Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2016-09-02

Perl 6 | Reference Documentation | Rakudo

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

All times shown according to UTC.

Time Nick Message
00:05 mcmillhj joined #perl6
00:05 labster joined #perl6
00:06 itaipu joined #perl6
00:17 mcmillhj joined #perl6
00:20 aries_liuxueyang joined #perl6
00:21 pierre_ joined #perl6
00:31 mcmillhj joined #perl6
00:31 BenGoldberg joined #perl6
00:42 mcmillhj joined #perl6
00:51 ugexe .tell pmurias re: `zef clone Foo::Bar` question from last week - this mostly exists already using either `zef fetch Foo::Bar` or `zef look Foo::Bar`. If it happens that Foo::Bar resolves to a .tar.gz (like from cpan) and not a git uri it will extract it for you as well
00:51 yoleaux ugexe: I'll pass your message to pmurias.
00:55 mcmillhj joined #perl6
01:01 bstamour /?
01:06 mcmillhj joined #perl6
01:06 ugexe ?$/
01:17 mcmillhj joined #perl6
01:20 itaipu joined #perl6
01:25 pierre_ joined #perl6
01:31 pierre_ joined #perl6
01:36 mcmillhj joined #perl6
01:49 bob777 joined #perl6
01:50 gfldex m: my @tests = { say 1; False }, { say 2; True }, { say 3; True }; say [&&] @tests;
01:50 camelia rakudo-moar d2b115: OUTPUT«2␤3␤True␤»
01:51 gfldex i would expect the first block to be executed as well
01:51 mcmillhj joined #perl6
01:51 gfldex and it's return value to be taken into consideration
01:54 pierre_ joined #perl6
02:06 BenGoldberg m: my @tests = { say 1; False }, { say 2; True }, { say 3; True }; @tests.say
02:06 camelia rakudo-moar d2b115: OUTPUT«[-> ;; $_? is raw { #`(Block|80637344) ... } -> ;; $_? is raw { #`(Block|80637416) ... } -> ;; $_? is raw { #`(Block|80637488) ... }]␤»
02:06 BenGoldberg m: my @tests = { say 1; False }, { say 2; True }, { say 3; True }; so $_ for @tests;
02:06 camelia rakudo-moar d2b115: OUTPUT«WARNINGS for <tmp>:␤Useless use of "so " in expression "so $_" in sink context (line 1)␤»
02:06 BenGoldberg m: my @tests = { say 1; False }, { say 2; True }, { say 3; True }; $_() for @tests;
02:07 camelia rakudo-moar d2b115: OUTPUT«1␤2␤3␤»
02:07 BenGoldberg m: my @tests = { say 1; False }, { say 2; True }, { say 3; True }; say all(@tests);
02:07 camelia rakudo-moar d2b115: OUTPUT«all(-> ;; $_? is raw { #`(Block|69009920) ... }, -> ;; $_? is raw { #`(Block|69009992) ... }, -> ;; $_? is raw { #`(Block|69010064) ... })␤»
02:07 BenGoldberg m: my @tests = { say 1; False }, { say 2; True }, { say 3; True }; say so all(@tests);
02:07 camelia rakudo-moar d2b115: OUTPUT«True␤»
02:07 BenGoldberg m: my @tests = { say 1; False }, { say 2; True }, { say 3; True }; say [&&] @tests;
02:07 camelia rakudo-moar d2b115: OUTPUT«2␤3␤True␤»
02:08 mempko joined #perl6
02:08 BenGoldberg m: my @tests = { say 1; False }, { say 2; True }, { say 3; True }; say [&&] !@tests;
02:08 camelia rakudo-moar d2b115: OUTPUT«False␤»
02:08 BenGoldberg m: my @tests = { say 1; False }, { say 2; True }, { say 3; True }; say [&&] |@tests;
02:08 camelia rakudo-moar d2b115: OUTPUT«2␤3␤True␤»
02:08 BenGoldberg I don't understand why any of those are getting called.
02:09 gfldex m: say Regex ~~ Callable;
02:09 camelia rakudo-moar d2b115: OUTPUT«True␤»
02:09 BenGoldberg Is it a bug?
02:09 gfldex that could be the reason
02:09 BenGoldberg ?
02:09 mcmillhj joined #perl6
02:09 * BenGoldberg doesn't get it.
02:12 gfldex my guess is there is a ~~ somewhere and that is getting the Block called
02:12 gfldex m: my @tests = { say 1; True }, { say 2; True }, { say 3; True }; multi sub tester(&test){ test() }; multi sub tester([]){ True }; multi sub tester([&test, *@rest]) { test() && tester(@rest) }; say so tester(@tests);
02:12 camelia rakudo-moar d2b115: OUTPUT«1␤2␤3␤True␤»
02:13 gfldex that works as a short circuiting &&
02:13 gfldex it's a bit long tho
02:14 canopus joined #perl6
02:14 Nelsen joined #perl6
02:14 Nelsen Good evening, mates
02:14 Nelsen I was wondering if there was a "Codecademy"-like site to learn Perl?
02:25 mcmillhj joined #perl6
02:32 itaipu joined #perl6
02:38 noganex_ joined #perl6
02:38 mcmillhj joined #perl6
02:42 zostay m: say "test.foo".trans('.' => '::');
02:42 camelia rakudo-moar d2b115: OUTPUT«test:foo␤»
02:42 zostay m: say "test.foo".trans(['.'] => ['::']);
02:42 camelia rakudo-moar d2b115: OUTPUT«test::foo␤»
02:42 zostay that seems like a bug
02:44 MasterDuke zostay: i got bit by that too, but it's in the docs
02:46 zostay then the docs are too subtle for me, i don't read that in the docs at all
02:48 MasterDuke well, it's not really spelled out here ( https://docs.perl6.org/routine/trans ), but it's shown in the examples
02:48 canopus joined #perl6
02:49 zostay which example?
02:50 gfldex examples ain't no roast
02:50 gfldex it's a bug
02:50 MasterDuke $str.=trans( [ '<'   , '>'   , '&' ] =>              [ '&lt;', '&gt;', '&amp;' ]);
02:50 gfldex the examples got it because they are meant to work with rakudo
02:50 gfldex zostay: please rakudobug
02:51 zostay that does not say that $str.=trans('<' => '&lt;', '>' => '&gt;'); shouldn't work...thx, gfldex
02:52 MasterDuke https://github.com/perl6/roast/blob/m​aster/S05-transliteration/trans.t#L55
02:53 MasterDuke (not that i like the current behavior, just playing devil's advocate here)
02:53 tala joined #perl6
02:53 zostay at the very least, the docs are incomplete... but i think there's an opportunity to call this a bug still ;)
02:55 gfldex it should either work or complain if it got a multi char replacement. Silently dropping stuff is LTA.
02:57 MasterDuke doesn't that match Perl 5's behavior?
03:00 zostay gfldex, that's what i just suggested on the bug i posted... either warn and fix docs or make the behavior consistent (with my personal vote being for the latter)
03:01 zostay i don't think there's an analogous function to .trans in perl5, tr/// does not have comparable calling conventions at all
03:01 zostay .trans() is awesome as a multiple string find and replace at the same time thing
03:13 canopus joined #perl6
03:22 mcmillhj joined #perl6
03:24 canopus joined #perl6
03:27 djbkd joined #perl6
03:39 canopus joined #perl6
03:43 pierre_ joined #perl6
03:49 rodarmor joined #perl6
03:50 mcmillhj joined #perl6
03:57 canopus joined #perl6
04:06 AndyBotwin joined #perl6
04:06 AndyBotwin joined #perl6
04:09 pierre_ joined #perl6
04:09 labster joined #perl6
04:09 mempko joined #perl6
04:22 mcmillhj joined #perl6
04:22 pierre_ joined #perl6
04:23 canopus joined #perl6
04:27 pierre_ joined #perl6
04:34 TimToady R3b0zak.
04:34 TimToady dur
04:35 TimToady so much for that password :)
04:36 grondilu lol
04:36 pierre_ joined #perl6
04:37 grondilu hey TimToady did you have a look at https://github.com/perl6/roast/issues/147 ?
04:39 mcmillhj joined #perl6
04:41 TimToady I don't see how to allow that, offhand, given the list is evaluated before anything is declared
04:41 TimToady something fiddly like that probably wants to use constants anyway
04:42 grondilu then it wouldn't be a list but a syntactic construct.  Like in C I suppose.
04:42 cpage_ joined #perl6
04:42 TimToady in p6 it's just a general list, which you could generate from any compile-time values
04:43 grondilu then allow a syntax that differ from the syntax that uses a list.
04:43 TimToady it seems like kind of a rare thing to want
04:43 grondilu :/
04:44 grondilu that may be true.
04:44 TimToady bein's you're the first person to ask for it in 15 years...
04:44 grondilu yet it bothered me badly that as I was looking at a C header I realized I could not make a straightforward translation.
04:46 grondilu (well, "bothered me badly" that was maybe strong an expression)
04:46 grondilu it was frustrating though.
04:47 nadim joined #perl6
04:47 grondilu also, it's not as if NativeCall has existed for 15 years, is it?
04:52 mcmillhj joined #perl6
04:53 ItayAlmog2805 joined #perl6
04:54 goldfax joined #perl6
05:22 wamba joined #perl6
05:26 mcmillhj joined #perl6
05:49 Woodi joined #perl6
05:50 brrt joined #perl6
05:56 mcmillhj joined #perl6
06:05 woolfy left #perl6
06:13 dami joined #perl6
06:17 firstdayonthejob joined #perl6
06:28 ShimmerFairy fwiw I ran into the same issue once when using enums in P6. The use case is typically to define "synonyms" for some reason or another.
06:29 ShimmerFairy I'd also like to tell enums not to pollute the namespace with unqualified names, while we're at it.
06:29 pierre_ joined #perl6
06:32 moritz ShimmerFairy: you might be able to do that with anonymous enums
06:33 jjido joined #perl6
06:33 moritz constant EnumName = enum <a b c>;
06:33 moritz or something like that
06:33 moritz haven't tested it
06:40 domidumont joined #perl6
06:41 mcmillhj joined #perl6
06:41 jjido joined #perl6
06:41 * grondilu is not sure about what difference there is between constant EnumName = enum <a b c> and enum EnumName <a b c>
06:42 grondilu m: constant FooBar = enum <Foo Bar>
06:42 camelia rakudo-moar d2b115: ( no output )
06:42 grondilu m: constant FooBar = enum <Foo Bar>; say FooBar.WHAT
06:42 camelia rakudo-moar d2b115: OUTPUT«(Map)␤»
06:42 grondilu m: enum FooBar <Foo Bar>
06:42 camelia rakudo-moar d2b115: ( no output )
06:42 grondilu m: enum FooBar <Foo Bar>; say FooBar.WHAT
06:42 camelia rakudo-moar d2b115: OUTPUT«(FooBar)␤»
06:44 domidumont joined #perl6
06:45 lizmat joined #perl6
06:48 jjido joined #perl6
06:51 ShimmerFairy moritz: not synonyms to the enum, but synonyms for its values. In C++, something like e.g.  enum class Feature { Off, OneChoice, TwoChoice, On = OneChoice };  for a feature that used to be binary.
06:54 shinobicl_ joined #perl6
06:54 shinobicl_ m: sub postcircumfix:<' '>($it, $arg, :$double) { say $double ?? 2 * $it * $arg !! $it * $arg }; 2'3'; 2'3':double
06:54 camelia rakudo-moar d2b115: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Unable to parse expression in postcircumfix:sym<' '>; couldn't find final $stopper ␤at <tmp>:1␤------> 3 ?? 2 * $it * $arg !! $it * $arg }; 2'3'7⏏5; 2'3':double␤»
06:55 ribasushi joined #perl6
06:56 mcmillhj joined #perl6
06:56 moritz ShimmerFairy: I wasn't talking about the synonyms aspects, rather about not polluting your namespace with short names
06:56 jjido joined #perl6
06:57 ShimmerFairy ah
07:02 zengargoyle joined #perl6
07:05 masak morning, #perl6
07:05 moritz \o masak
07:06 masak m: enum Feature <Off OneChoice TwoChoice>; constant On = OneChoice; say On
07:06 camelia rakudo-moar d2b115: OUTPUT«OneChoice␤»
07:06 ItayAlmog joined #perl6
07:08 ShimmerFairy m: enum Feature <Off OneChoice TwoChoice>; constant On = OneChoice; say Feature::On  # if you, like me, don't much care for the namespace pollution of 'traditional' enums
07:08 camelia rakudo-moar d2b115: OUTPUT«Could not find symbol '&On'␤  in block <unit> at <tmp> line 1␤␤Actually thrown at:␤  in block <unit> at <tmp> line 1␤␤»
07:10 masak you defined `On`, not `Feature::On`
07:11 masak there's nothing magical going on with that `constant` statement. it's just a name binding.
07:11 ShimmerFairy masak: right, I'm just pointing out that I'd _like_ to be defining Feature::On instead :)
07:12 masak m: enum Feature (Off => 0, OneChoice => 1, TwoChoice => 2, On => 1); say Feature::On
07:12 camelia rakudo-moar d2b115: OUTPUT«On␤»
07:13 ShimmerFairy The issue with that though is that it's not semantically correct, and that it requires remembering that you have to keep the two options in sync.
07:16 leont joined #perl6
07:17 masak you *almost* sound like you have a spec patch coming up :)
07:20 moritz and a roast and rakudo patch :-)
07:21 masak and lots of good feedback on the respective PRs \o/
07:22 ShimmerFairy fwiw, <TimToady> I don't see how to allow that, offhand, given the list is evaluated before anything is declared
07:22 ShimmerFairy and there's already a roast issue here :P  https://github.com/perl6/roast/issues/147
07:22 masak so, you're saying what you want to do is only impossible once? :P
07:24 djbkd joined #perl6
07:24 masak fwiw, grondilu's `enum ("foo", bar => foo);` can't work -- "foo" is just a string, and so there is no `foo` identifier declared when using it as a pair value
07:24 masak guess this is exactly what TimToady is pointing out
07:24 ShimmerFairy nah, I personally think that enums don't necessarily need to keep that behavior :) . In fact I think it would be nice if they worked like function signatures:  sub foo($a, $b = $a) { ... }
07:25 ShimmerFairy I can see why that example in particular doesn't work, though.
07:26 masak that `("foo", bar => foo)` thing is literally a list expression. much less magical than a parameter list.
07:26 masak enums currently ride on the fact that it gets lists in various forms.
07:26 ShimmerFairy right, but considering it's a 'package'-style thing to declare, maybe they could stand to be more magical?
07:28 ShimmerFairy masak: I ultimately think that Perl 6's enums are poorly designed. Sure they have type safety, and EnumClass($out-of-range) won't produce a valid defined object (like it does in C++), but they don't really feel worthwhile to me otherwise.
07:28 masak that opinion is fine.
07:29 masak I disagree, though.
07:29 masak you should have seen back when they *were* badly designed :P
07:30 ShimmerFairy masak: In C++ land, I always use 'enum class' over 'enum', where my only sticking point is that doing that kind of conversion I mentioned will live perfectly fine. But otherwise I like them :)
07:30 Juerd joined #perl6
07:32 djbkd joined #perl6
07:33 masak ShimmerFairy: maybe macros will be able to help you in the future.
07:33 masak I love being able to *say* that with enough confidence nowadays! :D
07:33 labster joined #perl6
07:35 leont ShimmerFairy: yeah, enum class is quite awesome :-)
07:35 ShimmerFairy masak: maybe so, as long as it's easy to fix broken-to-me language aspects, there's not much reason to complain :)
07:36 ShimmerFairy leont: agreed, the only issues I have are of the kind that you get for a low-level language: "type conversion doesn't check if the int is in range!"    "why won't you give me a postfix:<++> that DWIMs!?"
07:36 ShimmerFairy (and even on the latter one, I personally don't see it as a really big issue)
07:37 leont The main thing I miss is methods on enum classes ;-)
07:38 ShimmerFairy Well, introspection is a separate issue :P
07:38 ShimmerFairy (imagine if C++ had Signature and Namespace types :P)
07:39 stigo joined #perl6
07:39 leont Signatures can be largely done at compile-time, actually
07:42 ShimmerFairy Honestly, I find C++ to be just as interesting as Perl 6, but coming from the opposite direction on the low-level/high-level ladder
07:45 leont The interesting thing about C++ is how much it is still in development
07:45 leont C++11 like Perl6 took way too long, but it worth it
07:45 leont At least with the new versions they're iterative now. C++14 was a nice minor update and C++17 will bring new awesomeness
07:45 ShimmerFairy leont: and C++11 taught them to maintain their current faster release cycle too, so that's good :)
07:46 mcmillhj joined #perl6
07:51 mohae_ joined #perl6
07:53 llfourn joined #perl6
08:00 mcmillhj joined #perl6
08:10 zakharyas joined #perl6
08:13 fridim___ joined #perl6
08:16 pmurias joined #perl6
08:24 bjz joined #perl6
08:32 mcmillhj joined #perl6
08:32 rindolf joined #perl6
08:41 bobv joined #perl6
08:46 mcmillhj joined #perl6
08:53 pierre_ joined #perl6
08:54 bjz joined #perl6
08:57 pierre_ joined #perl6
08:59 bjz_ joined #perl6
09:04 mcmillhj joined #perl6
09:13 balajips joined #perl6
09:28 eliasr joined #perl6
09:37 RabidGravy joined #perl6
09:48 mcmillhj joined #perl6
09:52 El_Che wut? C++17? What shall I do with my 1996 book?
09:57 MorayJ joined #perl6
10:01 DrForr Burn it, of course. Everything magically corresponds to the new spec.
10:07 mcmillhj joined #perl6
10:12 TheLemonMan joined #perl6
10:17 user9 joined #perl6
10:18 mcmillhj joined #perl6
10:26 fridim___ joined #perl6
10:31 lizmat joined #perl6
10:35 mcmillhj joined #perl6
10:36 rurban joined #perl6
10:42 kaare__ joined #perl6
10:50 pierre_ joined #perl6
10:50 woolfy joined #perl6
11:02 dustinm`_ joined #perl6
11:02 MilkmanD1n joined #perl6
11:07 sakuya joined #perl6
11:09 mcmillhj joined #perl6
11:09 eyck joined #perl6
11:09 nine joined #perl6
11:09 imcsk8 joined #perl6
11:09 Ulti joined #perl6
11:09 hanekomu joined #perl6
11:09 mtj_ joined #perl6
11:09 jdv79 joined #perl6
11:09 broquaint joined #perl6
11:09 shadowpaste joined #perl6
11:09 xdbr joined #perl6
11:09 TheDir_ joined #perl6
11:09 jeek joined #perl6
11:09 gypsydave5 joined #perl6
11:09 jast joined #perl6
11:10 mtj__ joined #perl6
11:10 jervo joined #perl6
11:11 mtj_ joined #perl6
11:11 xinming joined #perl6
11:19 xinming_ joined #perl6
11:19 mcmillhj joined #perl6
11:22 andreoss joined #perl6
11:22 lizmat joined #perl6
11:25 iH2O joined #perl6
11:29 andreoss m: multi MAIN() { die $?USAGE }; multi MAIN(Str :$option) { say "hi "};
11:29 camelia rakudo-moar d2b115: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Variable '$?USAGE' is not declared␤at <tmp>:1␤------> 3multi MAIN() { die 7⏏5$?USAGE }; multi MAIN(Str :$option) { sa␤»
11:29 andreoss m: multi MAIN() { die &?USAGE }; multi MAIN(Str :$option) { say "hi "};
11:29 camelia rakudo-moar d2b115: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Undeclared name:␤    ?USAGE used at line 1␤␤»
11:31 andreoss m: multi MAIN() { die $*USAGE }; multi MAIN(Str :$option) { say "hi "};
11:31 camelia rakudo-moar d2b115: OUTPUT«hi ␤»
11:32 andreoss m: multi MAIN() { die $*USAGE }; multi MAIN(Str :$option) { say $option.perl };
11:32 camelia rakudo-moar d2b115: OUTPUT«Str␤»
11:32 andreoss m: multi MAIN() { die $*USAGE }; multi MAIN(Str:D :$option) { say $option };
11:32 camelia rakudo-moar d2b115: OUTPUT«Dynamic variable $*USAGE not found␤  in sub MAIN at <tmp> line 1␤  in block <unit> at <tmp> line 1␤␤»
11:33 andreoss how do i get this usage message?
11:36 iH2O left #perl6
11:39 woolfy joined #perl6
11:40 nwc10 joined #perl6
11:48 telex joined #perl6
11:49 _4d47 joined #perl6
11:52 fridim___ joined #perl6
11:52 _4d47 mm how can this method gets called ? https://github.com/rakudo/rakudo​/blob/nom/src/core/Array.pm#L419
11:53 mcmillhj joined #perl6
11:55 TheLemonMan _4d47, Array.new(VV, :shape(NN))
11:56 MasterDuke joined #perl6
11:59 _4d47 TheLemonMan: aww thanks, my eyes where reading Mu:D: instead of Mu:D
11:59 nwc10 left #perl6
11:59 MasterDuke andreoss: $?USAGE isn't implemented yet
11:59 MasterDuke and i believe MAIN will just call USAGE() if it can't dispatch. there's an autogenerated one, but if you create your own sub USAGE it will get called instead
12:01 lizmat joined #perl6
12:06 mcmillhj joined #perl6
12:10 elohmrow joined #perl6
12:14 woolfy joined #perl6
12:15 fridim___ left #perl6
12:22 sufrostico joined #perl6
12:23 _4d47 left #perl6
12:27 RabidGravy boom
12:28 MasterDuke big bada boom?
12:28 RabidGravy it's amazing how shutting down all those spreadsheet instances frees up memory
12:28 RabidGravy anyhow reboot
12:29 ab6tract joined #perl6
12:30 * ab6tract clickbaits http://blogs.perl.org/users/ab5tract/2​016/09/introducing-terminalprint.html
12:38 lizmat joined #perl6
12:44 TimToady bl 20
12:49 Xliff \o
12:49 Xliff Morning all.
12:49 Xliff https://gist.github.com/Xliff/f​ceb9354193e3b2d1d9c1968b991d6ec
12:49 Xliff ^ halp!
12:49 pierre_ joined #perl6
12:51 Xliff Oh. Lemme update that a sec. Because it's a weird edge case.
12:53 Xliff OK. Updated
12:55 nine Xliff: looks terrible (the error message)
12:56 nine Xliff: you could try package TheTest { ... }; sub EXPORT(+@a) { ... }
12:57 nine i.e. make sure TheTest is defined before you refer to it
12:58 ab6tract shouldn't EXPORT be inside the respective package?
12:58 mcmillhj joined #perl6
12:58 nine It belongs to the compilation unit and has to be in the top level.
12:58 nine AFAIK
12:59 Xliff nine: OK, one sec.
12:59 smls joined #perl6
13:00 ab6tract One day I'm going to have to learn me some of the distinctions between a compilation unit vs a file vs a 'unit module'
13:00 Xliff nine: No change.
13:00 ab6tract nine: so then if i do a 'unit class', the EXPORT will work because that's the top level of the compunit?
13:00 Xliff ab6tract: The original draft of the code had it that way, but that wouldn't parse.
13:01 El_Che ab6tract: looks empty
13:01 Xliff ab6tract: Or gave strange errors, depending.
13:01 ab6tract El_Che: the blog post?
13:01 El_Che ab6tract: yes
13:01 ab6tract El_Che: try a refresh please
13:02 ab6tract that was my fault
13:02 El_Che better :)
13:02 Xliff nine: Another weird thing... if I drop the signature from EXPORT, it still fails in the same manner.
13:02 Xliff For all 3 cases.
13:02 ab6tract thanks for pointing it out :)
13:04 Xliff Another odd thing: If I return an empty map from EXPORT, the error message goes away for all cases.
13:04 Xliff Should I RT?
13:04 ab6tract Xliff: I am using EXPORT::DEFAULT like this https://github.com/ab5tract/Terminal-Prin​t/blob/master/lib/Terminal/Print.pm6#L324
13:04 ab6tract and it seems to be working so far
13:05 Xliff ab6tract: Yes, but is that a map definition or a sub?
13:06 mcmillhj joined #perl6
13:06 Xliff On first glance, looked like a map. I will try.
13:06 ab6tract it is neither, AFAICT :)
13:06 ab6tract it seems to be autotriggered on import
13:07 ab6tract i'm just doing some binding into the OUR:: namespace there
13:08 Xliff SSDD
13:08 Xliff (still getting the error)
13:09 woolfy joined #perl6
13:12 RabidGravy joined #perl6
13:13 ab6tract Xliff: have you updated the gist?
13:15 cdg joined #perl6
13:15 Xliff About to. I think I found a work around.
13:16 maerciba5 joined #perl6
13:16 maerciba5 left #perl6
13:17 Xliff Updated.
13:20 Xliff Yup! Workaround applied to original code and is working! \o/
13:23 mr-foobar joined #perl6
13:25 skids joined #perl6
13:26 Actualeyes joined #perl6
13:28 ab6tract Xliff++
13:29 Xliff The more I write in P6, the more I find that the block notation is more reliable than using "unit ..."
13:30 mst hmm
13:30 mst this sort of sounds like maybe a case for finding all the tests using block notation and making sure there's a unit version
13:31 Xliff ¯\_(ツ)_/¯
13:31 Xliff m/...(>.<)...m/
13:32 Xliff Well, after this I get to move to nqp....
13:32 MilkmanDan joined #perl6
13:33 Xliff Still worried that's a bit over my head at this point, but hey...
13:33 TimToady sounds to me more like problems with how EXPORT works; 'unit' is conceptually very simple
13:37 TimToady unit should probably be disallowing those declarations before it, however, since the point of unit is to wrap the entire file as if it were a block declaration
13:39 jnthn Does unit sneakily introduce a lexical scope?
13:39 TimToady probably
13:40 jnthn That could be surprising in that a sub EXPORT is meant to be located in UNIT
13:40 TimToady it might be assuming wrongly that it's just about the first thing in the file
13:40 jnthn So folks doing `unit class Foo; sub EXPORT() { ... }` could get caught out
13:40 TimToady 'tis possible
13:41 TimToady would have to glare at the code some time after I wake up :)
13:41 RabidGravy heads up to anyone using chrome on FC24 (I think timotimo does,) if it offers an update of it, you are better just uninstalling and re-installing first
13:51 Xliff My original version of the module this popped up in had EXPORT located after the "unit" declaration.
13:52 Xliff I could never get that to work and the error messages were less than helpful when trying to figure out a solution.
13:52 Xliff Should I RT this issue?
13:52 Xliff Here's is a writeup, with solution... https://gist.github.com/Xliff/f​ceb9354193e3b2d1d9c1968b991d6ec
13:55 skids There is RT#127305
13:55 synopsebot6 Link:  https://rt.perl.org/rt3//Publ​ic/Bug/Display.html?id=127305
13:56 skids Lots of metaish modules already forgo the unit form.
13:56 TimToady there is precisely one test of 'sub EXPORT', and it doesn't involve 'unit'
13:56 TimToady it pulls in a file that reads:
13:56 TimToady class Foo {}
13:56 TimToady sub EXPORT(|){ { 'Bar' => Foo } }
13:57 TimToady which seems...a bit rudimentary
14:00 RabidGravy not exhaustive for sure
14:05 zacts joined #perl6
14:10 TimToady jnthn: yes, it calls <.newpad>, which it should probably avoid for a 'unit' declaration
14:13 * TimToady experiments with the simple fix first to see how much stuff blows up
14:13 mcmillhj joined #perl6
14:13 khw joined #perl6
14:15 TimToady .oO(phase one bug trials)
14:16 TimToady oh, heh, if I refrain from pushing a lexpad, I should also refrain from popping it...
14:17 timotimo RabidGravy: can you explain that a bit more? the chrome thing?
14:17 timotimo also, they no longer call it "Fedora Core" ;)
14:18 acrussell joined #perl6
14:18 TimToady Fedora Not-So-Core?
14:19 timotimo fedora shell?
14:19 timotimo fedora fruit?
14:19 RabidGravy timotimo, the upgrade seems to mess up the chain of links for the "alternatives" resulting in it disappearing completely
14:21 timotimo hah, that's fun
14:22 TimToady well, simple fix is not simple, insofar as it blows up tools/build/install-core-dist.pl
14:22 timotimo :|
14:22 TimToady maybe a 'make clean' issue
14:23 TimToady but if there's code that depends on a unit's OUTER meaning UNIT, we have a problem
14:27 TimToady we have a problem
14:28 TimToady getting a push on an NQPMu, probably something assuming OUTER gets to UNIT somewhere, I'm guessing
14:29 TimToady well, this will take at last another cuppa
14:29 Actualeyes joined #perl6
14:33 mspo ugh alternatives
14:35 timotimo with alternatives it seems like i'm always getting a 32bit wine binary but a 64bit wineserver binary served, or the other way around
14:35 timotimo every single time i want to run a windows program i have to shuffle the two around
14:35 timotimo it's hilariously bad
14:35 timotimo but perhaps it's because i get at least one wine update between two events of wanting to run any windows program
14:35 nwc10 joined #perl6
14:36 canopus joined #perl6
14:40 MetaZoffix joined #perl6
14:40 andreoss m: my $x = { $_++ }; $_ = 10 ;$x();$x();say $_;
14:40 camelia rakudo-moar 6dab6b: OUTPUT«12␤»
14:40 andreoss m: my $x = sub { $_++ }; $_ = 10 ;$x();$x();say $_;
14:40 camelia rakudo-moar 6dab6b: OUTPUT«10␤»
14:41 timotimo yeah, bare blocks don't get their own $_
14:41 timotimo m: my $x = -> { $_++ }; $_ = 10 ;$x();$x();say $_;
14:41 camelia rakudo-moar 6dab6b: OUTPUT«12␤»
14:41 timotimo like pointy blocks
14:44 TimToady well, fixed the NQPMu.push, but now install-core-dist merely runs forever
14:47 pmurias joined #perl6
14:47 nwc10 lizmat: domm is down to 5 minutes, so would be useful if your workshop could get ready to return
14:47 nwc10 (he's still speaking - not yet time to invade the room)
14:48 lizmat ok, I put out the message
14:49 pmurias ab5tract: using 'no precompilation' seems horrible
14:49 yoleaux 00:51Z <ugexe> pmurias: re: `zef clone Foo::Bar` question from last week - this mostly exists already using either `zef fetch Foo::Bar` or `zef look Foo::Bar`. If it happens that Foo::Bar resolves to a .tar.gz (like from cpan) and not a git uri it will extract it for you as well
14:51 * pmurias is also slightly upset people are using Terminal:: instead of the traditional Term::
14:52 TimToady I carped about that at the time, but nothing changed
14:52 timotimo pmurias: i think we "recently" got someone ranting about how Term collides with the term "term"
14:52 timotimo though i think if you had a term, you'd term it P6::Term or something, that'd be a better term for term
14:53 TimToady Term vs term is not even the same case, if people are worried about grammar collisions
14:54 TimToady "term" is such a generic term that I don't see any collision with people wanting to use that for a name root
14:54 timotimo what, modules are case sensitive? why can't i install ACME::MEOW :(
14:55 TimToady and Term has a long history of meaning "terminal" at the root of the namespace
14:56 TimToady next thing you know, we'll all be starting our programs with "IDENTIFICATION DIVISION" or so, just because longer is always clearer...
14:56 timotimo "THIS IS THE PLACE WHERE THE NAME OF THE PROGRAM GOES, BEN. PLEASE PUT THE NAME HERE THERE IS NO SUCH THING AS A PROGRAM NAME DIVISION GEEZ"
14:57 woolfy left #perl6
14:57 TimToady and PERFORM is obviously better than DO, and COMPUTE is obviously better than, well, nothing
14:58 timotimo every mathematical expression needs to start with COMPUTE
14:58 DrForr STORING RESULT INTO...
14:58 [Coke] WE HAVE THE BEST PROGRAM NAMES.
14:58 * timotimo afkbbl
14:58 TimToady anyway, I agree with pmurias++ here
14:58 lizmat joined #perl6
14:58 DrForr JOY IDENTIFICATION DIVISION
14:59 ugexe just use the sha1 as the module name
15:02 pmurias ugexe: unfortunately the modules end up as sha1 in stacktraces anyways
15:04 pmurias ab5tract: the only thing 'no precompilation' in Terminal::Print::Dimensions seems to be doing is wasting CPU time
15:05 nwc10 left #perl6
15:11 MetaZoffix m: class Meow { has $!s = Supplier.new; has $.sup = $!s.Supply; method do-things { $!s.emit: $_ for <a b c>; } }; react { with Meow.new { whenever .sup { $^m.say }; .do-things;  }; done }
15:11 camelia rakudo-moar e8d0d0: OUTPUT«a␤b␤c␤»
15:12 MetaZoffix Any idea why this isn't printing the stuff with whenevers in my code even tho it works in the example above? I got my react block https://github.com/zoffixznet/na/blob​/master/bin/release/cli-runner.p6#L6 and my Meow class that emits messages https://github.com/zoffixznet/na/b​lob/master/lib/NA/Releaser.pm6#L33
15:13 MetaZoffix And in the output I do get output from `say "Got stuff: $mes";` that's a line above the emits, but the whenevers don't react and don't print anything
15:13 Actualeyes1 joined #perl6
15:13 * [Coke] tries to figure out his bailador/start problem now that the latest crisis is over
15:14 canopus joined #perl6
15:17 MetaZoffix And even if I put $!shell-messenger.emit: "built stuff!!!!!"; above this `given` I get the message received :S weird https://github.com/zoffixznet/na/b​lob/master/lib/NA/Releaser.pm6#L19
15:18 smls m: say (1..*).rotor(1..*)[^5];  # Floyd's triangle
15:18 camelia rakudo-moar e8d0d0: OUTPUT«((1) (2 3) (4 5 6) (7 8 9 10) (11 12 13 14 15))␤»
15:18 smls m: say (1..*).rotor(1...*)[^5];  # Floyd's triangle
15:18 domidumont joined #perl6
15:19 smls ^^ Why does it work with a Range argument but not a Seq?
15:19 camelia rakudo-moar e8d0d0: OUTPUT«(timeout)»
15:19 smls Does it have to do with how parameters are passed to routines, or does `rotor` mishandle it?
15:20 MetaZoffix s: (1..*), "rotor", \(1...*)
15:20 SourceBaby MetaZoffix, Something's wrong: ␤ERR: Type check failed in binding to &code; expected Callable but got Method+{<anon|70743856>} (Method+{<anon|7074385...)␤  in sub do-sourcery at /home/zoffix/services/lib/CoreHackers-​Sourcery/lib/CoreHackers/Sourcery.pm6 (CoreHackers::Sourcery) line 42␤  in sub sourcery at /home/zoffix/services/lib/CoreHackers-​Sourcery/lib/CoreHackers/Sourcery.pm6 (CoreHackers::Sourcery) line 29␤  in block <unit> at -e l
15:20 MetaZoffix s: (1..*), "rotor", \(1..*)
15:20 SourceBaby MetaZoffix, Something's wrong: ␤ERR: Unhandled exception: While looking for '/home/zoffix/services/sourceabl​e/building-perl6/perl6.moarvm': no such file or directory␤
15:20 MetaZoffix great
15:20 MetaZoffix s: (1..*), "rotor", \(1..*)
15:20 SourceBaby MetaZoffix, Something's wrong: ␤ERR: Unhandled exception: Missing or wrong version of dependency 'src/Perl6/Grammar.nqp' (from 'gen/moar/m-main.nqp')␤   at <unknown>:1  (/home/zoffix/services/sourceable/building-pe​rl6/perl6.moarvm:<dependencies+deserialize>)␤
15:21 domidumont joined #perl6
15:21 Xliff smls: WAG - Eager vs lazy?
15:21 MetaZoffix m: my $w = 1...*; say $w.list[^5]
15:21 MetaZoffix there's yo problem
15:21 camelia rakudo-moar e8d0d0: OUTPUT«(timeout)»
15:21 Xliff m: my $w = 1..%; say $w.list[^5]
15:21 camelia rakudo-moar e8d0d0: OUTPUT«(Nil Nil Nil Nil Nil)␤»
15:22 [Coke] m: my $w := 1...*; say $w.list[^5]
15:22 camelia rakudo-moar e8d0d0: OUTPUT«(1 2 3 4 5)␤»
15:22 Xliff m: my $w = 1..*; say $w.list[^5]
15:22 camelia rakudo-moar e8d0d0: OUTPUT«(1 2 3 4 5)␤»
15:22 MetaZoffix smls: rotor takes a List:D, so for Seqs it goes through Any, which calls .list which loses laziness
15:22 Xliff \o/
15:22 Xliff My WAG was right!?
15:22 MetaZoffix smls: it may or may not be an LTA :) feel free to rakudobug
15:23 smls m: say (1..*).^find_method("rotor").signature
15:23 camelia rakudo-moar e8d0d0: OUTPUT«($: |c is raw)␤»
15:23 timotimo "WAG"?
15:23 Xliff Wild Assed Guess
15:23 timotimo oh
15:23 smls MetaZoffix: Where does it say it takes a List:D?
15:23 Xliff =)
15:24 MetaZoffix s: (1..*), "rotor", \(1..*)
15:24 SourceBaby MetaZoffix, Something's wrong: ␤ERR: Type check failed in binding to &code; expected Callable but got Method+{<anon|73357248>} (Method+{<anon|7335724...)␤  in sub do-sourcery at /home/zoffix/services/lib/CoreHackers-​Sourcery/lib/CoreHackers/Sourcery.pm6 (CoreHackers::Sourcery) line 42␤  in sub sourcery at /home/zoffix/services/lib/CoreHackers-​Sourcery/lib/CoreHackers/Sourcery.pm6 (CoreHackers::Sourcery) line 29␤  in block <unit> at -e l
15:24 MetaZoffix I hate that bug
15:24 MetaZoffix smls: in the source code: https://github.com/rakudo/rakudo​/blob/nom/src/core/List.pm#L1152
15:25 smls m: my $w = (1...*); say $w.list[^5]
15:25 MetaZoffix smls: oh, wait, I'm even looking at the wrong thing. It's the divisors that are the problem >_<
15:25 camelia rakudo-moar e8d0d0: OUTPUT«(1 2 3 4 5)␤»
15:25 smls ^^ MetaZoffix: Your timeout earlier was because missing parens
15:25 MetaZoffix right
15:26 lostinfog joined #perl6
15:26 MetaZoffix m: sub foo ( *@what ) { say @what[^5] }( 1...* )
15:26 camelia rakudo-moar e8d0d0: OUTPUT«(1 2 3 4 5)␤»
15:26 MetaZoffix m: sub foo ( *@what ) { say @what.is-lazy }( 1...* )
15:27 MetaZoffix :/
15:27 camelia rakudo-moar e8d0d0: OUTPUT«(timeout)»
15:27 MetaZoffix m: sub foo ( *@what ) { say @what.is-lazy }( 1...* )
15:27 timotimo potentially only works with +@?
15:27 camelia rakudo-moar e8d0d0: OUTPUT«(timeout)»
15:27 smls That one seems related to RT #128201
15:27 synopsebot6 Link:  https://rt.perl.org/rt3//Publ​ic/Bug/Display.html?id=128201
15:28 MetaZoffix m: sub foo ( +@what ) { say @what.is-lazy }( 1...* )
15:28 camelia rakudo-moar e8d0d0: OUTPUT«True␤»
15:28 MetaZoffix bisect: m: sub foo ( *@what ) { say @what.is-lazy }( 1...* )
15:28 MetaZoffix oh shit.. that was probably a bad idea
15:28 MetaZoffix bisect: m: Promise.in(2).then: {exit}; sub foo ( *@what ) { say @what.is-lazy }( 1...* )
15:29 MetaZoffix smls: seems to be yeah. 'cause .rotor is checking .is-lazy: https://github.com/rakudo/rakudo​/blob/nom/src/core/List.pm#L1154
15:29 * MetaZoffix &
15:33 grondilu joined #perl6
15:33 MetaZoffix joined #perl6
15:33 MetaZoffix I got distracted and totally forgot that I originally came here with my emit issue :) Any idea why it's not emitting?
15:36 timotimo what isn't emitting?
15:37 MetaZoffix timotimo: my code, about two screenfuls above. http://irclog.perlgeek.de/p​erl6/2016-09-02#i_13139346
15:37 timotimo thank you
15:37 MetaZoffix Even though small sample works... even though doing $!shell-messenger.emit: $mes; in another method works. Even though putting "say $mess" in the original method with .emit also prints stuff :S
15:39 bioduds joined #perl6
15:41 * timotimo very distracted
15:47 MetaZoffix figured it out by writing it like this: https://gist.github.com/zoffixznet​/ee5e2a1fdd4634711f03df60f79b5831
15:48 awwaiid hoelzro: Add yourself to the STL unsession I made for Rakudo Perl 6, https://github.com/strangeloop/​StrangeLoop2016/wiki/Unsessions
15:48 MetaZoffix Seems it's getting its head screwed when the method calls are inside the react{}
15:48 awwaiid and anyone else going to StrangeLoop as well :)
15:48 MetaZoffix Maybe I should try a Channel instead of Supplies in here
15:54 bjz joined #perl6
15:55 sakuya joined #perl6
15:56 bisectable6 joined #perl6
15:56 MetaZoffix Thanks for the help!
16:04 zacts joined #perl6
16:16 pierre_ joined #perl6
16:16 ab6tract pmurias: then someone fixed a bug in precompilation, because previous editions of this library would get a single value for terminal width and height
16:16 ab6tract pmurias: WRT to 'term' vs 'terminal' -- sorry, I cannot disagree more
16:17 ab6tract single value for each of width and height, which will be wrong if the terminal is resized after precompilation
16:17 ab6tract if you find that behavior to be in error, please submit an issue
16:17 ab6tract I think it would be a crying shame if P6 ended up with P5's namespacing by fiat
16:24 domidumont joined #perl6
16:25 ab6tract after some initial testing, I am seeing that removing no precompilation does indeed have the undesired effect of pinning the grid size of the first invocation
16:25 ab6tract pmurias: if you find it behaving otherwise, I would be interested to know about it
16:27 [Coke] .seen supernovus
16:27 yoleaux I saw supernovus 22 Apr 2016 23:27Z in #perl6: <supernovus> Well, I'm going to have to run. Have a great day/night everyone. I hope to fix up some of my long neglected libraries at some point when I'm not completely overloaded with work! :-)
16:28 ab6tract further, it seems to settle the size at precompilation time, which means it isn't even accurate to the size you had when you first ran it
16:28 [Coke] .seen ufobat
16:28 yoleaux I saw ufobat 27 Aug 2016 06:53Z in #perl6: <ufobat> yay :D
16:31 mcmillhj joined #perl6
16:37 zacts_pi joined #perl6
16:41 hoelzro awwaiid: done!
16:46 tailgate Is Pod::To::HTML a library that comes with rakudo-star? I tried deleting all copies of it on my computer, but it's still loading it
16:46 tailgate how do I find where it's getting the file from?
16:50 ab6tract .tell pmurias I've tried to make the the 'no precompilation' hit as minimal as possible. are you using a terminal client which automatically updates COLUMNS and ROWS in the environment for you as you resize?
16:50 yoleaux ab6tract: I'll pass your message to pmurias.
16:50 timotimo rakudo has Pod::To::Text, nothing else in Pod::
16:51 mcmillhj joined #perl6
17:03 mcmillhj joined #perl6
17:03 bioduds hey guys
17:04 bioduds im not able to rua a one liner
17:04 bioduds only a 2 liner
17:04 bioduds this is the best I can do so far
17:04 bioduds http://67.205.136.118/install.sh
17:05 sufrostico joined #perl6
17:05 timotimo how does it not work?
17:05 timotimo i mean, what's wrong?
17:08 timotimo hm. you set PREFIX, is that enough? i thought we have to --prefix in the configure line, too
17:08 timotimo and if you put stuff into /usr/local, it's not going to be able to install anything into the prefix
17:09 timotimo because you're asking the user to run it as "sh" not as "sudo sh"
17:09 timotimo and also:
17:09 timotimo compiling and such shouldn't be done as root
17:09 timotimo so maybe it'd be better to extract tha building into three steps so we can "sudo make install" only
17:09 mcmillhj joined #perl6
17:10 bioduds --PREFIX ?
17:10 bioduds how so?
17:10 timotimo no, --prefix
17:10 firstdayonthejob joined #perl6
17:10 timotimo i've never tried to use the prefix env var
17:11 timotimo also, i think if you don't "export" it, it won't land in the environment anyway
17:11 bioduds i believe that lies the magic of making all work, but I really dont quite understand what it does
17:11 MilkmanDan joined #perl6
17:12 bioduds what im testing now is to run the following line: curl http://67.205.136.118/install.sh | sh; . ~/rakudo/setpath.sh
17:12 bioduds as it would actually be 2 lines in one
17:12 bioduds that would be the 'ONE' liner
17:12 timotimo ok. at what point does it break?
17:12 bioduds it does not break
17:12 bioduds it installs fine
17:13 bioduds you just need to . ~/rakudo/setpath.sh manually after
17:13 timotimo oh
17:13 bioduds which Im hoping to have already on with ;. ~/rakudo/setpath.sh
17:14 bioduds testing now, if works, should be it
17:17 pierre_ joined #perl6
17:20 bioduds WORKED
17:20 bioduds ONE LINER IS HERE
17:20 bioduds its actually a 2-liner
17:20 bioduds so
17:21 bioduds right now you may do: curl http://67.205.136.118/install.sh | sh ;. ~/rakudo/setpath.sh
17:21 bioduds and it will install perfectly
17:24 mspo that looks totally safe
17:24 mspo sudo it?
17:24 bioduds no need
17:24 bioduds simply run it
17:25 mspo was joke :)
17:25 bioduds oh
17:25 bioduds :)
17:25 bioduds now, there need to be SSL on to have it on https://install.perl6.org
17:25 bioduds making the 'one liner': curl https://install.perl6.org | sh;. ~/rakudo/setpath.sh
17:26 bioduds which will work like a charm
17:26 bioduds anyone would care to test it on a mac?
17:26 sjohnson let's encrypt!
17:26 bioduds i believe it will work on mac too
17:27 mspo The certificate is only valid for the following names: design.perl6.org, doc.perl6.org, docs.perl6.org, modules.perl6.org, perl6.org, tablets.perl6.org, www.p6c.org, www.perl6.org
17:27 timotimo we've already got SSL for the rest of perl6.org, no?
17:27 bioduds im gonna start testing on a few other linux distros as well
17:28 mspo new firefox doesn't have a way to actually view the cert
17:28 mspo that's really annoying
17:28 sjohnson it does, you just gotta dig deeper.
17:28 mspo I just get the www page from install
17:28 timotimo protect the innocent users who don't know what ssl is
17:29 sjohnson design.perl6.org, doc.perl6.org, docs.perl6.org, modules.perl6.org, perl6.org, tablets.perl6.org, www.p6c.org, www.perl6.org
17:29 sjohnson looks like my list matches mspo.
17:30 bioduds im going to put up a test list and once its tested on all ill send you guys the table
17:37 japhb ab5tract: Thank you for writing Terminal::Print.  I've been putting off doing a couple projects (both work and personal) because I didn't feel like a giant yak shave.  You've just removed several layers of yak hair.  :-)
17:38 timotimo do we have a certificate behind where we download the rakudo star tarball from?
17:40 acrussell joined #perl6
17:47 bioduds no
17:47 bioduds there should be
17:47 bioduds https://rakudo.org/downloads/s​tar/rakudo-star-2016.04.tar.gz instead of http://rakudo.org/downloads/st​ar/rakudo-star-2016.04.tar.gz
17:51 mspo The certificate is only valid for the following names: host.pmichaud.com, www.host.pmichaud.com
17:58 leont joined #perl6
18:03 timotimo damn
18:03 kyclark joined #perl6
18:03 zakharyas joined #perl6
18:04 bioduds this whole ssl mob is gruesome
18:05 mspo where is perl6.org hosted?
18:05 labster joined #perl6
18:05 bioduds tests on ubuntu 14.04 LTS 64 bits with 2 cpus and 2gb ram ran ok
18:05 mspo p6c
18:05 mspo why isn't rakudo just uploading to there?
18:05 bioduds testing now with 14.04 1 cpu 1 gb ram
18:05 bioduds digital ocean vps's
18:05 bioduds gonna try them all
18:06 bioduds writing a sheet with results too
18:07 bioduds wow
18:07 bioduds 1gb failed
18:07 bioduds this install script is greedy man
18:08 DrForr Perl6 ain't Lua.
18:09 mspo yeah
18:09 bioduds testing CentOS 7.2 x64 2gb ram 2 cpus dig. ocean
18:10 timotimo yeah, the core setting is kind of big
18:12 kyclark Given a list of strings, how would I efficiently find all those starting with a given list of prefixes?
18:12 kyclark my @list = <foo bar baz quux>
18:12 kyclark my @want = <fo ba>
18:12 kyclark I feel something like " @list ~~ any(@want)"
18:13 kyclark Actually, I just need to do one string at a time, e.g., "foo ~~ any(@want)"
18:13 TestNinja joined #perl6
18:13 kyclark @want.map({ 'foo' ~~ /^$_/ })
18:15 TestNinja m: my @list = <foo bar baz quux>; my @want = <fo ba>; say @want.grep: *.starts-with: @list.any
18:15 camelia rakudo-moar e8d0d0: OUTPUT«()␤»
18:15 TestNinja orly
18:15 TestNinja m: my @list = <foo bar baz quux>; my @want = <fo ba>; say @want.grep: /^@list/
18:15 camelia rakudo-moar e8d0d0: OUTPUT«()␤»
18:15 TestNinja oh right :)
18:16 TestNinja m: my @list = <foo bar baz quux>; my @want = <fo ba>; say @list.grep: /^@want/
18:16 camelia rakudo-moar e8d0d0: OUTPUT«(foo bar baz)␤»
18:16 TestNinja m: my @list = <foo bar baz quux>; my @want = <fo ba>; say @list.grep: *.starts-with: @want.any
18:16 camelia rakudo-moar e8d0d0: OUTPUT«(foo bar baz)␤»
18:16 TestNinja \o/
18:17 kyclark Let me soak that in for a bit.
18:26 dj_goku Is anyone else having issues building 2016.08.1?
18:26 TestNinja dj_goku: what sort of issues?
18:27 kyclark TestNinja, that's some cool stuff.  Thanks.
18:27 sjoshi joined #perl6
18:27 TestNinja Any time.
18:27 dj_goku TestNinja: I can't scroll back give me a few minutes to try and rebuild it.
18:27 dj_goku currently trying 2016.07.1
18:29 dj_goku hmm might be a memory issue: make: *** [CORE.setting.moarvm] Error 137
18:29 dj_goku Command failed (status 512): make
18:29 TestNinja You need around 2GB for RAM+swap
18:30 dj_goku hmm
18:30 sjoshi joined #perl6
18:31 bioduds yep, around 2gb ram
18:31 bioduds from what my tests are showing
18:32 dj_goku well that stinks lol
18:32 TestNinja Just resize to get more swap :)
18:33 TestNinja I built on boxes with 1GB RAM and large spawp
18:33 dj_goku this is a production box. :D
18:34 TestNinja ooopppss....
18:34 TestNinja murdered committable6 with committable6: 2016.04 m: say "meow"; <a b c d b>.rotor: 1 => -2;
18:35 Xliff You bastid!
18:35 moritz TestNinja: it's only murder if the dead one is of the same species as you. So you're a bot? :-)
18:35 TestNinja I guess it's not comming back :/
18:35 Xliff Sorry. South Park ref. I couldn't help it.
18:35 dj_goku After a few things finish running I'll cycle a service that is eating some memory
18:35 TestNinja moritz: you never know ;)
18:38 Xliff "<TestNinja> m: my @list = <foo bar baz quux>; my @want = <fo ba>; say @list.grep: *.starts-with: @want.any"
18:38 Xliff Syntax Yum!
18:38 domidumont joined #perl6
18:38 bioduds not sure I can resize swap on Digital Ocean VPSs
18:39 mspo do you have shell access?
18:40 bioduds also, main reason for this one-liner p6 installer is to help out for beginners. so, we dont want to tell beginners they should actually go and change swap
18:40 bioduds yes, shell access
18:40 mst right, 1Gb isn't enough to build rakudo
18:40 bioduds out of the box, 2gb is needed
18:40 mspo bioduds: just make a swapfile
18:41 bioduds in docs, there could be a full explanation on how to change swap in order to make it install on 1gb ram
18:41 mspo dd if=/dev/zero of=/var/db/swap0 bs=1M count=1024 ; mkswap /var/db/swap0; swapon /var/db/swap0
18:42 bioduds I'll use this mspo after I test all linux distros i can, thanks :D
18:42 bioduds currently on centos 7.2 2b ram 64 bits
18:43 bioduds it is installing
18:43 bioduds all need to have build-essential set before
18:43 bioduds sudo apt install build-essential on ubutnu
18:43 bioduds sudo yum groupinstall 'Development Tools' on centos
18:43 mst I think ~development also works
18:44 bioduds im not sure how to check this in script
18:44 bioduds if I manage to have it tested
18:44 bioduds I can set a line to install it if it is not there I guess
18:46 avuserow bioduds: better is to check if the tool exists, maybe something like `command -v make >/dev/null || (echo no make && exit 1)`
18:46 bioduds ok, ill try that
18:46 bioduds cause if i can do this, the one-liner will work out of the box
18:47 bioduds since perl5 ships with main linux distros
18:47 bioduds i believe it ships with mac as well
18:47 andreoss left #perl6
18:51 bioduds min 1gb swap would be what? 1024?
18:51 dha joined #perl6
18:51 bioduds there could be a min reqs telling 2gb ram or 1gb min. 1024Mb swap
18:52 bioduds also, 512 RAM and swap 2048 would work?
18:54 [Coke] rakudo isn't on perl6.org because the Rakudo Perl 6 Compiler is a different thing from the Perl 6 Language Specification.
18:54 [Coke] and this underscores that difference.
18:55 leont joined #perl6
18:56 bioduds centos installed great
18:56 bioduds testing 1GB RAM now
18:57 committable6 joined #perl6
18:58 bioduds avuserow
18:58 bioduds how should i do this test in the shell script?
18:59 zacts joined #perl6
18:59 [Coke] I doubt there is a portable way for you to do that.
18:59 [Coke] or at least, an easy way.
19:00 bioduds i wont work on that i guess
19:00 bioduds maybe after testing
19:03 TestNinja m: %{Any} = a => 'B', c => 'b';
19:03 camelia rakudo-moar e8d0d0: OUTPUT«Odd number of elements found where hash initializer expected:␤Only saw: -> ;; $_? is raw { #`(Block|74651200) ... }␤  in block <unit> at <tmp> line 1␤␤»
19:03 TestNinja What's it on about?
19:03 bioduds also, installing it after checking that it does not exist would be platform dependent
19:03 TestNinja m: my %h{Any}; %h = a => 'B', c => 'b'; dd %h
19:03 camelia rakudo-moar e8d0d0: OUTPUT«Hash[Any,Any] %h = (my Any %{Any} = :a("B"), :c("b"))␤»
19:04 TestNinja m: my Any %{Any} = :a("B"), :c("b")
19:04 camelia rakudo-moar e8d0d0: ( no output )
19:04 TestNinja mkay
19:04 TestNinja m: %{Any} = a => "B", c => "b"
19:04 camelia rakudo-moar e8d0d0: OUTPUT«Odd number of elements found where hash initializer expected:␤Only saw: -> ;; $_? is raw { #`(Block|77295888) ... }␤  in block <unit> at <tmp> line 1␤␤»
19:04 TestNinja m: my %{Any} = a => "B", c => "b"
19:04 camelia rakudo-moar e8d0d0: ( no output )
19:04 bioduds testing on 1gb ram centos
19:04 TestNinja m: %{Any}
19:04 camelia rakudo-moar e8d0d0: OUTPUT«Odd number of elements found where hash initializer expected:␤Only saw: -> ;; $_? is raw { #`(Block|76513384) ... }␤  in block <unit> at <tmp> line 1␤␤»
19:05 TestNinja Ah, treats it as a coercer
19:07 iH2O joined #perl6
19:07 bioduds i should nevertheless ensure it is runnable
19:07 bioduds as avuserow pointed
19:08 bioduds also, check if min perl5 is there
19:08 bioduds do you guys know what is min. perl5 version needed for perl6 to be installed?
19:09 MasterDuke Configure.pl has 'use 5.010;'
19:12 sjoshi joined #perl6
19:12 timotimo bioduds: did you keep the PREFIX= part of the script?
19:13 TestNinja bioduds: 5.10.1
19:13 TestNinja (5.10.0 won't work)
19:15 bioduds fedora does not come with perl pre-installed
19:15 bioduds timotimo yes I did
19:15 bioduds right now im testing if i can detect make
19:15 bioduds also will test for perl
19:16 bioduds so 5.10.1 and above min
19:16 bioduds ok
19:18 timotimo cool
19:22 leont 5.8.8 may be a more convenient dependency (RHEL5), but OTOH that target group is unlikely to be interested in p6
19:23 mst I would be very surprised if tehre was somebody with 5.8.8 who both wanted p6 and wasn't capable of building a newer perl5
19:24 bioduds there are these legacy users but as you said, it would surprise to see if they wanted to "upgrade" to p6
19:26 bioduds good, i could test now
19:26 bioduds with
19:26 bioduds command -v make >/dev/null 2>&1 || { echo >&2 "I require make but it's not installed.  Aborting."; exit 1; }
19:27 leont True, I doubt it's much of a practical concern
19:27 bioduds i guess i can do the same for perl
19:29 bioduds good, worked
19:29 bioduds now I can prevent installation if no make and no perl is available and warn installers to install it
19:30 bioduds http://67.205.136.118/install.sh
19:30 [Coke] If i have a promise with a status of Broken, how can I find out what broke it?
19:30 [Coke] arglebargle. "cause"
19:32 TestNinja .result
19:33 TestNinja Well, the .result would have an Exception if it got broken by exception. Unsure about promises broken manually
19:33 timotimo m: say Promise.^attributes
19:33 camelia rakudo-moar 685cc5: OUTPUT«(Mu $!scheduler Mu $!status Mu $!result int $!vow_taken Mu $!lock Mu $!cond Positional @!thens Mu $!dynamic_context)␤»
19:34 timotimo m: say Promise.^methods(:local)
19:34 camelia rakudo-moar 685cc5: OUTPUT«(BUILD vow keep break result cause then start in at anyof allof Supply Str Numeric Bool scheduler status)␤»
19:34 timotimo ah, yes, cause is probably it
19:34 bioduds cool, installing on Fedora 24 x64 2gb ram
19:34 TestNinja :o
19:35 TestNinja "If the promise was broken, returns the result (or exception). Otherwise, throws an exception of type X::Promise::CauseOnlyValidOnBroken."
19:35 bioduds timotimo did you say I should remove PREFIX="/usr/local" ?
19:35 bioduds or leave it there?
19:35 timotimo remove
19:35 timotimo because i don't think it actually has any effect the way it's there
19:36 bioduds ok
19:36 timotimo because you said "curl the_script | sh" works, but it can't work if the prefix is really /usr/local
19:36 bioduds removed, going to test in a sec
19:36 TestNinja m: say (1.1, 3.1).all ~~ any(1.1, 2.1, 3.1)
19:36 camelia rakudo-moar 685cc5: OUTPUT«False␤»
19:37 TestNinja I excpected that to be True. How come it's false? Is it comparing two Junctions to each other instead of junctioning the values?
19:37 bioduds testing
19:38 [Coke] I have a promise $p that is broken. if i then try to print $p.cause.backtrace, I get Backtrace.new
19:38 espadrine joined #perl6
19:38 bioduds fedora is failing
19:38 bioduds Can't exec "/home/super/rakudo/rakudo-st​ar-2016.04/install/bin/moar": No such file or directory at tools/lib/NQP/Configure.pm line 450.  ===SORRY!=== No suitable MoarVM (moar executable) found using the --prefix (You can get a MoarVM built automatically with --gen-moar.)  Command failed (status 512): /usr/bin/perl Configure.pl --prefix=/home/super/rakudo/​rakudo-star-2016.04/install --backends=moar --make-install make: *** No targets
19:39 bioduds what is this?
19:39 TestNinja bioduds: it can't find moar
19:39 TestNinja m: say .elems == .grep: any(1.1, 2.1, 3.1) given (1.1, 3.1)
19:39 camelia rakudo-moar 685cc5: OUTPUT«True␤»
19:39 TestNinja This'll do, for me, I guess
19:39 timotimo right, the configure.pl is supposed to alreay install a moar when it goes on to build nqp
19:39 [Coke] looks like you didn't pass in --gen-moar
19:40 bioduds i did
19:40 bioduds perl Configure.pl --backend=moar --gen-moar
19:40 bioduds is there in the script
19:42 moritz then it looks like compiling or installing moar failed
19:42 bioduds perhaps there is no tar
19:42 MasterDuke --backends
19:42 bioduds fedora comes all naked it seems
19:42 moritz --gen-moar implies backend moar
19:43 bioduds no, there is tar
19:43 bioduds well, I dont understand then
19:43 bioduds No suitable MoarVM (moar executable) found using the --prefix
19:43 [Coke] did you look at the output of the failed build? or are you hiding it?
19:43 bioduds sorry guys, i dont understand what this means
19:44 [Coke] if you have it captured, nopaste it somewhere.
19:44 [Coke] probably that moar didn't build. you have to find the error message on the moar build.
19:44 bioduds let me try
19:44 [Coke] nopaste the output of your failed attempt to build (put it on gist.github.com or something)
19:45 [Coke] aaaarigh. give my bailador script a POST. fails in a promise, can't get the backtrace of the promise. again, fails, again, now fails different way.... again, now it works. wtf.
19:46 [Coke] (fresh copy? that still fails)
19:46 bioduds gonna leave fedora for later
19:47 bioduds heres what i got so far https://docs.google.com/spreadsheets/d/1dOAJEglf0D​9hV5doDADZJEMUIvWlmsRSBx8S118-3iY/edit?usp=sharing
19:48 avuserow bioduds: the snippet I gave you was shell code and seemed to work for me. Not a shell expert though.
19:48 bioduds yes avuserow it was almost correct
19:48 bioduds i just changed a bit
19:48 bioduds is working with: https://docs.google.com/spreadsheets/d/1dOAJEglf0D​9hV5doDADZJEMUIvWlmsRSBx8S118-3iY/edit?usp=sharing
19:48 bioduds opd
19:49 bioduds ops sorry
19:49 bioduds is working with: command -v make >/dev/null 2>&1 || { echo >&2 "make command is required to install Perl6. Please install it and try again."; exit 1; }
19:49 bioduds also I did the same to test for perl
19:49 bioduds command -v perl >/dev/null 2>&1 || { echo >&2 "perl is required to install Perl6 (Min. version 5.10.1). Please install it and try again."; exit 1; }
19:49 bioduds this works
19:49 bioduds and quits when none is available
19:50 [Coke] AIIIGH
19:50 iH2O left #perl6
19:54 [Coke] (that was for my async bailador stuff, nothing here. )
20:00 baest_ joined #perl6
20:08 [Coke] https://docs.perl6.org/routine/run should note if run is synchronous or not.
20:10 cdg joined #perl6
20:11 leont Yeah, that could be made more obvious, even if synchronous is the default in the rest of the world
20:14 [Coke] m: my $p = start { die 'oh bother' }; sleep 5; say ~$p.cause;
20:14 [Coke] m: my $p = start { die 'oh bother' }; sleep 5; say $p.cause;
20:14 camelia rakudo-moar 685cc5: OUTPUT«oh bother␤»
20:14 camelia rakudo-moar 685cc5: OUTPUT«oh bother␤  in block  at <tmp> line 1␤␤»
20:15 [Coke] there's a surprising Str/gist difference. :|
20:16 TEttinger joined #perl6
20:17 geekosaur makes sense to me, although something like $p.cause.payload or etc. might make more sense/be clearer
20:19 [Coke] should at least be documented on https://docs.perl6.org/type/Exception
20:19 cpage_ joined #perl6
20:22 El_Che travis rules: you can test perl6 docker images with it!
20:23 timotimo nice
20:24 El_Che https://travis-ci.org/nxadm/rakudo-pkg
20:28 jjido joined #perl6
20:35 AlexDani` joined #perl6
20:37 shzbot6959 joined #perl6
20:40 dha joined #perl6
20:41 MilkmanDan joined #perl6
20:43 bisectable6 joined #perl6
20:51 cdg joined #perl6
21:06 sufrostico joined #perl6
21:08 El_Che bioduds: I am trying to build native packages. I may have a go at creating fedora packages if I find the time. But you could have a go at the Centos7 pkg: https://github.com/nxadm/rakudo-pkg (the compile commands are in the docker dir, pkg_rakudo.sh for compiling, the Dockerfile for the prerequisites)
21:09 El_Che s/go/look
21:15 grondilu m: say (my %)<foo> // "oops";
21:15 camelia rakudo-moar 685cc5: OUTPUT«oops␤»
21:16 grondilu oh my bad
21:16 MilkmanDan joined #perl6
21:17 grondilu m: my %h; (%h<foo> && die "already defined") = "bar"; say %h<foo>
21:17 camelia rakudo-moar 685cc5: OUTPUT«bar␤»
21:17 grondilu nice
21:21 mcmillhj joined #perl6
21:23 * grondilu wonders if that's possible in P5
21:23 grondilu nope
21:24 grondilu m: (my $ and die) = "foo"
21:24 camelia rakudo-moar 685cc5: ( no output )
21:33 labster joined #perl6
21:39 harmil_wk I'm really getting into the weeds with this timeout issue. It shouldn't be taking me a week to figure out how to time out a test...
21:41 ugexe m: await Promise.anyof(start { sleep 100 }, start { sleep 3 });
21:42 camelia rakudo-moar 685cc5: ( no output )
21:42 harmil_wk ugexe: that can't interrupt a long-running call, though, right?
21:42 harmil_wk The sleep 100 just ends up running until you exit, doesn't it?
21:42 lizmat joined #perl6
21:43 harmil_wk I have 177 tests to run, and I don't want that many threads generating massive integers until I exit.
21:44 harmil_wk Java has a Thread.interrupt which would be ideal. Then I'd just Promise.anyof($timer, $test) and then $test.interrrupt if the timer hit.
21:47 harmil_wk m: await Promise.anyof(start { loop { sleep 1; say 1} }, start { sleep 1 }); say "Okay, your test is done"; sleep 2;
21:47 camelia rakudo-moar 685cc5: OUTPUT«1␤Okay, your test is done␤1␤1␤»
21:52 wamba joined #perl6
21:52 bioduds guys, what should be the place to ask for help if a user can't get perl6 to be installed?
21:53 bioduds I need to put that on the script if it for some unknown reason installation fails
21:54 cpage_ joined #perl6
21:55 harmil_wk If I'm reading what you wrote correct, bioduds, you're asking what you should tell people to do if they can't install Perl 6 as a prerequisite for running your script?
21:55 ugexe m: my $s = Supplier.new; my $sup = $s.Supply; await Promise.anyof(start { $sup.tap: { die(); }; for (^100) { say $_; sleep 1 } }, start { sleep 1 }); $s.emit(0); sleep 2; # some sort of kill switch i guess... this blows up the whole thing though
21:55 camelia rakudo-moar 685cc5: OUTPUT«0␤1␤Died␤  in block <unit> at <tmp> line 1␤␤»
21:56 bioduds yes, if they cant install with the script, where should they go for help
21:56 harmil_wk ugexe: oh, interesting... that's much better than the horrific signal handling I was trying to get working...
21:57 woolfy joined #perl6
21:57 bioduds should I tell them to go to chat? or faq?
21:57 bioduds perhaps https://docs.perl6.org/language/faq.html
21:57 harmil_wk bioduds: does your script try to install Perl6? I'm just trying to avoid the situation where your script has a problem and we direct people to a general Perl 6 resource...
21:57 bioduds yes it tries to install perl6
21:58 bioduds script is here: http://67.205.136.118/install.sh
21:58 bioduds it uses main perl6 installation proceedures
21:59 bioduds if it gets approved, should go run in https://install.perl6.org
21:59 harmil_wk I would say that the first line of defense should be the script maintainer.
21:59 frobisher joined #perl6
21:59 bioduds as top p6 guys here are looking on how would the best way to do this should be
22:00 bioduds I agree.
22:00 bioduds So, you think it'd be best if I say, please contact script developer - my email and also look at faq
22:00 stmuk bioduds: why doesn't it use the latest Rakudo Star?
22:00 bioduds this should be ok?
22:00 bioduds it is using latest
22:00 bioduds no?
22:01 stmuk rakudo-star-2016.07.tar.gz was the last
22:01 harmil_wk So something like, "if the script fails to install, see <location for script FAQ/bug reporting>, but if you have issues with the installed version of Perl 6 after installation, see <various Perl 6 references and FAQs>"
22:01 stmuk it should work if you just change it
22:01 bioduds let me change it to the latest then
22:01 bioduds thanks
22:02 dha We didn't have a release last month?
22:03 harmil_wk bioduds: note that you can get the latest release tag here: https://github.com/rakudo/star/releases
22:04 harmil_wk dha: star releases more slowly now that we're in non-zero-index-Greek-letters
22:05 dha Ah.
22:05 stmuk dha: nope I'm aiming for 3 monthly releases
22:05 harmil_wk Though I guess it's not Greek, since "c" isn't Greek... I'll get back to you when I understand the release naming :)
22:06 harmil_wk stmuk: is that "3-month"ly releases?
22:06 harmil_wk or 3 per month?
22:06 stmuk 1 per quarter (3 months)
22:06 harmil_wk that's what I thought, just clarifying
22:07 emdashcomma c is lunate sigma :)
22:07 bioduds yep, changed
22:07 bioduds going to test with latest now
22:12 abruanese joined #perl6
22:12 danlei joined #perl6
22:13 stmuk bioduds: "
22:13 stmuk oops
22:14 stmuk "You need gcc compiler installed in order to proceed" is only on linux (Mac will need clang installed)
22:15 danlei perl6 repl exits for me on some errors (e.g. when entering 'sub }'). this didn't happen with ubuntu's rakudo package, but with a fresh build (rakudobrew build moar). bug? if not, can it be disabled?
22:15 bioduds clang?
22:17 bioduds let me put that on the warnings then
22:19 danlei (2016.08-1-71-g685cc5a/MoarVM 2016.08)
22:21 jnthn danlei: Hm, curious, "sub }" and enter doesn't do it on what I've got locally, which is 2016.08.1-22-g6e76f7e
22:21 jnthn danlei: So sounds like a recent regression...please file a bug report.
22:21 bioduds i believe my fedora perl installation is the problem
22:22 bioduds in fedora install failure
22:22 bioduds i did sudo yum install perl
22:22 bioduds perhaps i shouldn't sudo i guess
22:22 jnthn (That describe is a tad misleading fwiw, 'cus I'm in a branch...)
22:22 bioduds so far ubuntu and centos installations were successful
22:24 stmuk bioduds: clang is an alternative compiler to gcc now used on Mac and FreeBSD and confusingly accessible as "gcc"
22:24 stmuk I'll try it on a Mac
22:24 bioduds great
22:24 bioduds i cant put up a mac vm, its too complicated
22:24 bioduds hackintosh
22:25 bioduds line now is: curl http://67.205.136.118/install.sh | sh;. ~/rakudo/setpath.sh
22:25 bioduds later hopefully will become: curl https://install.perl6.org | sh;. ~/rakudo/setpath.sh
22:25 danlei jnthn: ok, thanks. I'll report it.
22:26 bioduds great
22:26 bioduds if you can, send me results to bioduds@gmail.com
22:26 bioduds im curious to see if it will work on mac
22:26 cpage_ joined #perl6
22:26 El_Che bioduds: you should make the install dir overwritable by a ENV setting (as rakudobrew and perlbrew). Personally I hate stuff in my $HOME (they lib under ~/lib)
22:27 jnthn danlei: Thanks!
22:27 MilkmanDan joined #perl6
22:28 stmuk bioduds: that script should probably exit with a warning if the directory created by tar already exists
22:28 bioduds well, I was rm -rf ~/rakudo but i believe nine told me not to
22:28 stmuk sorry I mean the ~/rakudo directory
22:29 bioduds it should, I believe, clean up before install yes
22:29 bioduds I do rm -rf if it fails though
22:29 stmuk yeah I would favour testing for the existance of ~/rakudo, you might delete existing work under ~/rakudo
22:29 bioduds sure, cause otherwise it would fail
22:29 bioduds im putting that in
22:29 stmuk someone might be using that with programs existing there and have forgotten
22:29 El_Che with rm it's better to err on the safe side
22:30 bioduds correct
22:30 bioduds I'll think about it
22:30 bioduds gotta go now, be back soon!
22:30 bioduds thanks all guys
22:31 stmuk bioduds: you have an old 2016.04 reference there
22:33 dalek perl6-most-wanted: cd3e174 | (Tom Browder)++ | most-wanted/modules.md:
22:33 dalek perl6-most-wanted: update
22:33 dalek perl6-most-wanted:
22:33 dalek perl6-most-wanted: add a new WIP module: Net::IP
22:33 dalek perl6-most-wanted: delete an unneeded module: Tes::Number::Delta (all function now included in core module Test)
22:33 dalek perl6-most-wanted: review: https://github.com/perl6/perl6​-most-wanted/commit/cd3e1746b1
22:38 zacts joined #perl6
22:40 cpage_ joined #perl6
22:41 AlexDaniel m: my $c = Channel.new; $c.send(42); $c.close; say WHAT $c.list
22:41 camelia rakudo-moar fbeadb: OUTPUT«(List)␤»
22:41 labster joined #perl6
22:44 AlexDaniel m: my $c = Channel.new; start { $c.send(42); sleep 3; $c.close; }; say $c.list; say now - BEGIN now
22:44 camelia rakudo-moar fbeadb: OUTPUT«(42)␤3.0148542␤»
22:44 AlexDaniel m: my $c = Channel.new; start { $c.send(42); sleep 3; $c.close; }; say $c.list.WHAT; say now - BEGIN now
22:44 camelia rakudo-moar fbeadb: OUTPUT«(List)␤0.0146833␤»
22:44 AlexDaniel I don't understand… How can a List behave like that?
22:45 masak TimToady: questions from charging into the unknown: do you think `is parsed` should be made to work for non-macro subs? do you think `is parsed` has a place on (say) `constant` declarations and/or routine parameters? if possible, please answer modulo any reservations you may have to the viability of `is parsed`. ;)
22:45 dalek joined #perl6
22:48 masak (to my surprise, `is parsed` is turning out to be solving a lot of interesting problems in 007, at least on paper)
22:49 jnthn masak: Do you handle the parse regex having access to the tokens of the grammar, ooc?
22:52 perlawhirl joined #perl6
22:53 perlawhirl hi perlers
22:53 mcmillhj joined #perl6
22:54 perlawhirl ab5tract: Need to update you META6 for Terminal::Print to add OO::Monitors as a dependency
22:54 perlawhirl .seen ab5tract
22:54 yoleaux I saw ab5tract 13 Aug 2016 16:17Z in #perl6: <ab5tract> Gar, how did those RHS parens slip in.. Nevermind, my reasoning was based on a grave misunderstanding
22:55 * TimToady is still deeply suspicious of 'is parsed', but has to run off to play the drums now
22:56 perlawhirl .tell ab5tract You need to update you META6 for Terminal::Print to add OO::Monitors as a dependency
22:56 yoleaux perlawhirl: I'll pass your message to ab5tract.
22:59 bioduds stmuk, no, i changed already
22:59 bioduds to 2016.07
23:02 AlexDaniel oh by the way
23:02 AlexDaniel commit: pre-glr,HEAD my @a = <a b c>; .say for 42, @a
23:02 committable6 AlexDaniel, ¦«pre-glr»: 42␤a␤b␤c␤¦«HEAD»: 42␤[a b c]
23:02 masak jnthn: it's not implemented yet, but (far as I can see) in 007, having access to the tokens of the grammar is higher on the list than having access to lexically defined regexes.
23:02 AlexDaniel commit: 2015.07 say ‘Hello from the past!’
23:02 committable6 AlexDaniel, ¦«2015.07»: Hello from the past!
23:03 masak jnthn: an `is parsed` regex will operate with access to the grammar corresponding to the language currently in effect.
23:04 masak or, perhaps better to say, the parser currently in effect.
23:05 masak heh, I asked TimToady to answer modulo reservations, and all I got was the reservations :P
23:13 skids joined #perl6
23:29 pierre_ joined #perl6
23:30 kurahaupo joined #perl6
23:33 cpage_ joined #perl6
23:33 mcmillhj joined #perl6
23:34 cdg joined #perl6
23:35 cpage_ joined #perl6
23:48 pierre___ joined #perl6
23:54 zengargoyle joined #perl6
23:55 japhb Isn't that precisely the answer that modulo *can't* give?

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

Perl 6 | Reference Documentation | Rakudo