Camelia, the Perl 6 bug

IRC log for #perl6, 2014-09-21

Perl 6 | Reference Documentation | Rakudo | Niecza | Specs

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

All times shown according to UTC.

Time Nick Message
00:01 dalek specs: 9b8da72 | (Stéphane Payrard)++ | S03-operators.pod:
00:01 dalek specs: corrected meaning of .^ operator
00:01 dalek specs: review: https://github.com/perl6/specs/commit/9b8da72128
00:02 xenoterracide_ joined #perl6
00:03 cognome joined #perl6
00:06 aborazmeh joined #perl6
00:36 colomon joined #perl6
01:05 colomon joined #perl6
01:06 BenGoldberg m: &dd.WHAT.say
01:06 camelia rakudo-moar e63078: OUTPUT«(Sub)␤»
01:07 BenGoldberg m: dd(2..3);
01:07 camelia rakudo-moar e63078: OUTPUT«No such method 'name' for invocant of type 'Range'␤  in sub dd at src/gen/m-CORE.setting:2561␤  in block <unit> at /tmp/QNp14g1IjH:1␤␤»
01:08 BenGoldberg m: my $r = 2..3; dd($r);
01:08 camelia rakudo-moar e63078: OUTPUT«$r = 2..3␤»
01:08 BenGoldberg m: dd my $r = 2..3;;
01:08 camelia rakudo-moar e63078: OUTPUT«$r = 2..3␤»
01:08 FROGGS_ joined #perl6
01:13 eupcan277 joined #perl6
01:15 raiph joined #perl6
01:36 timotimo dd should probably check if there's a .VAR available
01:36 timotimo as in $the_thing.VAR ~~ Scalar or something
01:38 japhb Was someone trying to reach me?  This window had a highlight, but I don't see anything in my scrollback (which doesn't go back very far in clock time, you guys must have been chatty the last couple days.  :-)
01:42 timotimo hey japhb
01:46 timotimo have you seen my latest benchmarks?
01:46 japhb timotimo: Last I saw were from early this week.
01:46 timotimo http://t.h8.lv/p6bench/2014-​09-20-history_rakudo_2.html - check out the spikes upwards for push-related benchmarks
01:46 timotimo FROGGS has been notified about the problems with the json parse benchmarks and was investigating and improving that somewhat earlier today
01:46 japhb timotimo: Woah, nice!
01:46 japhb And I was just about to ask about the JSON parsing.  :-)
01:47 ilbot3 joined #perl6
01:47 Topic for #perl6 is now »ö« Welcome to Perl 6! | http://perl6.org/ | evalbot usage: 'perl6: say 3;' or rakudo:,  niecza:, std:, or /msg camelia perl6: ... | irclog: http://irc.perl6.org | UTF-8 is our friend!
01:48 japhb Which would be really nice on both points
02:00 masak morning, #perl6
02:01 japhb o/
02:27 colomon_ joined #perl6
02:31 * masak is intrigued by https://rt.perl.org/Ticket/Display.html?id=122810
02:33 jnthn morning, #perl6
02:35 jnthn masak: That one looks pretty odd...
02:35 ingy morning jnthn
02:37 masak before I had read through all of it, I had already done the experiment and golfed it down to the same linecount (5655) as in the bug report.
02:37 araujo joined #perl6
02:41 noganex_ joined #perl6
02:48 xenoterracide_ joined #perl6
02:57 xenoterracide_ joined #perl6
03:05 xenoterracide_ joined #perl6
03:09 xenoterracide_ joined #perl6
03:17 eupcan277_d joined #perl6
03:30 xenoterracide_ joined #perl6
03:31 jnthn Curiously,
03:31 jnthn type src\gen\m-CORE.setting | perl6-m -e "(bag slurp.words)"
03:31 jnthn Too many arguments in flattening array.
03:32 jnthn sub bag(*@args --> Bag) { Bag.new(|@args) }
03:33 jnthn eek, flattening a slurpy means grabbing all its elems in flattening context...
03:33 jnthn Which will be slow and hit limits.
03:34 jnthn And method new already slurps so there's no point to the flattening there at all
03:39 * jnthn finally manages to also download the gutenburg book from the ticket
03:42 jnthn yay, with my changes it works, and gives results in 6.4s
03:42 * jnthn profiles it out of curiosity
03:49 japhb How is it looking?
03:49 jnthn Most time is spent constructing the bag, I think, but that involves doing the words bit too
03:54 jnthn An obvious optimization in sort gets us to 5.8s.
04:00 jnthn The GC is quite fun to watch in the profile
04:00 jnthn While we're building the bag some objects are being retained
04:01 jnthn Once we reach the later part of the program where it's just sorting, we're promoting nothing.
04:02 jnthn Anyway, seems I have a bug fix and two performance improvements. I'll take that. Spectest time. :)
04:03 jnthn Building the profiler is one of the best time investments I've made of late... :)
04:03 eupcan277 joined #perl6
04:18 dalek rakudo/nom: a8267b4 | jonathan++ | src/core/List.pm:
04:18 dalek rakudo/nom: Bring a comment in line with reality.
04:18 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/a8267b418c
04:18 dalek rakudo/nom: 0d4bc70 | jonathan++ | src/core/set_operators.pm:
04:18 dalek rakudo/nom: Don't fatten args to Bag.new/Mix.new.
04:18 dalek rakudo/nom:
04:18 dalek rakudo/nom: They already slurp, so there's no reason to do this. In fact, there's
04:18 dalek rakudo/nom: a very good reason not to: flattening the args may produce a huge
04:18 dalek rakudo/nom: number of them, which is painful performance wise and hits limits
04:18 dalek rakudo/nom: on JVM and on MoarVM too. This appears to fix the issue reported in
04:18 dalek rakudo/nom: RT#122810.
04:18 synopsebot Link: https://rt.perl.org/rt3//Publ​ic/Bug/Display.html?id=122810
04:18 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/0d4bc7031f
04:19 dalek rakudo/nom: 39a299f | jonathan++ | src/core/List.pm:
04:19 dalek rakudo/nom: Optimize sorting block.
04:19 dalek rakudo/nom:
04:19 dalek rakudo/nom: Since we know we can never be < 0 because we're sorting an indices
04:19 dalek rakudo/nom: array, we can go directly to .at_pos.
04:19 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/39a299fae0
04:42 eupcan277_y joined #perl6
05:20 eupcan277_y joined #perl6
05:24 japhb Yes, time spent on the profiler is pretty much near the top of your good ideas this summer.  ;-)
05:24 japhb (And that's saying something!)
05:51 nwc10 T-0:50
05:51 nwc10 http://www.spacex.com/webcast/
05:52 nwc10 as in 20 seconds now
05:58 kaare_ joined #perl6
06:22 telex joined #perl6
06:24 uniejo joined #perl6
06:26 mberends joined #perl6
06:35 mr-foobar joined #perl6
06:54 masak https://twitter.com/SpaceX​/status/513572559611105280 is truly gorgeous.
06:58 jnthn Wow!
06:59 jnthn long exposure ftw
07:06 darutoko joined #perl6
07:25 FROGGS_ timotimo: no, was sleepy... will do that today
07:26 dalek star: f21eb1c | (Tobias Leich)++ | tools/star/release-guide.pod:
07:26 dalek star: volunteer for this month's release
07:26 dalek star: review: https://github.com/rakudo/star/commit/f21eb1c7c1
07:33 jnthn FROGGS_++
07:38 jaffa4 joined #perl6
08:00 denis_boyun joined #perl6
08:04 spider-mario joined #perl6
08:05 mr-foobar joined #perl6
08:16 mr-foobar joined #perl6
08:30 Juerd joined #perl6
08:32 kjs_ joined #perl6
08:36 FROGGS_ jnthn: there is an nqp::coerce_sI, but why isnt there an nqp::coerce_si?
08:37 FROGGS_ I'd like to use that in .MATCH
08:43 Juerd joined #perl6
08:46 denis_boyun_ joined #perl6
08:51 denis_boyun joined #perl6
08:52 synopsebot joined #perl6
08:53 [Coke] joined #perl6
08:53 masak joined #perl6
08:53 PerlJam joined #perl6
08:55 Util joined #perl6
08:56 leont joined #perl6
08:57 tadzik joined #perl6
08:57 pmichaud joined #perl6
08:59 sergot joined #perl6
09:02 Ven joined #perl6
09:04 hugme joined #perl6
09:04 Juerd joined #perl6
09:04 dalek joined #perl6
09:08 mr-foobar joined #perl6
09:14 kjs_ joined #perl6
09:15 SamuraiJack_ joined #perl6
09:30 dalek rakudo/optimatch: 215672b | (Tobias Leich)++ | src/core/Cursor.pm:
09:30 dalek rakudo/optimatch: first (untested) optimized version of .MATCH
09:30 dalek rakudo/optimatch: review: https://github.com/rakudo/rakudo/commit/215672bd71
09:31 SamuraiJack_ joined #perl6
09:43 FROGGS_ why do we have native nums in that bloody $curcap?
09:49 denis_boyun joined #perl6
09:55 denis_boyun joined #perl6
10:04 SamuraiJack_ joined #perl6
10:12 erkan joined #perl6
10:26 nine ggoebel11111113: I'm not sure if that would be really useful. Right now you couldn't use any Perl 6 modules in your script if Inline::Perl5 were executing it right from the start. Since you need to have Perl 5 installed to use Inline::Perl5, it would be more useful to specify perl 5 in your shebang line.
10:28 nine ggoebel11111113: But now that I think of it: accessing Perl 6 modules from Perl 5 code running in Inline::Perl5 would be rather simple. I'd just have to create a virtual "Perl6" module with useful methods like "use" and "invoke" same as I do it with Perl6::Object already.
10:34 kjs_ joined #perl6
10:34 denis_boyun_ joined #perl6
10:34 dalek rakudo/nom: 9aa8f6e | (Elizabeth Mattijsen)++ | src/core/IO/Spec/Win32.pm:
10:34 dalek rakudo/nom: Fix another case of double slurpy
10:34 dalek rakudo/nom:
10:34 dalek rakudo/nom: There shouldn't be any more in the core now.
10:34 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/9aa8f6e442
10:44 carlin ... It's alive!
10:44 carlin https://github.com/carbin/stomp
10:45 lizmat carlin++
10:46 lizmat m: say 0 ...^ *>5   # why is this warning ?
10:46 camelia rakudo-moar 39a299: OUTPUT«Code object coerced to string (please use .gist or .perl to do that)  in block <unit> at /tmp/0Jy5yxhEiv:1␤␤0 1 2 3 4 5␤»
10:47 lizmat r: say 0 ...^ *>5
10:47 camelia rakudo-{parrot,moar} 39a299: OUTPUT«Code object coerced to string (please use .gist or .perl to do that)  in block <unit> at /tmp/tmpfile:1␤␤0 1 2 3 4 5␤»
10:47 camelia ..rakudo-jvm 39a299: OUTPUT«Can't call method "syswrite" on an undefined value at /home/p6eval/jvm-rakudo/eval-client.pl line 32.␤»
10:47 lizmat at least it's consistent
10:51 Alina-malina joined #perl6
11:03 Ven joined #perl6
11:05 jaffa4 has anyone tried to program together?
11:05 jaffa4 using some kind of tool?
11:05 jaffa4 to fix bugs in Perl6 compiler?
11:06 * lizmat has only pair-programmed a few times IRL
11:08 jaffa4 irl?
11:08 lizmat In Real Life
11:08 jaffa4 do you have other lives as well?
11:08 lizmat some might argue I also have a life online
11:09 * grondilu thinks that's the point of the expression "IRL"
11:09 jaffa4 virtula life
11:16 dalek rakudo/nom: e71a9c1 | (Elizabeth Mattijsen)++ | src/core/Variable.pm:
11:16 dalek rakudo/nom: Give a better error for 'my int $a is default(*)'
11:16 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/e71a9c117f
11:20 lizmat TimToady: S02:1409 (Each) feels like a fossil
11:20 synopsebot Link: http://perlcabal.org/syn/S02.html#line_1409
11:21 lizmat at least I see no other mention of it apart from there, and no sign of any implementation work
11:21 lizmat shall I eradicate Each ?
11:24 lizmat or, now that we have more async stuff in the core, is it time to start looking at Each / Junction orderedness / unorderedness
11:24 lizmat as atm I think Junctions are actually behaving as Each's ?
11:42 mattp_ joined #perl6
11:46 vendethiel carlin: https://github.com/carbin/stomp/​blob/master/lib/Stomp/CLI.pm6#L9
11:46 vendethiel why <>?
11:47 lizmat because he doesn't like quotes ?
11:47 carlin I prefer <> over ""
11:48 pmurias joined #perl6
11:49 nine Anyone have any idea about the fasibility of ggoebel11111113's idea to run scripts through Inline::Perl5 automatically if there's no indication that a script is written in Perl 6? S01 seems to require this form of support.
11:51 pmurias it's only supposed to happen when perl6 is invoked as perl
11:51 yoleaux 20 Sep 2014 20:28Z <raiph> pmurias: Does it make sense to remove https://github.com/perl6/simple-tests ?
11:55 nine pmurias: so I've read. Which would be fine with me. If invoked as perl and the file ending is not .pl6 or .pm6 and there's no 'use v6;' or equivalent automatically load Inline::Perl5 and use it to run the script.
11:55 spider-mario joined #perl6
11:56 pmurias but currently very few/nobody runs perl6 as perl
11:57 yeahnoob joined #perl6
11:58 virtualsue joined #perl6
11:59 nine pmurias: no wonder. It doesn't have any advantages and breaks Perl 5 scripts. But if we transparently support Perl 5 as required by S01 and lower startup time to comparable levels, this will be an important step to reunite the Perls and give up the sister languages nonsense :)
12:00 vendethiel https://github.com/carbin/stomp/b​lob/master/lib/Stomp/CLI.pm6#L218 could use MAIN :p
12:01 kurahaupo joined #perl6
12:05 pmurias the only way to give up the sister languages nonsense it to improve the implementation and ecosystem of Perl 6 to a point where it surpasses Perl 5
12:06 carlin vendethiel: could too, will have to look at that
12:06 vendethiel pmurias: well, got to give it some time
12:06 lizmat pmurias++  :-)
12:08 dalek specs: 3ceec03 | (Elizabeth Mattijsen)++ | S02-bits.pod:
12:08 dalek specs: Remove a bit of BagHash fossil
12:08 dalek specs:
12:08 dalek specs: Bags return pairs in list context now, as mentioned above.
12:08 dalek specs: review: https://github.com/perl6/specs/commit/3ceec03d9b
12:08 nine pmurias: I disagree. as soon as we become downward compatible again, we can start calling Perl 6 the newer version of Perl and Perl 5 the legacy version of Perl. Newer versions of anything don't necessarily have to be better at everything. Though of course it helps adoption a lot if they are.
12:10 jnthn FROGGS_: I think writing prefix:<~> in NQP actually gets you the MoarVM coerce_si equivalent
12:12 dalek specs: 7101879 | (Elizabeth Mattijsen)++ | S02-bits.pod:
12:12 dalek specs: Elaborate a bit about MixHash
12:12 dalek specs: review: https://github.com/perl6/specs/commit/7101879539
12:12 jnthn FROGGS_: As in, ~$foo
12:12 lizmat TimToady: seems like PairSeq is a bit of a fossil
12:12 lizmat unless you want to call the .list of Hash/Bag/Mix that ?
12:13 jnthn lizmat: iirc, the each thingy is some kind of junction-like construct that retains ordering
12:13 lizmat which is how Junction atm is implemented, no?
12:14 jnthn It's certainly not promised to be that way, though.
12:14 lizmat realize that, maybe we should randomize it before 6.0 to make sure we don't get more dependent code on that out there
12:15 jnthn Worth considering
12:17 jnthn http://perlcabal.org/syn/S09.html#Each calls out each as conjectural
12:19 vendethiel :o. does that work??
12:19 vendethiel m: (^10).each.say; #prolly not
12:19 camelia rakudo-moar 9aa8f6: OUTPUT«No such method 'each' for invocant of type 'Range'␤  in block <unit> at /tmp/RgMLbzixye:1␤␤»
12:20 lizmat hmmm...  so is that LHF ?
12:21 jnthn lizmat: I don't see it being hard, though it's probably a bit under-spec'd
12:21 FROGGS_ jnthn: I want to get an int from a str
12:21 jnthn Doesn't say what happens in boolen context for example
12:21 jnthn FROGGS_: oh...
12:21 lizmat agree about the underspeccing, hence my question to TimToady
12:21 FROGGS[mobile] joined #perl6
12:22 jnthn I'd be surprised if a more concrete spec for it yielded a heck of a lot more work though
12:22 zakharyas joined #perl6
12:23 lizmat well, unless perceived clarity != real clarity, so rather see the spec clarified first  :-)
12:23 jnthn FROGGS_: Ah...
12:24 jnthn m: nqp::bind(my $int $i, '42'); say $i;
12:24 camelia rakudo-moar 9aa8f6: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/elFLTw0JSfâ�¤Unable to parse expression in argument list; couldn't find final ')' â�¤at /tmp/elFLTw0JSf:1â�¤------> [32mnqp::bind(my $int [33mâ��[31m$i, '42'); say $i;[0mâ�¤    expecting any of:…»
12:24 jnthn oops
12:24 jnthn m: nqp::bind(my int $i, '42'); say $i;
12:24 camelia rakudo-moar 9aa8f6: OUTPUT«This type cannot unbox to a native integer␤  in block <unit> at /tmp/aCwCpCYDT_:1␤␤»
12:24 jnthn m: nqp::bind(my int $i, nqp::unbox_s('42')); say $i;
12:24 camelia rakudo-moar 9aa8f6: OUTPUT«42␤»
12:24 jnthn FROGGS_: That's a possible way
12:25 jnthn Agree it ain't entirely convenient
12:26 FROGGS[mobile] jnthn: don't has to be :o)
12:30 ggoebel11111113 nine: nice to hear that the run as perl5 by default migration aspect of http://perlcabal.org/syn/S01.html may be within reach. Seems like a pretty significant deliverable. Is this still a prerequisite for 6.0?
12:31 lizmat ggoebel11111113: good question, one I hope will be answered after APW
12:31 nine lizmat: looking forward to that discussion :)
12:31 pmurias nine: a big reason why we call Perl 6 a sister language instead of the newer version is that calling Perl 5 a legacy version hurts it's perception a lot
12:33 nine pmurias: currently, yes, absolutely. I think the sister languages thing was and probably still is necessary and good. But in the long term I'd really like to but it away and bring the Perl's together. It's a main motivation for writing Inline::Perl5 for me.
12:33 nine s/but/put/
12:33 gtodd and perl5 keeps borrowing and porting ideas back
12:33 lizmat well, some p5pers are leaving P5
12:33 lizmat the latest rurban: from a mail earlier today: "So I'll stop watching this mess now, goodbye 5.22 and p5p."
12:35 vendethiel what's going on in p5p?
12:35 pmurias rurban has been conflicted with p5p for a long time
12:35 lizmat yes, but it now seems he has given up on P5
12:35 ggoebel11111114 joined #perl6
12:35 pmurias he seems to be conflicted with MoarVM and diakopter
12:35 lizmat or at least p5p
12:36 lizmat some at p5p will say "good riddance"
12:37 lizmat I would not be one of them
12:40 woolfy rurban++
12:41 masak I think any animosity across the p5-p6 line is senseless and in the end a kind of own goal.
12:42 masak given that bubble-outsiders can barely tell p5ers and p6ers apart, having infighting across that line is only going to look like... structural weakness in the ranks.
12:43 ggoebel anything which makes the line thinner++
12:43 ggoebel and easier to cross in order to get work, play, and fun done...
12:43 masak what the p5 people should focus on is to be accepting of and curious about p6 while preserving their own cultural heritage.
12:43 lizmat going back to ggoebel's question wrt to starting in P5 land
12:44 lizmat I don't really see that happening unless we have "use v5"
12:44 lizmat or an Inline::Perl6 in P5 land
12:45 masak what the p6 people should focus on is to be awesome and bring more of the future into equal distribution... while trying not to say "we're the next version" until it already doesn't need saying.
12:45 gtodd I want p5 to adopt more bits of p6 ... like p6 formats :-)
12:45 masak gtodd: there's a CPAN module.
12:46 gtodd yup and it works great ... should be the default ;-)
12:46 gtodd there's rperl and p2 etc
12:48 masak and moe.
12:48 masak I don't have a sense of which state moe got left in.
12:48 ggoebel looks like perl4 is still hanging on... https://lintian.debian.org/tags/scri​pt-uses-perl4-libs-without-dep.html
12:48 raiph joined #perl6
12:49 tinyblak joined #perl6
12:49 gtodd isn't rurban involved with p2 and rperl?
12:50 lizmat p2 I know for sure, rperl I'm not so sure about
12:50 ggoebel IMHO perl6 is more likely to get installed (or linked) to /usr/bin/perl ...if s01 is implemented
12:52 lizmat what is a PairSet ?   seems like a fossil in S02:1691
12:52 synopsebot Link: http://perlcabal.org/syn/S02.html#line_1691
12:52 jnthn lizmat: Never heard of it :)
12:53 lizmat "Likewise, C<%x> may be bound to any object that does the C<Associative> role, such as C<Pair>, C<PairSet>, C<Set>, C<Bag>, C<Mix>, or C<Capture>.
12:53 jnthn Yeah. Looks like fossil
12:53 jnthn Just toss it or pick a more apporpriate example type
12:53 jnthn Also
12:53 jnthn uh, no also :)
12:54 dalek specs: f9ac5c9 | (Elizabeth Mattijsen)++ | S02-bits.pod:
12:54 dalek specs: Remove reference to PairSet, considered a fossil
12:54 dalek specs: review: https://github.com/perl6/specs/commit/f9ac5c9d9f
12:58 lizmat "    $^foo       self-declared formal positional parameter"
12:58 lizmat the ^ twigil ???
13:00 FROGGS[mobile] yes
13:01 lizmat ah, of courtse
13:01 lizmat course
13:01 lizmat duh  :-)
13:02 jnthn That one's even long implemented... :)
13:02 lizmat the phrasing is what too me off guard
13:02 lizmat *took
13:04 kaare_ joined #perl6
13:07 timotimo o/
13:09 lizmat timotimo o/
13:09 lizmat m: 'my $s = anon sub tract($n) { say $n; tract($n-1) if $n }; $s(5)
13:09 lizmat m: my $s = anon sub tract($n) { say $n; tract($n-1) if $n }; $s(5)
13:09 camelia rakudo-moar e71a9c: OUTPUT«(timeout)»
13:09 camelia rakudo-moar e71a9c: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/Wx9jTYZ8p7â�¤Undeclared routine:â�¤    tract used at line 1â�¤â�¤Â»
13:10 lizmat this should work according to S02:2035
13:10 synopsebot Link: http://perlcabal.org/syn/S02.html#line_2035
13:10 jnthn aye, nyi
13:10 lizmat is that a NYI or a rakudobug of anon sub
13:10 jnthn It was added to spec some time after anon subs worked in Rakudo, sa I remember.
13:10 jnthn LHFish
13:11 lizmat :-)
13:13 dalek specs: 886b1ec | (Elizabeth Mattijsen)++ | S02-bits.pod:
13:13 dalek specs: Make examples more cut/pastable
13:13 dalek specs: review: https://github.com/perl6/specs/commit/886b1ec4ac
13:15 dalek nqp-js: 46fc11b | (Pawel Murias)++ | src/vm/js/ (4 files):
13:15 dalek nqp-js: Implement nqp::{open,tellfh,readlinefh​,readallfh,printfh,closefh}
13:15 dalek nqp-js: review: https://github.com/pmurias​/nqp-js/commit/46fc11ba6d
13:15 dalek nqp-js: a7ad086 | (Pawel Murias)++ | src/vm/js/QAST/Compiler.nqp:
13:15 dalek nqp-js: Use a :sideffects modifier instead of add_sideffect_op.
13:15 dalek nqp-js: review: https://github.com/pmurias​/nqp-js/commit/a7ad086e51
13:15 dalek nqp-js: cf8a087 | (Pawel Murias)++ | src/vm/js/QAST/Compiler.nqp:
13:15 dalek nqp-js: Extra blank line.
13:15 dalek nqp-js: review: https://github.com/pmurias​/nqp-js/commit/cf8a087f35
13:15 dalek nqp-js: dc54025 | (Pawel Murias)++ | src/vm/js/QAST/Compiler.nqp:
13:15 dalek nqp-js: Refactor for the DRY principle.
13:15 dalek nqp-js:
13:15 dalek nqp-js: Omit the code generation callback for ops that just call a nqp.op.$opname
13:15 dalek nqp-js: review: https://github.com/pmurias​/nqp-js/commit/dc54025dff
13:16 pmurias masak: I think moe was abandoned
13:17 masak lizmat: looks like PairSet was supposed to go away as of cf42c47703cc126f6cad2c82d5d923e8ab00a85e
13:18 dalek specs: 281d8f1 | (Elizabeth Mattijsen)++ | S02-bits.pod:
13:18 dalek specs: Mention handling circular references in .perl
13:18 dalek specs: review: https://github.com/perl6/specs/commit/281d8f1877
13:18 masak lizmat: also, looking at the commits involved, I'm not at all sure it wasn't all a misspelling of "PairSeq"...
13:18 lizmat ah, hhmmm....
13:19 lizmat could be, but PairSeq also seems a fossil
13:19 masak yeah.
13:19 jnthn There once was a type called Seq, which is no more.
13:19 masak I currently lean towards "not a typo, but was renamed one to the other"
13:19 masak m: say Seq
13:19 camelia rakudo-moar e71a9c: OUTPUT«(Seq)␤»
13:19 BenGoldberg joined #perl6
13:19 masak Seq lives on as an empty class in the setting ;)
13:19 * masak .oO( ghost type )
13:19 lizmat and quite a few menitions in the spec
13:20 lizmat *mentions
13:20 masak I don't particularly like the meme "kill it with fire"... but at this point, I don't think we need the Seq type.
13:20 pmurias masak: looking at the moe tests it looks like a Perl5/Perl6 hybrid or some very early version of Perl 6
13:20 masak pmurias: aye.
13:21 masak pmurias: I get the feeling stevan realized that his project basically ended up borrowing a lot of ideas from Perl 6.
13:23 dalek specs: 66e6fe9 | (Elizabeth Mattijsen)++ | S03-operators.pod:
13:23 dalek specs: Assume PairSeq was typo for fossil PairSet
13:23 dalek specs: review: https://github.com/perl6/specs/commit/66e6fe9432
13:23 lizmat #dammitstevan  :-)
13:24 teodozjan joined #perl6
13:24 masak something like "a project that wants to be a linear combination of p5 and p6, but looks to p6 for sanity tiebreakers... ends up converging on p6"
13:25 masak or at least it becomes p6 without the risky/interesting parts, like roles and grammars.
13:25 jnthn m: say Seq.^mro # morbid curiosity...
13:25 camelia rakudo-moar e71a9c: OUTPUT«(Seq) (List) (Iterable) (Cool) (Any) (Mu)␤»
13:25 dalek nqp-js: ffc7791 | (Pawel Murias)++ | tools/build/ (2 files):
13:25 dalek nqp-js: Update the Makefile, so that it reinstalls the runtime when the new runtime files change.
13:25 dalek nqp-js: review: https://github.com/pmurias​/nqp-js/commit/ffc7791e26
13:26 jnthn Yeah, I think Seq's continued existence in Rakudo is fossil
13:26 teodozjan hi, is there any fast way to list loaded classes/modules? I tried to use @*INC but I didn't find any usable method on CompUnitRepo
13:26 nine lizmat: why would Inline::Perl6 be a prerequisite?
13:27 lizmat jnthn: I will remove Seq from setting first in one revertable commit, and then remove from spec later
13:27 jnthn lizmat: wfm
13:29 Ven joined #perl6
13:29 masak jnthn:
13:29 masak m: use MONKEY_TYPING; augment class Mu { multi method gist(Mu:U $type:) is default { "({$type ~~ Any ?? $type.^name !! "M" ~ "o" x 100})" } }; say Seq.^mro
13:29 camelia rakudo-moar e71a9c: OUTPUT«(Seq) (List) (Iterable) (Cool) (Any) (Mooooooooooooooooooooooooooooooooo​ooooooooooooooooooooooooooooooooooo​oooooooooooooooooooooooooooooooo)␤»
13:29 tadzik :D
13:31 timotimo m)
13:31 lizmat .oO( in dutch, the pronunciation of "Mu" very much sounds like the word for "tired" )
13:36 dalek rakudo/nom: 3eae3c3 | (Elizabeth Mattijsen)++ | src/core/ (3 files):
13:36 dalek rakudo/nom: Eradicate Seq with fire
13:36 dalek rakudo/nom:
13:36 dalek rakudo/nom: Oddly enough, src/core/Seq.pm was not included in an backend setting.
13:36 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/3eae3c3fcf
13:37 jnthn m: Mu.^set_name('M' ~ 'o' x 100); say Mu # anything you can do, I can do meta :P
13:37 camelia rakudo-moar e71a9c: OUTPUT«(Mooooooooooooooooooooooooooooo​oooooooooooooooooooooooooooooooooooooo​ooooooooooooooooooooooooooooooooo)␤»
13:37 dalek roast: c37d1b7 | (Elizabeth Mattijsen)++ | S02-types/WHICH.t:
13:37 dalek roast: Eradicate Seq with fire
13:37 dalek roast: review: https://github.com/perl6/roast/commit/c37d1b7e76
13:42 lizmat m: say (item <a b c>, <d e f>).WHAT   # would have expected Parcel here, not Array
13:42 camelia rakudo-moar e71a9c: OUTPUT«(Array)␤»
13:44 jnthn m: say (item <a b c>, <d e f>).perl
13:44 camelia rakudo-moar e71a9c: OUTPUT«["a", "b", "c", "d", "e", "f"]␤»
13:45 lizmat multi sub item(*@a)         { my $ = @a }
13:46 masak ooh, .^set_name
13:46 lizmat I would have expected a parcel of 2 parcels there
13:46 jnthn lizmat: Itemized, presumably?
13:46 jnthn lizmat: uh, the outer parcel I mean...
13:46 raiph joined #perl6
13:46 lizmat yes
13:46 jnthn m: say (<a b c>, <d e f>).WHAT
13:46 camelia rakudo-moar e71a9c: OUTPUT«(Parcel)␤»
13:46 jnthn opos
13:47 vendethiel the call binds loosely :)
13:47 jnthn m: say (<a b c>, <d e f>).item.WHAT
13:47 camelia rakudo-moar e71a9c: OUTPUT«(Parcel)␤»
13:47 jnthn m: say (<a b c>, <d e f>).item.perl
13:47 camelia rakudo-moar e71a9c: OUTPUT«$(("a", "b", "c"), ("d", "e", "f"))␤»
13:47 jnthn lizmat: Like this I guess... :)
13:47 lizmat yup
13:47 jnthn I think that makes sense
13:47 jnthn Can be done with the vmargarray trick
13:48 jnthn Same one infix:<,> uses I guess
13:48 lizmat vmargarray ??
13:48 lizmat missp?
13:48 jnthn my sub infix:<,>(|) is pure { nqp::p6parcel(nqp::p6argvmarray(), nqp::null());
13:48 jnthn }
13:49 jnthn bah, linebreak got lost.
13:49 vendethiel m: .say for @((<a b c>, <d e f>).item)
13:49 camelia rakudo-moar e71a9c: OUTPUT«a␤b␤c␤d␤e␤f␤»
13:49 jnthn We may want to define
13:49 jnthn multi item(\x) { x.item }
13:50 jnthn multi item(|) { my $ = nqp::p6parcel(nqp::p6argvmarray(), nqp::null()); }
13:50 jnthn I think those two between them would give sane results. Worth a try, at lesat...
13:50 lizmat trying
13:50 jnthn lizmat++
13:52 lizmat x.item doesn't make sense, as Mu.item just returns x
14:00 jnthn lizmat: That's a totally fine answer in general.
14:00 jnthn lizmat: item 42 should just be 42
14:01 jnthn lizmat: Things that don't behave like items should already be overriding item to do the righ tthing.
14:01 jnthn *right
14:01 jnthn m: say (1,2,3).item.perl # as Parcel does, for example
14:01 camelia rakudo-moar e71a9c: OUTPUT«$(1, 2, 3)␤»
14:02 lizmat hmmm... we seem to have a spectest testing that (item $a, $b) makes an Array
14:02 lizmat so I can assume the spectest is faulty
14:03 jnthn Yeah, that feels a bit off to me.
14:03 lizmat that was the only spectest issue
14:04 jnthn OK, in that case I think the change is kosher
14:14 gtodd jnthn: lizmat: it appears that rakudo/rakudo/commit/a8267b418c does fix RT#122810 (possibly my first bug report) but I'm trying to grok the commit message: "Don't fatten args to Bag.new/Mix.new. They already slurp, so there's no reason to do this ..."
14:14 synopsebot Link: https://rt.perl.org/rt3//Publ​ic/Bug/Display.html?id=122810
14:15 lizmat s/fatten/flatten
14:15 jnthn omg, did I really type fatten?
14:15 lizmat yesh
14:15 gtodd fatten should be flatten ;-) but ... hehe so they were getting flattened twice in some sort of loop
14:15 jnthn *lol*
14:15 jnthn fail
14:16 gtodd jnthn: sometimes a typo is just a typo ... sometimes it's a freudian slip :-D
14:16 jnthn gtodd: the issue is sub foo(*@a) { Foo.new(|@a) }, then call foo(@a_really_huge_array) will go off and try to make a callsite that is array-sized.
14:17 jnthn Which is a hugely inefficient thing to do once the array gets large.
14:17 gtodd oh ... for no reason ?  anyway yeah it fixes it and makes (or something else does) it faster
14:17 jnthn Reading being that the slurpy flattens.
14:17 jnthn *Reason
14:17 gtodd excellent ... my bug report worked!! :-D
14:17 jnthn I should probably give up on typing today. :)
14:18 jnthn Glad it's faster. It ran in < 6s on my laptop here, after I did the .sort opt too...
14:18 gtodd heh I have no idea what my old rt btcard account even was but so far with my new ID bugs reported resulting in a fix I'm batting 100 heheh
14:19 jnthn Talking of bug reports, RT#122756 is a fun one...
14:19 synopsebot Link: https://rt.perl.org/rt3//Publ​ic/Bug/Display.html?id=122756
14:19 jnthn The analysis suggested by the subject of the ticket is actually wrong
14:19 gtodd oopsie I changed the command  perl6 -e '(bag slurp.words.grep( *.chars > 5 )).pairs.sort( *.value).reverse[^10].say'; probably the reason
14:19 gtodd why it was faster
14:19 jnthn m: role Foo {}; say .WHAT for pi but Foo, [pi] but Foo;
14:19 camelia rakudo-moar e71a9c: OUTPUT«(Num+{Foo})␤(Num)␤»
14:20 ajr joined #perl6
14:20 dalek rakudo/nom: aac70ae | (Elizabeth Mattijsen)++ | src/core/ (2 files):
14:20 dalek rakudo/nom: item() returns Parcel, like .item already did
14:20 dalek rakudo/nom:
14:20 dalek rakudo/nom: This part of the Seq cleanup
14:20 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/aac70aeb54
14:20 jnthn The reason it's Num the second time around is because the but does .clone() on what it receives
14:21 jnthn And so we lose the itemization and flatten the but'ed thing...
14:21 dalek roast: 0bc42d6 | (Elizabeth Mattijsen)++ | S03-operators/context.t:
14:21 dalek roast: item() returns a Parcel, not an Array
14:21 dalek roast: review: https://github.com/perl6/roast/commit/0bc42d6df9
14:22 jnthn It's not entirely clear to me where to fix this. We typically have a policy that .method busts its way through item distinctions.
14:22 jnthn Real hinges on
14:22 jnthn m: say [1,2,3].clone().perl
14:22 camelia rakudo-moar e71a9c: OUTPUT«Array.new(1, 2, 3)␤»
14:22 jnthn Is that correct, or should it be [1, 2, 3]?
14:23 jnthn TimToady: ^^ could use your brane for a moment, if you can spare it :)
14:23 gtodd RT question ... as the reporter can/should I mark the bug closed or do you have workflow /method for that?
14:23 jnthn gtodd: Typically we like to have tests for things
14:23 jnthn gtodd: Can makr it testneeded
14:24 gtodd hmm ok and then somehow find a test that confirms RT#122810 is fixed and then close ... ok
14:24 synopsebot Link: https://rt.perl.org/rt3//Publ​ic/Bug/Display.html?id=122810
14:25 jaffa4 I submitted a bug report, any idea when it is going to be fixed?
14:25 gtodd so should I put "testneeded" in the subject line and change status to "fixed"?
14:26 gtodd ermm "resolved" ?
14:26 jaffa4 its number is  #122815
14:26 synopsebot Link: https://rt.perl.org/rt3//Publ​ic/Bug/Display.html?id=122815
14:26 * nine thinks that as long as a bug doesn't have a test it cannot be resolved because nothing makes sure it _stays_ resolved.
14:28 mr-foobar joined #perl6
14:28 gtodd yeah ... ok I will just add a comment
14:28 gtodd and leave it as "open"
14:29 jnthn jaffa4: I actually have that one open in a browser tab. The answer, I *think*, is that you need to put the "is repr" thing on the pre-decl. But we really should catch the use of it on the actual declaration and point out that it isn't going to work, instead of just ignoring it.
14:30 PZt joined #perl6
14:31 jnthn jaffa4: So, likely conclusion is that the program is incorrect as it stands, but Rakudo was wrong to not point this out.
14:31 jaffa4 so it also needs to be a cpointer, jnthn
14:31 jnthn I didn't have time to try it, but
14:31 TimToady maybe .cone is really .CLONE
14:31 TimToady *l
14:31 jnthn class TopWindow is repr('CPointer') {...} should do it
14:32 jaffa4 what is repr, by the way?
14:32 jnthn repr = representation = how this type should be represented in memory
14:33 jnthn It needs to go on the first mention, since that's where we create the type object, and we need to create that so we've something to install so mentions of TopWindow will be recognized.
14:34 jnthn Note it's only needed in the pre-decl, so you don't need to have it in both places.
14:35 SamuraiJack_ joined #perl6
14:35 jnthn Let me know if that doesn't work out, but I'm pretty sure that's what is happening.
14:35 jnthn TimToady: English spoken loudly enough will be understood anywhere? ;)
14:35 TimToady SI!
14:36 jnthn TimToady: I don't think it needs to be macro-ish, in so far as a method can happily say method clone(\self-with-cont:) { } to recover what was actually the invocant.
14:37 jaffa4 I get Unable to parse class definition -> ------> class Ctrl ⏏s repr('CPointer') {...}
14:37 jnthn Did you actually type is, or s?
14:38 jnthn m: class Ctrl is repr('CPointer') { ... }; class Ctrl { }
14:38 camelia rakudo-moar 3eae3c: ( no output )
14:38 jnthn m: class Ctrl is repr('CPointer') { ... }; class Ctrl { }; say Ctrl.REPR
14:38 jaffa4 yes, that is typo.. so not an error
14:38 camelia rakudo-moar 3eae3c: OUTPUT«CPointer␤»
14:39 jnthn TimToady: Mostly I'm just wondering if we want to make clone not lose itemization, or if it's but that wants to change...
14:39 kst joined #perl6
14:39 * TimToady isn't awake enough to have an opinion on that
14:39 jnthn Or if there's a third thing I haven't thought of. I'm pretty sure the behavior reported in the RT is wrong though.
14:39 timotimo m: class Test is repr('CPointer') { }; say Test.REPR.WHAT.perl
14:39 camelia rakudo-moar 3eae3c: OUTPUT«Cannot find method 'perl': no method cache and no .^find_method␤  in block <unit> at /tmp/Iobu26_UTg:1␤␤»
14:39 jnthn np :)
14:39 timotimo m: class Test is repr('CPointer') { }; say Test.REPR.WHAT
14:39 camelia rakudo-moar 3eae3c: OUTPUT«(Str)␤»
14:40 timotimo we don't really have objects to represent representations
14:40 jnthn No
14:40 jnthn They're just string identifiers
14:40 jnthn Things have to bottom out somewhere... :)
14:41 timotimo aye
14:41 mr-foobar joined #perl6
14:41 timotimo at least we can do stuff like class Foo is repr(Otherclass.REPR) like this :)
14:42 jnthn Well, and once the S12 package decl stuff is in place we can also consider cpointer Foo { } as the declaration too :)
14:42 jnthn EXPORTHOW::DECLARE iirc :)
14:42 timotimo ah, sure
14:42 jaffa4 jhthn: your suggestion seems to work
14:43 jnthn jaffa4: OK, good. I'll keep the ticket handy, 'cus I want to put something in the compiler so others don't fall into the same trap and have things silently not work...
14:49 kjs_ joined #perl6
15:00 SamuraiJack joined #perl6
15:04 mr-foobar joined #perl6
15:11 jnthn Time for some rest...'night
15:12 timotimo gnite jnthn
15:16 lizmat good night. jnthn
15:18 lizmat cycling&
15:30 xenoterracide_ joined #perl6
15:31 _rubio_terra_ joined #perl6
15:45 vendethiel .oO( we need a haskell backend because we have liquid types )
15:49 leont joined #perl6
15:53 vendethiel (actually, refinement types)
15:57 xenoterracide_ joined #perl6
16:13 kjs_ joined #perl6
16:18 erkan joined #perl6
16:18 erkan joined #perl6
16:19 xenoterracide_ joined #perl6
16:24 dalek rakudo/optimatch: f56eb0f | (Tobias Leich)++ | src/core/Cursor.pm:
16:24 dalek rakudo/optimatch: use nqp::bind to coerce str to int, jnthn++
16:24 dalek rakudo/optimatch: review: https://github.com/rakudo/rakudo/commit/f56eb0f298
16:34 toysrough joined #perl6
17:03 cognome joined #perl6
17:05 liztormato joined #perl6
17:13 eupcan277_y joined #perl6
17:16 Ven joined #perl6
17:16 brrt joined #perl6
17:26 TimToady liztormato: Each is not redunant with Junction, since junctions do something in a boolean context, while each does something in list context
17:27 TimToady *redundant needs redundant d's
17:27 liztormato ;-)
17:28 liztormato That was not really clear to me from the spec
17:28 TimToady btw, we'll be commuting to NJ over the next ten days, so connectivity is likely to be spotty out in the wilds
17:28 TimToady well, it's marked conjectural till someone implements it :
17:28 TimToady :)
17:29 liztormato But maybe it's conjecturally specced too vaguely to get implemented
17:30 liztormato Commuting to NJ by car I assume?
17:31 TimToady yup
17:31 TimToady note to would-be burglars, there are still people at home too. :)
17:32 liztormato Wow. It's on my bucket list to do a cross country ride like that
17:32 TimToady well, it's my bucket list to hit all the states I've never been in, and that I'll do this trip, if all goes according to plan
17:33 liztormato Cool!
17:40 TimToady "LHFish" <-- Low Hanging Fish?
17:42 ponbiki joined #perl6
17:43 TimToady Parcel has basically taken over the tuple semantics from Seq, so we don't need Seq anymore
17:43 vendethiel NJ?
17:43 liztormato Ishisms everywhere!
17:43 liztormato New Jersey
17:44 TimToady my son Aron has relocated to IAS at Princeton, so we're taken them their car, which they left behind in SoCal
17:44 liztormato TimToady: I will fight Seq with fire in the spec
17:46 TimToady afk &
17:48 vendethiel .tell masak I'm gonna make macros to join our `where`s and that: http://goto.ucsd.edu/~rjhala/liquid/haskell/b​log/blog/2013/01/01/refinement-types-101.lhs/
17:48 yoleaux vendethiel: I'll pass your message to masak.
17:48 vendethiel :P
17:51 dalek nqp-js: 4bb73ad | (Pawel Murias)++ | src/vm/js/ (2 files):
17:51 dalek nqp-js: Implement nqp ops: symlink, link, unlink, setencoding.
17:51 dalek nqp-js:
17:51 dalek nqp-js: Make nqp::stat with an EXISTS flag return true when it encounters a symlink that points to a missing file.
17:51 dalek nqp-js: review: https://github.com/pmurias​/nqp-js/commit/4bb73ade66
17:53 xenoterracide_ joined #perl6
17:54 liztormato Continuing cycling&
17:54 liztormato left #perl6
18:10 kjs_ joined #perl6
18:16 Esgorhannoth joined #perl6
18:16 eupcan277_y_h joined #perl6
18:18 Esgorhannoth left #perl6
18:22 molaf__ joined #perl6
18:22 telex joined #perl6
18:24 brrt left #perl6
18:36 dwarring joined #perl6
18:37 virtualsue joined #perl6
18:39 pmurias is it possible that an say is executed but it doesn't get to print anything because of buffering and an exception
18:39 pmurias on oldish MoarVM
18:42 dalek perl6-bench: b980a1d | (Geoffrey Broadwell)++ | EXPECTED:
18:42 dalek perl6-bench: Mark EXPECTED out of date and call for volunteers to update it
18:42 dalek perl6-bench: review: https://github.com/japhb/pe​rl6-bench/commit/b980a1daa8
18:45 anaeem1 joined #perl6
18:53 ifim joined #perl6
19:01 lizmat pmurias: could be...  jnthn or brrt or timotimo might have a better idea
19:01 japhb Something is still goofy with rc-forest-fire on current nqp-m and perl6-m.  If I do a nice large field, like `perl6-m perl6/rc-forest-fire 79 40 100`, it's slow enough to see that sometimes the fire doesn't propagate to all the places it should.  And under nqp-m, it doesn't take long for the screen to get jumpy and then corrupted.
19:01 japhb TimToady, jnthn: ^^^
19:04 [Sno] joined #perl6
19:04 eupcan277_y_h_h joined #perl6
19:06 guru joined #perl6
19:07 timotimo pmurias: yes, indeed
19:08 timotimo pmurias: pipe it into a file or strace it
19:08 timotimo that'll force it to sync it out
19:12 vendethiel http://colinm.org/language_checklist.html
19:13 lizmat m: say Bag(3,8,[2,(9,3)],8) >>->> 1   # according to S03:4415, this should be Bag(2,7,[1,(8,2)],7)
19:13 synopsebot Link: http://perlcabal.org/syn/S03.html#line_4415
19:13 camelia rakudo-moar aac70a: OUTPUT«"2 9 3" => 0, "3" => 0, "8" => 1␤»
19:14 lizmat I feel the spec is wrong, TimToady do you agree ?
19:15 woolfy Just saw somebody post this on the Perl Developers page on Facebook: "If programming languages were vehicles": http://s3.crashworks.org.s3-website-us-east-1.amaz​onaws.com/if-programming-languages-were-vehicles/
19:15 lizmat specifically, because Bags take items: one could wonder if it shouldn't fail on trying to  [2,(9,3)]
19:15 lizmat - 1
19:15 woolfy I could not resist (I tried for some seconds) to post this: "The hippy Volkswagen Van is of course just for the Perl 5 folks.  The Perl 6 folks are represented by a Prius Plug-In."   http://www.toyota.nl/s3/tme-toyotaone-pro​d/toyotaone/nlnl/toyota-prius-plugin-2012​-exterior-tme-001-full_tcm-22-44038.jpg
19:17 lizmat .oO( especially now we can plug in Perl 5 with Inline::Perl5  :-)
19:18 lizmat taking this analogy further: the batteries are still small, but they will grow over time and dependence on old fuels like gasoline will be a thing of the past
19:20 raiph .tell hoelzro I've added a comment following up on your bug reports at https://questhub.io/realm/perl/​quest/52d6b4dcf6d3292b1a00007e
19:20 yoleaux 20 Sep 2014 22:33Z <grondilu> raiph: I'm not sure what this is.  It's not mine anyway.  Still, I notice this uses Parrot-specific code.  If it can be shown that the pure P6 version is as least as good, it should prevail, shouldn't it?
19:20 yoleaux raiph: I'll pass your message to hoelzro.
19:20 grondilu How exactly is Inline::Perl5 different from v5?  Is it just syntax or is it completely different?
19:21 raiph grondilu: completely different
19:21 grondilu ok
19:21 raiph Inline::Perl5 calls out to an embedded version of regular perl
19:22 grondilu ok but that's how it works.  Could Inline::Perl5 be used to implement v5?
19:22 lizmat no
19:23 lizmat Inline::Perl5 will always depend on a Perl5 lib existing
19:23 raiph v5 is a brand new compiler, written in P6
19:23 lizmat v5 will only need the grammar and associated actions in Perl 6 land
19:29 grondilu sure but wouldn't an implementation of v5 using the perlapi be a valid implementation?
19:29 * grondilu assumes Inline::Perl5 does use the perl5api
19:30 lizmat grondilu: use v5 has nothing to do with perl5 the executor
19:30 TimToady [1,(8,2)] flattens inside
19:30 ifim hi guys, I was wondering what place is the most up-to-date reference to learn Perl 6 from ground up
19:31 grondilu ifim: S02
19:31 grondilu https://github.com/perl6/specs
19:32 ifim grondilu: thanks!
19:32 * TimToady --> Carson City, hopefully &
19:32 lizmat .oO( Bonanza! )
19:32 pmurias grondilu: Inline::Perl5 could be used to implement the 'use v5;#p5 code here' construct
19:33 vendethiel ifim: try http://learnxinyminutes.com/docs/perl6/
19:33 pmurias grondilu: there is a perl5 impelementation (written in v5) called v5
19:33 vendethiel pmurias: written in p6 :p
19:34 leont joined #perl6
19:34 pmurias yes
19:37 raiph grondilu: Does your newer digest module do all the things the older one did?
19:38 eupcan277_y_h_h_ joined #perl6
19:38 grondilu raiph: I don't know.  IIRC it doesn't but I'd have to check.
19:39 grondilu but above all IIRC the interface is *very* different so it would not be compatible.
19:41 grondilu is there a simple way to check which module in the ecosystem depends on a given module?
19:41 _slade_ joined #perl6
19:41 SamuraiJack joined #perl6
19:42 raiph grondilu: simple: just move it to CPAN and use CPAN's nice new dependency display ;)
19:42 dalek specs: 82431de | (Elizabeth Mattijsen)++ | S (4 files):
19:42 dalek specs: Eradicate Seq
19:42 dalek specs: review: https://github.com/perl6/specs/commit/82431dec3b
19:43 * grondilu is sorry to say he does not like CPAN
19:43 raiph grondilu: I'm thinking the outcome re the old digest is that it will be deleted. i'm mostly wondering about cherry picking from it
19:44 grondilu raiph: do you want to be added as a contributer to my version of Digest?
19:45 Psyche^ joined #perl6
19:46 dalek roast: 2d166a0 | (David Warring david.warring@gmail.com)++ | S05-grammar/protoregex.t:
19:46 dalek roast: another fudged test for RT #120146
19:46 dalek roast: review: https://github.com/perl6/roast/commit/2d166a0f41
19:46 synopsebot Link: https://rt.perl.org/rt3//Publ​ic/Bug/Display.html?id=120146
19:47 grondilu I've added FROGGS and retupmoca lately, I guess I can add you as well
19:47 raiph grondilu: I'm just following up on https://questhub.io/realm/perl/​quest/534571eabbd0bed24e00006d
19:48 grondilu oh, ok
19:48 * grondilu doesn't quite feel like engaging in a quest right now (excess of laziness)
19:51 grondilu what is https://github.com/perl6/modules anyway??
19:51 grondilu I mean is there someone using this?
19:51 raiph grondilu: a stale repo (hasn't been updated in 5 years). fyi the old module is just one screenful of which I'd imagine just a few lines might be worth cherry picking, if any
19:52 * grondilu looks at it more closely
19:52 vendethiel raiph: no, it's just PIR code
19:52 vendethiel grondilu: ^
19:53 grondilu indeed, and not much of it.
19:53 grondilu I think you can just forget about this.
19:53 raiph OK, that repo needs to just go
19:53 raiph thanks
19:54 virtualsue joined #perl6
19:55 lizmat TimToady: nice ride through the Eldorado National Forest, hope you don't get bothered by any forest fires
20:03 [Sno] joined #perl6
20:03 dalek roast: 3747123 | (David Warring david.warring@gmail.com)++ | S05-grammar/protoregex.t:
20:03 dalek roast: typo
20:03 dalek roast: review: https://github.com/perl6/roast/commit/3747123d77
20:03 denis_boyun__ joined #perl6
20:05 bjz joined #perl6
20:06 nbrown joined #perl6

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

Perl 6 | Reference Documentation | Rakudo | Niecza | Specs