lizmat's 'perl5 is the new perl4' lightning talk was basically a giant fuck you to my entire community and put me off for quite a while
21:18 tony-o mst: when was that?
21:18 jnthn Well, I see we're back to whining, so I'll go have fun. :)
21:18 mst jnthn: meh
21:18 mst jnthn: you whined I didn't help earlier
21:18 mst jnthn: I told you why :P
21:18 jnthn :P
21:19 ajr_ $*OS and $*OSNAME both fail to identify the OS. Any other suggestions, anyone?
21:19 mst also, y'know, if anybody'd included the perl5 toolchain team in any of this we could've helped it not be terrible in the first place
21:20 mst jnthn: after chasing you into talking to the moose team and that helping, I sort of thought people might've worked that out when they got to the CUR implementation parts
21:20 timotimo ajr_: $*KERNEL and $*DISTRO?
21:20 mst and then I didn't notice until later
21:20 mst cos I also suck
21:21 awwaiid never too late
21:21 avar So maybe it's time for a perl6-porters?:)
21:22 ufobat joined #perl6
21:22 timotimo we did have #perl6-toolchain and #perl6-release
21:22 timotimo no clue if they're actively being used
21:22 ajr_ timotimo: those discriminate between Linuxes - how does one identify Windows (generically)?
21:22 ugexe m: say $*DISTRO.is-win
21:22 camelia rakudo-moar bc187b: OUTPUT«False␤»
21:23 ajr_ Failing that, what's the generic name for those variables, so I can find them in the docs?
21:23 timotimo right
21:23 mcmillhj joined #perl6
21:23 timotimo .is-win is the right one
21:23 gfldex ajr_: you poke around in the registry
21:23 mst timotimo: yes, we have #perl6-toolchain because I got fucked off and created it.
21:23 timotimo it won't tell you "windows" if you're running ubuntu on your windows 10, though
21:24 timotimo at least i don't think it will
21:24 timotimo but with ubuntu-inside-win10, you aren't as royally fucked as you are on windows on general
21:24 timotimo so maybe it shouldn't give True for .is-win
21:24 ajr_ The original swamp I was trying to drain was telling the user what EOF character to use to terminate input
21:24 gfldex ajr_: https://docs.perl6.org/language/​variables#Compile-time_variables
21:25 timotimo ajr_: did you know we do some automatic newline stuff For You?
21:25 mst timotimo: depends if they've fixed stat() yet
21:25 ItayAlmog joined #perl6
21:25 timotimo mst: i'm personally not going to touch windows for development with a quite long stick, so ... :)
21:25 timotimo i'm not even following the intricate details of that thing
21:25 mst ok, well, basically, at the moment it's more 'differently fucked'
21:26 ajr_ gfldex - thanks
21:26 lizmat joined #perl6
21:26 ajr_ timotimo: not sure I understand that
21:26 jdv79 this is the first i'm hearing of a p5 is the next p4.  odd.
21:26 tony-o was p4 the next p3?
21:27 timotimo ajr_: what exactly?
21:27 jdv79 i started using perl at about 5.6 or whatever was just before so i have no idea
21:27 mst jdv79: it was a lightning talk, the theme of which was "you're obsolete and shit and we're you're only hope for the future"
21:28 ajr_ The "automatic newline" stuff
21:28 mst I was not particularly happy about it
21:28 mst you're welcome to go find a video, I'm not going to discuss it further
21:28 nadim_ joined #perl6
21:28 avar mst: What day was it on?
21:29 mst it was a while ago, and I don't even remember which conference, and honestly I've done my best to forget about it because I'd like to get over it and keep trying to bring the two communities together in spite of the periodic acts of idiocy like that from both sides, especially give right now I think the perl5 community is producing acts of idiocy of such forms rather more often ;)
21:29 timotimo ajr_: ah, ok. i don't know how exactly it works nowadays; i think we translate at program boundary and inside perl6, \n refers to a "logical newline" and for the other meanings you use \x20 (or what ever the right ascii code is) and optionally the \x?? for \r
21:30 avar Ah, I thought you mean this YAPC that just finished
21:30 timotimo maybe perl6 is only now becoming interesting enough for regular perl5 users that they need to panic or something
21:31 ajr_ timotimo: I entirely agree with you about development on Windows, but there's a lot of it about, and I'm trying to write generic training examples.
21:31 jdv79 mst: thanks for your extensive efforts, btw.  much appreciated.
21:31 timotimo right
21:31 timotimo i wonder how hard it'd be to get a camelia running on windows so we can just try shit out on the channel
21:32 timotimo it'd potentially be expensive to get a windows host?
21:32 timotimo hm. isn't windows 10 for vms available for free?
21:32 woolfy joined #perl6
21:32 mcmillhj joined #perl6
21:33 timotimo i really don't want to stick a hand in that particular hornet's nest, though
21:33 nadim joined #perl6
21:36 timotimo is there a pure-perl5 implementation of git that'd allow us to fetch our submodules during build time and maybe also for panda and such?
21:37 timotimo hm. actually ... shouldn't we just be able to use github's "give me a zip file of this repo at this commit" as a replacement for basically everything we use git for during build and panda installs?
21:37 avar yes, you should
21:37 ugexe zef can already do that for module installs
21:38 timotimo very good. can you point that out on the mailing list thread i just posted to?
21:39 timotimo Message-Id: d7a537d8-a98d-111f-cbc0-57a7587efef1@wakelift.de
21:40 ugexe pure perl5 is possible via UA and simple api-url transforms
21:40 ugexe https://github.com/ugexe/Perl6-Distri​bution--Common--Remote/blob/master/li​b/Distribution/IO/Remote/Github.pm6 uses this method
21:40 timotimo ugexe: i meant a pure-perl git implementation :)
21:40 timotimo which would require parsing blobs, trees, and commits
21:41 ugexe you can fetch all of those from the web ai i believe
21:42 timotimo ah, yes
21:42 timotimo only if your server has recently run git-update-webserver-data-something-something
21:42 timotimo but git probably has something on their server that implements the whole thing itself
21:42 timotimo rather than relying on the stuff living on the filesystem
21:42 timotimo at least i'd do it along those lines
21:42 ugexe https://api.github.com/repos/rakud​o/rakudo/git/trees/nom?recursive=1 # for brevity
21:43 timotimo oh, interesting!
21:48 ajr_ This my $EOF = "CTRL-" ~ ($*DISTRO.is-win ?? "Z" !! "D"); appears to do what I want - at least on Linux - thanks
21:48 timotimo ah, nice.
21:51 mcmillhj joined #perl6
21:52 ItayAlmog is it ok that my compiler will have a limit of 10 params when creating and using subs or methods?
21:53 ajr_ left #perl6
21:53 timotimo that sounds like an extremely tight limit
21:56 timotimo m: sub a($,$,$,$,$,$,$,$,$,$,$,$,​$,$,$,$,$,$,$,$,$,$,$,$,$) { }; a(1,2,3,4,5,1,2,3,4,5,1,2,​3,4,5,1,2,3,4,5,1,2,3,4,5)
21:56 camelia rakudo-moar ebf34a: ( no output )
21:56 timotimo why would there be a static limit?
21:56 vendethiel .oO( there's only so much RAM )
21:57 yoleaux 22 Jul 2016 23:32Z <cog_> vendethiel: thx for the URL
21:57 tbrowder .tell smls: thanks! the "+$0" did the trick
21:57 yoleaux tbrowder: What kind of a name is "smls:"?!
21:57 vendethiel yoleaux: it's about time you learn how to remove that trailing :.
21:57 ItayAlmog I am using (as of right now) 10 registers which are for general use as the way to pass params
21:57 vendethiel ItayAlmog: cdecl will pass arguments via stack starting from the 8th one, IIRC :)
21:58 timotimo oh hey ven
21:58 timotimo how are you doing?
21:58 vendethiel pretty good! how about you?
21:58 tbrowder .tell smls thanks! the "+$0" did the trick
21:58 yoleaux tbrowder: I'll pass your message to smls.
21:58 timotimo a bit meh
21:58 vendethiel aw
21:58 ItayAlmog This can be a good Idea using the stack for parameters, I will try and make so it will use both :D
21:58 ItayAlmog thanks
21:58 vendethiel np, gl :-)
21:59 vendethiel timotimo: what have you been up to?
21:59 timotimo this and that
21:59 vendethiel you mean $. and $!
21:59 timotimo un-bitrotting my GTK::Simple-based shooter recently
21:59 timotimo i wanted to use our scintilla module for the code input thingie, but it doesn't seem to have an API for "what are the contents?" right now
22:00 timotimo so that was ... not helpful :D
22:00 kjs_ joined #perl6
22:00 vendethiel I remember that bit
22:00 * vendethiel gets a PTSD flashback everytime he reads "GTK::Simple" :-P
22:01 timotimo sorry about that :(
22:01 vendethiel don't be! it was so much fun.
22:01 timotimo i'm not saying it's your fault for using Mac OS X, but ... :\
22:02 vendethiel yeah, I'm sorry for wanting a desktop. :P
22:02 mst Mac OS X makes for a great OS for a TV
22:02 mst I don't understand people who try to use it as a development machine ;)
22:03 vendethiel I don't understand people who try to use linux as a desktop machine :).
22:03 mst neither do I, I'm sat at a windows+cygwin system with a freebsd VM
22:03 vendethiel you can be happy you managed to watch a video only after 5 hrs of fiddling with your settings, I can be happy stuff apple wants to work works :P
22:03 dogbert17 joined #perl6
22:03 mst what?
22:03 mst don't accuse me of being one of them
22:04 vendethiel sorry; sorry!
22:04 * mst switched back to windows because cygwin's X11 support was better than OS X
22:04 mst meanwhile, I have managed to hit exactly 140 characters and have just tested this on a fresh user on a dev server
22:04 mst https://twitter.com/shadowcat_​mst/status/769294212365946883
22:04 mst I might be rather proud of that tweet :)
22:05 masak \o/
22:05 masak mst++
22:05 vendethiel (and there's no |sh ! mst++)
22:05 timotimo \o/
22:05 mst that's one of the few things in this world that actually triggers my gag reflex.
22:06 vendethiel echo "let the hate"|you
22:12 leont joined #perl6
22:50 timotimo vendethiel: did you read about the many multithreading stability improvements made recently?
22:51 skids joined #perl6
22:56 vendethiel timotimo: yes!
22:57 vendethiel I'm reading pretty much every single commit made by you++, lizmat++, jnthn++ (and a few others :-).)
22:58 vendethiel I managed to sneak in Perl 6 in code generation, internal tools and some other areas. But nothing "running" yet...
23:01 vendethiel timotimo: I still have a project that just crashes (moarvm gc: invalid status) using GTK::Simple. Did that ever happen to you?
23:01 vendethiel (using supply.schedule-on and GTK::Simple::Scheduler)
23:06 timotimo not once :(
23:07 vendethiel i do have a repro repo: https://github.com/vendethiel/moarv​m-gcerr-gtk/blob/master/isolated.p6 if you ever get some time to look at it :)
23:08 zacts joined #perl6
23:09 timotimo that still crashes with very latest everything, i take it?
23:10 vendethiel it was crashing around a month ago, before I left for some holidays. I'll need to re-try with latest everything.
23:13 timotimo huh, the $c-o supply won't even ever emit anything at all?
23:14 vendethiel it loads for a few (dozens) seconds, then when the window starts loading, it prints "Invalid GC status observed" and crashes
23:14 Lucas_One joined #perl6
23:15 timotimo it seems a bit smelly that you're changing the value of a TextView before actually running the app itself?
23:15 AngeloMichael joined #perl6
23:15 timotimo the loading for a few dozen seconds could very well be startup overhead
23:15 vendethiel yeah
23:15 timotimo i recently made a measurement where a program with nothing but "use GTK::Simple" spends about 50% of its time in re-parsing the same json string over and over again a few hundred times
23:16 timotimo but that's only the first time running after installing
23:16 timotimo nine said we could change the implementation so it wouldn't have to parse the json stuff even a single time
23:16 vendethiel sounds amazing :-)
23:16 jnthn vendethiel: Some of the things I've fixed in the last couple of weeks stand a chance of fixing the memory corruption it's detected there.
23:17 vendethiel jnthn: that's amazing to hear! I'll be busy with a C school project this weekend, but I'll try to see if it's now working
23:17 jnthn vendethiel: I found a few nasty things going on in a couple of concurrency constructs where GC safe-points were mishandled
23:17 MilkmanDan joined #perl6
23:19 olinkl joined #perl6
23:20 Possum joined #perl6
23:25 zacts joined #perl6
23:28 kipd joined #perl6
23:39 mithaldu_ joined #perl6
23:42 tbrowder p6: my uint $n = 4
23:42 camelia rakudo-moar 6a8278: ( no output )
23:43 tbrowder p6: my uint64 $n = 3; say $n
23:43 camelia rakudo-moar 6a8278: OUTPUT«3␤»
23:44 AlexDaniel not sure why would you do that today though
23:44 AlexDaniel m: my uint64 $n = 3; say $n; $n -= 5; say $n
23:44 camelia rakudo-moar 6a8278: OUTPUT«3␤-2␤»
23:51 tbrowder I need to convert an Int to an unsigned in to compare output from a hash program to a p6 hash function
23:53 avuserow_ m: my uint64 $n = 3; say $n; $n -= 5; say $n; say $n.WHAT
23:53 camelia rakudo-moar 6a8278: OUTPUT«3␤-2␤(Int)␤»
23:54 avuserow_ never mind, that didn't do what I thought it would
23:54 tbrowder the p6 Digest::xxHash module wraps the C xxhash library and converts unsigned ints to p6 Ints; I want to compare the two outputs somehow
23:58 Skarsnik_ err, why $n type get changed
23:59 tbrowder p6: my uint $n = 7; say sprintf "%x", $n
23:59 camelia rakudo-moar 6a8278: OUTPUT«7␤»
23:59 Skarsnik probably because uint64 is a bastard type unlike unit?

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

Perl 6 | Reference Documentation | Rakudo