Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2014-01-24

Perl 6 | Reference Documentation | Rakudo

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

All times shown according to UTC.

Time Nick Message
00:00 jnthn timotimo: We shouldn't do it to the bytecode stream, that's what we want to mmap
00:00 jnthn timotimo: We may find there are other places (like annotations) where we can do it.
00:00 jnthn lue: But that's just a driver. First thing it does is load a module
00:00 timotimo right, annotations sounds good, too
00:01 timotimo it has lots of line numbers, aye?
00:01 jnthn aye
00:01 jnthn lue: https://github.com/jnthn/rakudo-debugger/blob/master/bin/perl6-debug.nqp is the thing I'm wanting to bring into the Rakudo repo. https://github.com/jnthn/rakudo-debugger/blob/master/lib/Debugger/UI/CommandLine.pm is the one to stay where it is.
00:01 timotimo sigh. yeah, i'll have to do that
00:02 timotimo and tomorrow i'll spend some time with a friend exploring gdb python scripts
00:02 timotimo i may end up making lots of neat things for exploring moarvm processes that way
00:02 lue I'll have to look at the debugger's code to understand how it's all connected.
00:02 timotimo it's all connected! don't you see !?!
00:03 lue Sounds like an interesting task; no guarantees I'll actually accomplish it though :P
00:04 jnthn timotimo: Object graphs usually are ;-)
00:06 timotimo jnthn: i didn't finish the var-sized-int thing yet because it involves some byte-order tricking and stuff. i should probably just prototype it in p6 and then port to C or something :)
00:06 jnthn timotimo: Yes, we need to be careful about that.
00:06 jnthn I think the bytecode and/or serialization format spec have something on that.
00:07 timotimo it has one specific endianness
00:08 timotimo hm. does the bytecode dumper work ATM anyway?
00:11 jnthn Yes.
00:12 jnthn I've been using it quite a bit recently while doing code-gen improvements.
00:12 timotimo i should really check that out, too.
00:14 timotimo jnthn: how do you feel about having ansi escapes for formatting optionally available in the bytecode dumper?
00:15 timotimo actually, that could be a perl6 program instead
00:15 jnthn timotimo: oh, I almost always redirect it to a file and read it that way.
00:15 timotimo and it could plausibly emit 24 bit colors to make all register names easily distinguishable
00:16 timotimo since they all look so similar
00:16 jnthn Feels like a lot of effort for something we probably look at rarely...
00:16 jnthn But yeah, you could hack a script up to do it.
00:17 timotimo well, since i don't work on a "what's useful" basis anyway ... ;)
00:17 jnthn :)
00:18 timotimo moarvm makes me pretty happy. maybe i'll even become a moar hacker, too
00:20 lue jnthn: "@ARGS.unshift('-MDebugger::UI::CommandLine');"    Is that how perl6-debug uses the CommandLine by default? Would -MDebugger::UI::Qt  override it?
00:21 timotimo it won't override it
00:21 timotimo er, at least i think so
00:21 timotimo how does adding multiple -M things work?
00:23 lue I get the feeling that it's some hardcoding that needs to be fixed.
00:23 timotimo you can inspect the @ARGS and only unshift that module if no other frontend has been declared
00:25 lue yeah. jnthn: if perl6-debug.nqp goes in the rakudo repo, I'm not sure how to not also include CommandLine.pm (or *some* sort of default frontend)
00:26 timotimo hm. from where you are, can you easily inspect what modules are installed?
00:27 lue timotimo: yeah, my ~/.perl6 is perfectly accessible :)
00:27 timotimo er, at run-time i mean :)
00:27 [Coke] anyone know mark fowler?
00:27 [Coke] (done with the presentation)
00:27 timotimo not me
00:27 timotimo was it good? :)
00:28 [Coke] needed moar time
00:29 lue timotimo: I'd prefer if stuff in rakudo/rakudo didn't hope you had certain things installed :/
00:29 lue s/things/Perl 6 modules/
00:29 timotimo hmm.
00:30 timotimo how about Pod::To::? it's solved similarly, but it does ship one module with rakudo itself
00:31 dalek roast: e72febe | (Pepe Schwarz)++ | S12-methods/private.t:
00:31 dalek roast: Adjusted tests for calling a method private to a role.
00:31 dalek roast: review: https://github.com/perl6/roast/commit/e72febee1a
00:31 dalek roast: eb23ef6 | (Pepe Schwarz)++ | S32-exceptions/misc.t:
00:31 dalek roast: Added tests for X::TypeCheck::Argument thrown by the Optimizer.
00:31 dalek roast: review: https://github.com/perl6/roast/commit/eb23ef6018
00:31 lue timotimo: That's probably how it would turn out. But it means either moving CommandLine.pm with the .nqp file (which jnthn said he didn't want), or creating a new, minimal interface.
00:32 lue Actually, it would have to be a minimal interface, jnthn's uses Term::ANSIColor...
00:32 timotimo right.
00:32 jnthn lue: Yes, that's how it does it right now. It is hard coded. The idea is to introduce a configuration file at some point.
00:33 timotimo jnthn: do i get the go-ahead to add 256colors and 24bit colors to Term::ANSIColor?
00:33 jnthn lue: I know that you won't get a functioning debugger without installing a UI
00:33 timotimo or would that go into Term::ANSIColor::Extended or something?
00:33 jnthn timotimo: It's tadzik's module... :)
00:33 lue jnthn: the config would probably be somewhere like ~/.perl6/<version>/etc/ (substitute for your installation's install location), right?
00:34 jnthn lue: Or just a .perl6_debug in your home directory.
00:34 stevan_ joined #perl6
00:34 timotimo oh, it is!
00:34 timotimo tadzik: ^ :)
00:36 lue jnthn: IIUC, you eventually want a config file that by default calls for the default UI, and that can be edited by the user. And the .nqp would do no hardcoded default stuff, aside from maybe creating the config file if not found.
00:50 timotimo [Coke]: did you get more time or did you have to abort early?
00:54 [Coke] sped through the last 1/3
00:54 [Coke] spent more time talking about history and community, though, so 'was fine.
00:57 timotimo fair enough
00:57 jnthn lue: Well, put the hardcoded thing in first, but yes, that's the goal.
00:57 jnthn lue: I don't think it should create the config file
00:57 timotimo for the next time i'll try to figure out (or maybe delegate to hoelzro) if you wan get pygments to output something google docs can read
00:57 [Coke] timotimo++
00:57 timotimo bedtimo now :)
00:58 jnthn lue: I do think we should go so far as catching the default UI module not being found and tell the user they need to install it. But that won't happen to Star uses
00:58 jnthn *users
00:58 lue jnthn: I was thinking it as a sort of "oh, there's no config file, I'll just recreate the default one and continue." But I can see not doing that.
01:03 Mouq joined #perl6
01:03 SubTerra joined #perl6
01:04 Alina-malina joined #perl6
01:06 jeffreykegler left #perl6
01:32 d4l3k_ joined #perl6
01:32 baest_ joined #perl6
01:32 Ayiko joined #perl6
01:43 lue joined #perl6
01:47 jnthn Time for some rest...'night
01:48 TimToady o/
01:57 xinming joined #perl6
02:07 TimToady revised http://rosettacode.org/wiki/Fractran#Perl_6 to derive first 20 primes (not the fastest way to find primes...)
02:07 psch g'night #perl6
02:07 TimToady o/
02:12 jnap joined #perl6
02:20 diakopter O_O
02:51 Mouq joined #perl6
02:52 xenoterracide joined #perl6
03:06 Alina-malina joined #perl6
03:26 skids joined #perl6
03:53 jeffreykegler joined #perl6
04:11 kaare_ joined #perl6
04:31 kaleem joined #perl6
04:39 Mouq joined #perl6
04:40 slavik joined #perl6
05:00 dalek perl6-roast-data: ccc5c4a | coke++ | / (6 files):
05:00 dalek perl6-roast-data: today (automated commit)
05:00 dalek perl6-roast-data: review: https://github.com/coke/perl6-roast-data/commit/ccc5c4a087
05:03 SamuraiJack joined #perl6
05:16 jeffreykegler left #perl6
06:11 moritz good morning #perl6
06:16 Resinator joined #perl6
06:27 Mouq joined #perl6
06:28 [Sno] joined #perl6
06:34 hoverboard joined #perl6
06:40 moritz yes, rakudo-j build works again for me (it hang some days ago)
06:52 zby_home_ joined #perl6
06:55 drin_m joined #perl6
06:59 dayangkun joined #perl6
07:21 dayangkun joined #perl6
07:37 hoverboard left #perl6
07:37 darutoko joined #perl6
07:38 Exodist joined #perl6
07:40 kaleem joined #perl6
07:59 varna joined #perl6
08:03 varna left #perl6
08:04 Resinator left #perl6
08:08 zakharyas joined #perl6
08:13 FROGGS joined #perl6
08:15 Mouq joined #perl6
08:17 dmol joined #perl6
08:22 Alula joined #perl6
08:29 dalek rakudo/nom: deb19ac | (Steve Mynott)++ | Configure.pl:
08:29 dalek rakudo/nom: add --backends=ALL option to Configure.pl
08:29 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/deb19acbe7
08:29 dalek rakudo/nom: 8ae2ded | (Tobias Leich)++ | Configure.pl:
08:29 dalek rakudo/nom: Merge branch 'stmuk-nom' into nom
08:29 dalek rakudo/nom:
08:29 dalek rakudo/nom: Conflicts:
08:29 dalek rakudo/nom: Configure.pl
08:29 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/8ae2ded4c1
08:46 pdcawley joined #perl6
08:49 SamuraiJack joined #perl6
08:56 daniel-s_ joined #perl6
09:05 cognominal joined #perl6
09:07 pdcawley joined #perl6
09:11 hoelzro morning #perl6
09:11 hoelzro jnthn: excellent post
09:11 moritz aye, jnthn++
09:12 hoelzro timotimo: do you know if Google Docs supports embedding raw HTML?
09:16 telex joined #perl6
09:16 Mouq joined #perl6
09:26 xenoterracide joined #perl6
09:28 FROGGS jnthn++ # that post even motivates me for $work, well done :o)
09:30 FROGGS can somebody judge if this will do no harm? https://github.com/perl6/nqp/pull/153
09:32 moritz iirc we use https because there are corporate firewalls that block git://
09:32 FROGGS true
09:32 xinming joined #perl6
09:32 FROGGS I think I had this issue
09:32 moritz so we should really have a --git-protocol=(git|https|ssh) option
09:33 moritz I'll comment on the pull request
09:33 FROGGS moritz++
09:34 SamuraiJack joined #perl6
09:36 masak antenoon, #perl6
09:37 Sqirrel joined #perl6
09:37 moritz \o masak
09:38 moritz masak++ # Rakudo release
09:41 masak oh, jnthn++ blugged.
09:41 FROGGS masak++ # indeed
09:44 JimmyZ all++
09:45 jnthn morning o/
09:46 jnthn masak: blogged or blug...you can't ablaut and morphology, that's just greedy :P
09:46 pecastro joined #perl6
09:47 tadzik good morning
09:48 FROGGS morning
09:50 fhelmberger joined #perl6
09:52 masak jnthn: says the guy who still writes "less specification tests" in his (otherwise excellent!) blog post. :P
09:53 FROGGS is... is that wrong?
09:53 FROGGS should it be written as "fewer..." ?
09:53 masak unless "specification tests" are a mass noun.
09:54 masak but they're not; they're discrete units.
09:54 JimmyZ jnthn: s/hare/have/, I think
09:54 * masak .oO( s/hare/bunny/ )
09:55 FROGGS hehe
09:55 timotimo o/
09:55 FROGGS hi timotimo
09:56 hoelzro hi Timo
09:58 masak hi hi timo timo
09:59 rindolf joined #perl6
10:01 jnthn JimmyZ: Thanks, fixed :)
10:02 * jnthn throws a mass noun at masak
10:03 * jnthn approves some blog comments
10:04 timotimo jnthn: i think it's supposed to be "a lot of information on hand" rather than "to hand"
10:08 jnthn timotimo: Hm, "to hand" feels much more natural to me here.
10:08 timotimo fair enough :)
10:09 jnthn timotimo: http://www.thefreedictionary.com/to+hand # to hand: ... 2. In one's possession.
10:10 jnthn "on hand" seems equally valid, mebbe it's a dialect thing.
10:10 timotimo ah, thanks
10:43 * masak throws an ablut morphologue at jnthn
10:45 nwc10 masak: yes, I noticed that one :-)
10:47 FROGGS very nice comments these are
10:47 jnthn more awesome in Febuary we should
10:48 6JTAA10LQ joined #perl6
10:49 FROGGS we shall do or do not, there is no try
10:49 jnthn That's 'cus you can write CATCH anywhere you like
10:49 * masak .oO( yo da, I know you don't like try, do I put some do in your do so you can do or do not )
10:50 timotimo FROGGS: so far only from people who are close to perl6 anyway :P
10:50 FROGGS timotimo: because the naysayers dunno what to write :o)
10:51 timotimo naysayers always get back up
10:51 timotimo they don't need a reason to naysay, they'll always come up with even the tiniest thing if they feel the need to
10:52 FROGGS did I mention that I don't really care about these ppl? (I do care about complaints if they are alright though)
10:54 timotimo and you're right to.
10:54 timotimo i have not yet reached that level of zen
10:57 sitaktif joined #perl6
10:57 FROGGS ohh, I have two little kids
10:59 timotimo :3
11:00 masak little kids are naysayers? I'm glad I learned this in time.
11:00 moritz unlyess you offer them some candy
11:00 moritz it's very rare for them to say nay to that.
11:01 timotimo well, i hope you (and all other perl6ers with little kids) succeed in growing them into kind and strong people :)
11:04 FROGGS masak: no, it is just that you either get some ease of mind, or you tend to shout at them all the time
11:08 Alula joined #perl6
11:16 masak FROGGS: I can easily imagine that learning to deal with kid dynamics helps being better at handling open-source naysayer dynamics :)
11:17 FROGGS :o)
11:21 cosimo anything dynamics really :)
11:23 masak "anything dynamics". I should put that on my business card.
11:23 masak I'm really interested in anything dynamics these days.
11:26 ribasushi woolfy1: hehe (re make a DBIx::Class for Perl 6) - we aren't really there yet wrt DBIish etc
11:26 ribasushi woolfy1: however 1) DBIC is pure-perl (with one exception which is easily work-aroundable) and 2) I am planning future jdbc bindings
11:27 ribasushi so ideally probably likely ostensibly allegedly it should wourk out of the box once we get a p5 parser up ;)
11:27 ribasushi moritz: ^^ you'll agree yes? :)
11:29 masak ribasushi++ # ideally probably likely ostensibly allegedly
11:29 woolfy1 ribasushi: I don't care much about the name of the module (well, maybe a bit for the ease of converting programs from Perl 5 to Perl 6)
11:30 masak as long as people find the module, the name is not that important.
11:30 woolfy1 So whatever you and Tim Bunce and Jens and others are going to do at the QA Hackathon (or elsewhere) that would expand the DBI-possibilities of Perl 6, it would be greatly appreciated
11:30 woolfy1 ribasushi: you are coming to Lyon, aren't you?
11:31 woolfy1 ribasushi: yay, according tohttp://act.qa-hackathon.org/qa2014/search?country=de you are coming
11:32 ribasushi some folks were misguided enough to invite me >.<
11:32 woolfy1 :-)
11:32 mavcunha joined #perl6
11:33 moritz agreed, contributions to DBIish would be greatly appreciated
11:33 moritz if I can get a list of participants (and their github IDs), I can give them a commit bit in advance
11:33 woolfy1 moritz: well, if at some point the -ish could be dropped and Perl 6 would have damned real good DBI, that would be appreciated too
11:34 ribasushi at perl-qa most of us are likely going to be busy fighting over the p5 toolchain, so I don't have high hopes
11:34 moritz woolfy1: aye, but history shows that good solutions usually develop out of -ish solutions, so I'm all for the evoluationary appraoach
11:34 ribasushi moritz: where do I get perlsix-ish? :)
11:38 denisboyun joined #perl6
11:40 masak ribasushi: https://github.com/rakudo/rakudo ;)
11:42 daniel-s_ joined #perl6
11:46 woolfy1 ribasushi: when you would make DBI stuff easier to convert, that would be a plus anyway
11:47 ribasushi what we really need is for diakopter to wrap up his grant, then everything will fall into place >.>
11:59 timotimo hoelzro: you can copypaste html with markup from another tab into google docs apparently
12:01 atroxaper joined #perl6
12:11 hoelzro mkay, that could work
12:11 hoelzro I'm not sure what all of the output that pygments supports are
12:12 timotimo it surely supports html output
12:15 fhelmberger joined #perl6
12:19 bjz_ joined #perl6
12:23 colomon Overnight failures in DB::Model::Easy and DBish
12:23 FROGGS colomon: linky?
12:23 colomon http://75.134.170.198:3001/report
12:24 FROGGS ahh, I was hoping to see the output
12:25 colomon I don't think emmentaler stores that
12:26 FROGGS sad
12:32 mavcunha joined #perl6
12:35 denis_boyun_ joined #perl6
12:36 timotimo wow, that's really a big amount of modules not working
12:46 colomon timotimo: this is actually one of the best stretches of modules *working* we've had since I started keeping track last June.
12:47 timotimo oh!
12:48 moritz mlenz@mlenz-workstation:~/p6$ panda install Nativecall
12:48 moritz Could not download module metadata: Could not find symbol '&INET' in method update at /home/mlenz/p6/rakudo/install/languages/perl6/site/lib/Panda/Ecosystem.pm:65
12:48 moritz wtf?
12:48 colomon you're running Moar
12:49 moritz I'm not
12:49 moritz panda is
12:49 colomon at least, that was the source of that bug for someone… (FROGGS?) yesterday
12:49 moritz I said 'perl6-p bootstrap.pl', and IMHO that should mean I run perl6-p
12:50 hoelzro is dir($dir) supposed to yield an IO::Path for '..'?
12:50 moritz ah, but maybe panda simply invokes perl6
12:50 moritz hoelzro: no
12:51 hoelzro ok, that's what I thought
12:51 * hoelzro found a bug
12:51 colomon timotimo: I think it would be really great to have a module cleanup "hackathon" sometime soon.  I'd be a third of the failures are easy fixes, and another third are ancient abandonware that should probably be removed from the ecosystem.
12:51 moritz colomon: +1
12:52 moritz ok, I've alias perl6 to perl6-p
12:52 moritz now panda won't install NativeCall for me :(
12:52 colomon that said, I have very little time, and what time I have right now might be better spent getting the smoke tester running on JVM too.
12:53 moritz t/03-simple-returns.t .. /usr/bin/ld: cannot find -lgmp
12:53 moritz collect2: error: ld returned 1 exit status
12:53 moritz do I need libgmp to run zavolaj? or should that come with parrot? or not needed at all?
12:53 * moritz confused
12:53 timotimo hoelzro: where did you find that bug? :)
12:53 hoelzro ./htmlify.pl in doc
12:54 hoelzro recursive-dir goes up to parent dirs
12:54 jnthn moritz: No, you shouldn't need it; it'd sounds like -lgmp got into the flags coming from Parrot somehow
12:54 moritz p: say dir('/usr/local')
12:54 camelia rakudo-parrot 8ae2de: OUTPUT«IO::Path</usr/local/src> IO::Path</usr/local/etc> IO::Path</usr/local/bin> IO::Path</usr/local/man> IO::Path</usr/local/include> IO::Path</usr/local/sbin> IO::Path</usr/local/share> IO::Path</usr/local/games> IO::Path</usr/local/lib>␤»
12:54 jnthn moritz: Though Rakudo woulda been build with the same flags...
12:54 moritz jnthn: rakudo built fine :-)
12:55 moritz when I install libgmp-dev, everything works fine
12:55 moritz maybe it was installed when I last configured parrot
12:55 jnthn maybe, yeah
12:55 hoelzro and now I seem to have caught it in an infinite loop
12:55 hoelzro although it might returning '.' as well
12:56 moritz hoelzro: htmlify worked fine for me on perl6-p last I tried (not too long ago; like, this year)
12:56 hoelzro it's on perl6-m
12:56 moritz ah
12:56 mtk joined #perl6
12:57 hoelzro I know what I'm doing tonight!
12:58 * jnthn wonders if it's memory usage on perl6-m will be any better
12:58 jnthn *its
12:58 hoelzro jnthn: that's what I'm trying to find it =)
12:59 hoelzro *find out
12:59 timotimo i would like that.
12:59 hoelzro so the dir() problem only occurs if the arg to dir() is *not* the cwd
12:59 hoelzro (afaict)
12:59 moritz m: say dir('/usr/local/')
12:59 camelia rakudo-moar 8ae2de: OUTPUT«IO::Path<src> IO::Path<etc> IO::Path<bin> IO::Path<man> IO::Path<include> IO::Path<sbin> IO::Path<share> IO::Path<games> IO::Path<lib>␤»
13:00 moritz m: say dir('/usr/local/')>>.absolute.join: ', '
13:00 camelia rakudo-moar 8ae2de: OUTPUT«/home/p6eval/src, /home/p6eval/etc, /home/p6eval/bin, /home/p6eval/man, /home/p6eval/include, /home/p6eval/sbin, /home/p6eval/share, /home/p6eval/games, /home/p6eval/lib␤»
13:00 * grondilu ecountered a weird error with r-m while trying to test his Modular module:
13:00 grondilu ===SORRY!===
13:00 grondilu Object conflict detected during deserialization.
13:00 grondilu (Probable attempt to load two modules that cannot be loaded together).
13:00 timotimo grondilu: that's also what keeps froggs from testing his v5 on moarvm
13:00 moritz r: say dir('/usr/local/')>>.absolute.join: ', '
13:00 camelia rakudo-moar 8ae2de: OUTPUT«/home/p6eval/src, /home/p6eval/etc, /home/p6eval/bin, /home/p6eval/man, /home/p6eval/include, /home/p6eval/sbin, /home/p6eval/share, /home/p6eval/games, /home/p6eval/lib␤»
13:00 camelia ..rakudo-parrot 8ae2de, rakudo-jvm 8ae2de: OUTPUT«/usr/local/src, /usr/local/etc, /usr/local/bin, /usr/local/man, /usr/local/include, /usr/local/sbin, /usr/local/share, /usr/local/games, /usr/local/lib␤»
13:01 moritz looks rather broken on r-m
13:01 hoelzro that's how it looks to me
13:03 timotimo lots of opportunities to get into moar development :)
13:04 hoelzro indeed
13:04 JimmyZ :)
13:04 hoelzro this is a good bug
13:04 hoelzro dir('lib/X') yields an IO::Path.new('lib/X')!
13:04 hoelzro so checking for '.' didn't work
13:06 hoelzro alright, got it working
13:06 hoelzro 9% memory on my VM (2.5 GB memory)
13:06 hoelzro that's a big win, imo
13:07 hoelzro er, that was just to get the file list
13:07 hoelzro too good to be true!
13:08 moritz the whole thing took about 12min on my machine
13:08 moritz hoelzro: so I was surprised it was *so* quick o your machine ;-)
13:08 hoelzro =)
13:08 hoelzro 1/6 done
13:08 hoelzro 12%
13:09 Mouq joined #perl6
13:10 hoelzro 1/2 done, 18% memory
13:14 hoelzro 7 minutes, and my machine didn't run out of memory!
13:14 hoelzro \o/
13:16 jnthn hoelzro: How does that compare with r-p on your machine?
13:17 jnthn hoelzro: Or DNF?
13:17 hoelzro jnthn: r-p runs out of memory
13:17 moritz \o/
13:17 hoelzro this is huge everyone
13:17 hoelzro a big thanks and congrats!
13:17 hoelzro I'm testing right now to see how much memory is used by program completion
13:21 sjn 9% of 2.5GB?
13:21 sjn nice!
13:21 hoelzro sjn: just for the first step =)
13:21 sjn well, ok
13:21 hoelzro I'm getting a more accurate reading now
13:21 hoelzro should be ~ 30%
13:21 sjn hm.
13:22 moritz m: say 0.09 * 2.5 * 1024
13:22 camelia rakudo-moar 8ae2de: OUTPUT«230.4␤»
13:23 hoelzro 32.4%
13:23 hoelzro not bad.
13:24 jnthn m: say 0.324 * 2.5 * 1024
13:24 camelia rakudo-moar 8ae2de: OUTPUT«829.44␤»
13:24 sjn less than 1gb is definitely a plus :)
13:24 xenoterracide joined #perl6
13:25 jnthn At some point, we can lose an extra sizeof(void *) off every object and STable in Moar
13:28 * masak .oO( thus cleaning out the STables )
13:29 arnsholt Eight bytes an object on 64-bit, sounds like a pretty good win
13:31 arnsholt What's needed to remove that pointer?
13:31 arnsholt
13:32 jnthn arnsholt: GC hackery
13:32 FROGGS yay
13:32 FROGGS -.-
13:32 jnthn arnsholt: Essentially, realizing the forwarder pointer can overwrite the first 8 bytes of the old collectable's body
13:33 FROGGS jnthn: then better you do that first *g*:
13:33 FROGGS <grondilu> (Probable attempt to load two modules that cannot be loaded together).
13:33 jnthn ah, yeah...
13:34 jnthn I need to write course material for $dayjob right now, but I can look at that one tonight.
13:34 arnsholt jnthn: Oh, that's neat =D
13:34 jnthn Was JSON::Tiny affected by it also?
13:34 FROGGS ohh, that sounds super sweet!
13:34 FROGGS jnthn: yes
13:34 arnsholt The old object is garbage, so might as well use it for something useful
13:34 jnthn arnsholt: Yeah, but it makes things harder to debug also, which is why I didn't do it yet. :)
13:34 arnsholt Heh. Makes sense
13:34 jnthn arnsholt: Or rather, why I didn't do it that way in the first place.
13:35 jnthn Now we have nwc10++ the torturer, though...and a bunch of issues found/fixed... )
13:36 sctt joined #perl6
13:36 FROGGS yeah, otherwise "we" would have spent half a year to locate these errors
13:36 moritz would it be feasible/maintable to do it as a configure switch?
13:36 jnthn aye
13:36 jnthn moritz: Very possibly.
13:36 moritz so that if you want to debug things, you can still get the easier debuggable path
13:37 jnthn moritz: I think nwc10 may have figured out exactly how to do it, while I only thought so far as "I know it'll be possible" :)
13:37 arnsholt Yeah, I was about to ask about that too, if it'd be possible to have an #ifdef for it
13:37 jnthn My gut feeling is yes, it's possible, 'cus we don't meddle with ->forwarder in many places at all.
13:39 ajr joined #perl6
13:42 FROGGS jnthn: btw, do you have a suggestion what I might do next?
13:42 kaleem joined #perl6
13:42 FROGGS somehow I am unable to get labels for nqp-j in place...
13:43 sctt_ joined #perl6
13:43 takesako joined #perl6
13:43 bibifuc_ joined #perl6
13:43 jnthn FROGGS: Fix dir? ;)
13:43 FROGGS maybe I should care about getting the S11 stuff in place instead?
13:44 FROGGS dir... ohh, I thought hoelzro is looking into that
13:44 LordVorp joined #perl6
13:44 jnthn Oh, ok :)
13:44 jnthn Yes, S11 stuff is also a worthy task.
13:44 moritz I think hoelzro looked into a workaround to use in htmlify
13:44 FROGGS hoelzro: are you?
13:44 FROGGS ahh
13:44 jnthn As is further triage of r-m spectest fails.
13:44 jdv79 joined #perl6
13:45 FROGGS yeah
13:45 yeltzooo joined #perl6
13:45 mattp_ joined #perl6
13:45 renormalist joined #perl6
13:45 jeffreykegler joined #perl6
13:46 revdiablo joined #perl6
13:46 jtpalmer joined #perl6
13:47 moritz http://irclog.perlgeek.de/perl6/2014-01-22#i_8157190 has a bug partially triaged
13:47 PZt joined #perl6
13:52 ggoebel1112 joined #perl6
13:55 robins joined #perl6
13:58 k1lldash9 joined #perl6
13:59 k1lldash9 perl6: say 3;
13:59 k1lldash9 'perl6: say 3;'
13:59 k1lldash9 hrm why is evalbot not working, am I doing something wrong?
13:59 FROGGS p: say 3
14:00 k1lldash9 p: say 3
14:00 FROGGS r: say 3
14:00 k1lldash9 hrm
14:00 k1lldash9 r: say 3
14:00 FROGGS ?
14:00 k1lldash9 no response
14:00 FROGGS camelia is not here
14:00 k1lldash9 ohh
14:00 moritz should I change that?
14:01 jnthn moritz: If you can figure out why it's down... :)
14:01 jnthn Seems it vanished with a ping timeout some minutes ago...
14:01 moritz ooh, I thought this was #moarvm :(
14:02 jnthn We have camelia in #moarvm already...well, we did ;)
14:02 camelia joined #perl6
14:02 moritz m: say 'hi'
14:03 hoelzro FROGGS: yeah, I found a workaround
14:03 hoelzro FROGGS: and I am looking at it
14:03 hoelzro but not until tonight
14:03 masak perl6: say 3;
14:03 camelia rakudo-moar 8ae2de: OUTPUT«hi␤»
14:03 camelia rakudo-parrot 8ae2de, rakudo-jvm 8ae2de, rakudo-moar 8ae2de, niecza v24-109-g48a8de3: OUTPUT«3␤»
14:04 * masak gave a quick introduction to Perl 6 yesterday at the end of his Perl course
14:04 moritz masak++
14:04 masak it's always nice to talk about Perl 6 -- especially on company time ;)
14:07 colomon masak+
14:07 colomon masak++
14:07 masak
14:08 jnthn .u  ㋡
14:08 yoleaux U+32E1 CIRCLED KATAKANA TU [So] (㋡)
14:08 jnthn haha
14:09 masak \㋛/
14:09 k1lldash9 perl6: say 3;
14:09 camelia rakudo-parrot 8ae2de, rakudo-jvm 8ae2de, rakudo-moar 8ae2de, niecza v24-109-g48a8de3: OUTPUT«3␤»
14:09 k1lldash9 you rock masak
14:10 masak r: say "you rock too, k1lldash9" # :)
14:10 camelia rakudo-parrot 8ae2de, rakudo-jvm 8ae2de, rakudo-moar 8ae2de: OUTPUT«you rock too, k1lldash9␤»
14:10 masak rm: say "I hear you have really fast startup these days. you rock, too!"
14:10 camelia rakudo-moar 8ae2de: OUTPUT«I hear you have really fast startup these days. you rock, too!␤»
14:10 masak yes. wow.
14:10 masak consider me a Rakudo-Moar hacker from now on.
14:11 moritz 0.25s on my workstation
14:11 masak yay
14:11 jercos really brings moar to the table.
14:11 moritz compared to 0.6s for r-p
14:12 moritz and 3.9s for p-j
14:12 kbaker_ joined #perl6
14:13 atroxape_ joined #perl6
14:14 jercos On the other hand, my dealings with prime numbers pretty much have required JVM use, given parrot consumes resources like crazy when trying to work with moderately large lists of numbers reduced from larger lists, and JVM doesn't have that problem at all.
14:14 jercos Gotta test this code on moar I guess...
14:14 * jercos scampers off to git
14:14 hoelzro FROGGS: are you looking at dir? I think this is something I could actually fix =)
14:17 FROGGS hoelzro: go ahead
14:17 hoelzro excellent
14:17 FROGGS I've not started yet
14:19 hoelzro cool
14:23 FROGGS r: sub circumfix:<<` `>>(*@args) { @args.join('-') }; say `3, 4, "f"`
14:23 camelia rakudo-moar 8ae2de: OUTPUT«===SORRY!===␤Cannot find method 'flat'␤»
14:23 camelia ..rakudo-parrot 8ae2de, rakudo-jvm 8ae2de: OUTPUT«3-4-f␤»
14:25 FROGGS m: sub circumfix:<<` `>>(*@args) { @args }; say `3, 4, "f"`
14:25 camelia rakudo-moar 8ae2de: OUTPUT«===SORRY!===␤Cannot find method 'flat'␤»
14:27 FROGGS Perl6::Optimizer line 673
14:28 xenoterracide joined #perl6
14:28 colomon jnthn++ did work tracking that down yesterday, I think?
14:29 FROGGS does not seem like, at least it is not fixed in nom
14:31 colomon tracking it down, not fixing it.  :)
14:31 FROGGS so I might read backlog? :o)
14:31 colomon yes.
14:31 colomon for clues.  ;)
14:31 jnthn I don't think I made any progress on that one
14:32 jnthn It was a whateverstar bug that I fixed last night.
14:32 moritz I concluded that it puts a Mu into the QAST node somwhere
14:32 moritz where? no idea
14:32 jnthn If I was gonna look into that one, I'd start by doing --target=ast
14:32 jnthn And maybe --target=parse
14:32 moritz maybe add some assertions to QAST::Node.new() and .push
14:32 jnthn My gut instinct is that it's a mis-parse at the heart of it.
14:32 jnthn Which then goes on to produce bogus AST
14:33 colomon FROGGS: http://irclog.perlgeek.de/perl6/2014-01-22#i_8157190 # starts here
14:33 timotimo jnthn: how do we distinguish an object's beginning from a forwarder written over it? by checking if it's in the new nursery already-allocated range?
14:33 FROGGS there is a QAST::Want that gets a null
14:35 airdisa joined #perl6
14:37 jnthn timotimo: No, can use a flag bit in the header
14:38 timotimo fair enough
14:38 timotimo that does sound pretty good
14:38 timotimo but our objects are rather big anyway
14:41 FROGGS - QAST::Op(call &say) say `3, 4, \"f\"`
14:41 FROGGS -
14:41 FROGGS there it is
14:42 jnthn That looks wrong, yes.
14:42 FROGGS :o)
14:43 d^_^b joined #perl6
14:43 d^_^b joined #perl6
14:44 dalek nqp: 8adf37d | (Tobias Leich)++ | src/QAST/Node.nqp:
14:44 dalek nqp: show "(NQPMu)" in ast dump because this tends to be bad
14:44 dalek nqp: review: https://github.com/perl6/nqp/commit/8adf37d188
14:45 jercos well, despite moar's obviously faster startup time (indicating that it had in fact, determined that 3 is a factor of the target in half a second >.>) it took an extra 40 seconds to factor the target 4-almost prime :|
14:45 jercos which appears to be mostly due to the fact that the JVM parallelized some stuff. not really sure what, but user time was higher than real >.>
14:45 FROGGS probably the gc
14:46 jercos in any case, given parrot's completely useless on the same code, it's nice to have another option now :D
14:46 FROGGS jercos: but if you would collect tiny snippets that are slow, we could use these for testing
14:46 jnthn jercos: Got a link to the code, ooc?
14:47 jercos Eh. It's not pretty, and has been my learn-to-maybe-perl6 project... I'll toss it in a gist.
14:49 jercos https://gist.github.com/jercos/8598678
14:49 jercos test was calling factor(36196657233).say
14:50 jercos that particular number being chosen for having a small number of factors over a (relastively) wide range
14:50 jercos (3 11 5927 185063)
14:50 jercos Obviously the idea the code represents is a little rediculous, and there's a lot there that is... well, completely useless due to a realization made after writing it :p
14:51 skids joined #perl6
14:51 FROGGS jercos: no need to excuse :o)
14:56 timotimo it's so refreshing to see perl6 code highlighted on gist
14:57 colomon wait, what?
14:57 Mouq joined #perl6
14:57 jercos It helpfully automatically selected perl6 highlights when I told it the filename was Bigprime.pm6 too :)
14:58 ggoebel1112 joined #perl6
14:58 timotimo yup
14:58 btyler joined #perl6
14:59 colomon \o/
14:59 colomon what's it using to parse for highlights?
14:59 FROGGS it uses hoelzro :o)
14:59 rindolf joined #perl6
14:59 colomon he must work quickly!
15:01 colomon it looks like the perl highlight is not smart enough to change to p6 if "use v6" starts the file?
15:02 * colomon is looking at ABC's code, which uses .pm and seems to be getting very incorrect p5 highlighting.
15:02 timotimo mhh :(
15:02 hoelzro colomon: there's a patch for that upstream
15:02 colomon hoelzro++
15:02 hoelzro it'll only take another year to get to GH =/
15:02 colomon hoelzro: how are you doing this?
15:02 hoelzro well
15:02 hoelzro first you update pygments
15:02 btyler joined #perl6
15:02 hoelzro and the maintainer quickly and happily picks up your changes
15:03 hoelzro then you send a PR to GitHub's pygments.rb module
15:03 hoelzro put it in the oven
15:03 hoelzro and set the timer to 4 months
15:03 hoelzro then, after that, you do the same for the linguist module
15:03 hoelzro (can you tell I'm a little bitter about the whole thing? =P )
15:04 hoelzro I'm just glad we have *some* support on GH =)
15:04 colomon how accurate is your code in pygments?
15:04 hoelzro accurate as far as P5 vs P6?
15:04 colomon as far as understanding p6
15:05 * colomon has been living with incorrect p6 syntax highlighting in his editor for AGES
15:05 hoelzro I say with pride that it's pretty good =)
15:05 hoelzro it highlights Grammar.pm perfectly
15:05 hoelzro the first bug was found by FROGGS last week
15:05 hoelzro pygments is a great library
15:06 hoelzro very well designed
15:06 FROGGS and you did a greta job with it :o)
15:06 FROGGS great
15:06 hoelzro thanks!
15:06 * colomon wonders if this will bring hoelzro++'s changes to his text editor: https://github.com/acangiano/pygments-textmate-bundle
15:06 FROGGS nqp-m: my role bar[$name] { method ::($name)() { 42 } }; my class Foo { }; my $Foo := Foo.new; $Foo := $Foo.HOW.mixin($Foo, bar.HOW.curry(bar, 'baz')); say($Foo.baz)
15:06 camelia nqp-moarvm: OUTPUT«42␤»
15:07 hoelzro colomon: you may have to install pygments from Hg
15:07 hoelzro give it a shot!
15:07 colomon to get the very latest?
15:07 hoelzro mhmm
15:07 jnap joined #perl6
15:07 * hoelzro checks releases
15:08 hoelzro nope, not formally released yet =/
15:08 hoelzro they do yearly releases, though
15:08 hoelzro and that time is approaching (2 weeks from now)
15:09 moritz \o/
15:09 tadzik I think modules need a way to specify things that are backend-dependent
15:09 tadzik otherwise, MIME::Base64 has trouble building
15:10 FROGGS tadzik: you are thinking of pragma if?
15:10 tadzik FROGGS: I dunno yet
15:11 timotimo i agree that something like that is needed
15:12 FROGGS nqp: my $a; my $b; $a := $b := 1; say($a); say($b)
15:12 camelia nqp-moarvm, nqp-jvm, nqp-parrot: OUTPUT«1␤1␤»
15:12 FROGGS but how? you can't compile source code to ast if you hit pir on the jvm for example
15:12 FROGGS but require-ing a module is rather slow
15:12 timotimo agreed. it'd probably have to be something at the compunit level then?
15:13 timotimo use Foo::Bar::$*VM<name>
15:13 timotimo that'd still be at compile time
15:13 FROGGS $?VM rather
15:13 timotimo also $?VM
15:13 timotimo yes
15:13 FROGGS +1
15:13 * colomon is installing hg
15:16 colomon tadzik: I was thinking MIME::Base64 should be pure p6 and/or NativeCall?
15:17 tadzik colomon: I dunno. In the perfect world, maybe :)
15:17 FROGGS it could easily delegate to jvm internals
15:17 FROGGS or to a moarvm op
15:18 * colomon is a little frustrated that Pygments "Download and installation" page doesn't include any actual information about building the code.
15:20 colomon sudo python setup.py install # it turns out
15:22 FROGGS nqp-m: my role bar[$name,$sub] { method ::($name)($/) { make '&' ~ $sub } }; my class Foo { }; my $Foo := Foo.new; $Foo := $Foo.HOW.mixin($Foo, bar.HOW.curry(bar, 'circumfix', 'circumfix:<<` `>>')); my $a := NQPMatch.new; say($Foo.circumfix($a))
15:22 camelia nqp-moarvm: OUTPUT«&circumfix:<<` `>>␤»
15:22 FROGGS why does that work here?
15:23 FROGGS ohh
15:24 dalek panda: 640d849 | tadzik++ | t/fakestate:
15:24 dalek panda: Get rid of warnings in tests, timotimo++
15:24 dalek panda: review: https://github.com/tadzik/panda/commit/640d849503
15:24 FROGGS https://gist.github.com/FROGGS/35233e86e1648ff732d9
15:24 FROGGS jnthn: in case you are interested ----^
15:27 xenoterracide joined #perl6
15:28 timotimo i'm not sure if i've done this wrong or not
15:28 timotimo i'm still getting missing or wrong version of dependency $foobar when launching moar from somewhere other than nqp/ (or rakudo/)
15:28 * timotimo nukes install/ again
15:29 FROGGS only circumfix is wrong, infix do work
15:29 diakopter what's in your path
15:34 timotimo diakopter: a tree has fallen onto my path!
15:34 diakopter STOMP
15:36 timotimo now it works again
15:36 timotimo maybe i didn't nuke install/ on this system yet
15:41 awraa joined #perl6
15:42 robinsmidsrod joined #perl6
15:47 FROGGS r: sub term:<foo> { 42 }; say foo # same bug
15:47 camelia rakudo-moar 8ae2de: OUTPUT«===SORRY!===␤Cannot find method 'flat'␤»
15:47 camelia ..rakudo-parrot 8ae2de, rakudo-jvm 8ae2de: OUTPUT«42␤»
15:48 jnthn Ah, so it's soemthing about action methods?
15:48 jnthn As in, the mixed in action methods?
15:48 FROGGS yeah
15:48 FROGGS exactly
15:49 FROGGS but my example (above) does pretty much the same but works
15:49 jnthn r: BEGIN say $*ACTIONS.HOW.name($*ACTIONS); sub term:<foo> { 42 }; BEGIN say $*ACTIONS.HOW.name($*ACTIONS);
15:49 camelia rakudo-parrot 8ae2de, rakudo-jvm 8ae2de, rakudo-moar 8ae2de: OUTPUT«Perl6::Actions␤Perl6::Actions+{TermAction}␤»
15:50 jnthn r: sub term:<foo> { 42 }; BEGIN say $*ACTIONS.HOW.method_table($*ACTIONS).keys
15:50 camelia rakudo-parrot 8ae2de, rakudo-jvm 8ae2de, rakudo-moar 8ae2de: OUTPUT«term:sym<foo>␤»
15:50 FROGGS hmmm
15:52 jnthn r: sub term:<foo> { 42 }; BEGIN $*ACTIONS.HOW.trace-on($*ACTIONS); foo
15:52 camelia rakudo-moar 8ae2de: OUTPUT«Calling term:sym<statement_prefix>␤Calling termish␤Calling EXPR␤Calling statement␤Calling termish␤Calling EXPR␤Calling statement␤»
15:52 camelia ..rakudo-parrot 8ae2de, rakudo-jvm 8ae2de: OUTPUT«Calling term:sym<statement_prefix>␤Calling termish␤Calling EXPR␤Calling statement␤Calling term:sym<foo>␤Calling termish␤Calling EXPR␤Calling statement␤»
15:53 jnthn Well, there's a difference...
15:54 jnthn r: sub term:<foo> { 42 }; BEGIN nqp::getlexdyn('$/').CURSOR.HOW.trace-on($*ACTIONS); foo
15:54 camelia rakudo-parrot 8ae2de, rakudo-jvm 8ae2de, rakudo-moar 8ae2de: ( no output )
15:54 hoelzro hmm
15:55 hoelzro I wonder how much memory building the docs would take on a 32-bit Moar?
15:55 jnthn Anyway, seems the issue is issue is that it doesn't call the action method.
15:55 jnthn For some reason
15:57 jnthn FROGGS: Found it for ya
15:58 jnthn FROGGS: Compare compilation of rxtype pass on JVM vs Moar
15:58 jnthn FROGGS: Note the "elsif +@($node) == 1 {" branch is missing for Moar.
15:58 jnthn FROGGS: Which is where it'd be getting the name of the action method when it's late-bound
15:58 FROGGS ohh!
15:58 jnthn hoelzro: Well, every pointer would be smaller...
15:59 hoelzro exactly
15:59 jnthn ...and I suppose we have plenty of those :)
16:00 hoelzro I'll have to try that later tonight
16:08 * FROGGS compiles something
16:13 FROGGS I love it:
16:13 FROGGS Using /home/froggs/dev/nqp/install/bin/nqp-p (version 2014.01-1-g8adf37d / Parrot 5.9.0).
16:13 FROGGS Using /home/froggs/dev/nqp/install/bin/nqp-j (version 2014.01-1-g8adf37d / OpenJDK 1.7.0_51).
16:13 FROGGS Using /home/froggs/dev/nqp/install/bin/nqp-m (version 2014.01-1-g8adf37d / MoarVM 2014.01-1-g6e85b9c).
16:14 * FROGGS wonders when a..z is full
16:16 FROGGS looks like I did it wrong: Error while compiling op if (source text: "$is_term {\n            my role Term[$meth_name, $op] {\n                token ::($meth_name) { $<sy..."): cannot stringify this
16:17 regreg joined #perl6
16:23 MarkFowler joined #perl6
16:23 treehug88 joined #perl6
16:24 MarkFowler Morning
16:24 FROGGS hi MarkFowler
16:24 MarkFowler [Coke]: what was the slide URL?
16:24 MarkFowler (that was presented at Albany.pm last night)
16:28 MarkFowler ah, got it
16:41 FROGGS when I do: my $name := $*QASTCOMPILER.as_mast($node[0], :want($MVM_reg_str))
16:42 FROGGS I thought $name would be a MAST::SVal
16:42 jnthn No
16:42 FROGGS it is a register
16:42 jnthn as_mast always returns a MAST::InstructionList
16:42 FROGGS ohh, right
16:42 stevan_ joined #perl6
16:42 jnthn I don't think you can rely on it being constant here either.
16:43 FROGGS hmmm
16:43 [Coke] MarkFowler: hio.
16:43 jnthn If we actually knew it statically we'd just have put it in .name
16:43 [Coke] I'm probably going to PDF that and post the PDF to the list. (and archive it on the albany.pm site, for lack of a better location)
16:44 MarkFowler github's presentation sharing dooh dah
16:44 MarkFowler https://speakerdeck.com/
16:45 Mouq joined #perl6
16:45 PerlJam [Coke]: How'd your presentation go last night?
16:47 FROGGS I think I understood (in case it compiles)
16:47 MarkFowler PerlJam: it was very well received
16:47 masak [Coke]++
16:47 PerlJam [Coke]++ indeed
16:48 timotimo i have a small python gdb plugin for printing the contents of a MVMString
16:48 FROGGS cool!
16:49 FROGGS jnthn: Incompatible MROs in P6opaque rebless :o(
16:49 FROGGS ohh wait
16:49 FROGGS I messed with the P6::Grammar
16:49 FROGGS man, I am stupid
16:50 * japhb reaches realtime in epic backlogging ... and now for replies ...
16:50 FROGGS hi japhb # testing
16:51 FROGGS aahh! Cannot invoke this object (REPR: P6str, cs = 0)
16:54 FROGGS jnthn: I really thought this would work: https://gist.github.com/FROGGS/35233e86e1648ff732d9
16:55 japhb re: http://irclog.perlgeek.de/moarvm/2014-01-22#i_8157565 : Those variable-length int encodings work decently if most ints are small, but you can't be sure exactly how big they will be (i.e. by having the schema at hand).  For negative numbers, they tend to do trickery like zig-zag encoding of ints into uints, or moving the sign bit from the msb to the lsb and inverting the other bits if the sign bit is set.
16:55 japhb (So as to get zero bits at the top)
16:55 japhb FROGGS: ok "testing"
16:55 FROGGS :o)
16:56 jnthn FROGGS: nqp::push(@flags, $Arg::str); # indicates a constant str
16:56 FROGGS ohh
16:57 ajr joined #perl6
16:59 japhb yoleaux: tell raydiak re: http://irclog.perlgeek.de/perl6/2014-01-21#i_8152416 , yes I've done a bit of work on the vector lib, just getting vector-vector and vector-scalar operations in place.  The code for similar operators is generated from template to avoid repetition and obscure copy-pasta errors affecting only one variation of one op somewhere.
16:59 japhb yoleaux: ping
16:59 yoleaux japhb: pong! (0 queued, 0 running)
17:00 FROGGS without its name I think
17:00 japhb Huh, doesn't yoleaux do messages?
17:00 japhb OK, trying
17:00 FROGGS dalek needed its name
17:00 japhb tell raydiak re: http://irclog.perlgeek.de/perl6/2014-01-21#i_8152416 , yes I've done a bit of work on the vector lib, just getting vector-vector and vector-scalar operations in place.  The code for similar operators is generated from template to avoid repetition and obscure copy-pasta errors affecting only one variation of one op somewhere.
17:00 japhb Nah, no love
17:00 jnthn .tell ?
17:00 yoleaux jnthn: I don't know what you want me to say to ?.
17:00 japhb .tell raydiak re: http://irclog.perlgeek.de/perl6/2014-01-21#i_8152416 , yes I've done a bit of work on the vector lib, just getting vector-vector and vector-scalar operations in place.  The code for similar operators is generated from template to avoid repetition and obscure copy-pasta errors affecting only one variation of one op somewhere.
17:00 yoleaux japhb: I'll pass your message to raydiak.
17:00 FROGGS .right
17:01 FROGGS .yaw siht saw tI
17:01 japhb There we go.  That kind of annoyance already doesn't happen in Net::IRC, btw, I fixed that one early on.
17:02 ajr_ joined #perl6
17:03 japhb .tell raydiak The code needs some massaging to make it sane to work on together, but if you're up for it, I can find some tuits to throw at that.
17:03 yoleaux japhb: I'll pass your message to raydiak.
17:04 FROGGS jnthn: am I supposed to pass 0 as a flag? $Arg::obj does not do either
17:04 TimToady r: .say for (^5).combinations(3);
17:04 camelia rakudo-parrot 8ae2de, rakudo-jvm 8ae2de: OUTPUT«No such method 'combinations' for invocant of type 'Range'␤  in block  at /tmp/tmpfile:1␤␤»
17:04 camelia ..rakudo-moar 8ae2de: OUTPUT«Cannot find method 'combinations'␤  in block  at /tmp/tmpfile:1␤␤»
17:04 ajr_ joined #perl6
17:05 japhb timotimo: re:  http://irclog.perlgeek.de/perl6/2014-01-21#i_8152652 , I'm fine with perl6-bench displaying more information about progress and expected completion time, as long as care is taken that such features don't interfere with benchmark accuracy (by stealing resources for the UI).
17:05 jnthn FROGGS: oh wait, that's not the problem at all...sorry
17:05 FROGGS hehe, okay
17:05 japhb timotimo: With the current design of perl6-bench, it might even be reasonable to make guesses about expected completion for quick-starting backends.  For JVM it may be harder.
17:06 FROGGS jnthn: I do a bit food hunting now, so we can maybe talk about that later
17:06 FROGGS I guess that is better for you too :o)
17:07 FROGGS[mobile] joined #perl6
17:07 jnthn FROGGS[mobile]: Yeah, it looks right...
17:07 jnthn FROGGS[mobile]: yes, sure
17:07 japhb skids: re: http://irclog.perlgeek.de/perl6/2014-01-21#i_8152921 -- link to "Sum" module?
17:08 djanatyn joined #perl6
17:08 japhb skids: Also, if you'd like to add tests to perl6-bench, I welcome PRs.  :-)
17:08 ajr_ joined #perl6
17:09 japhb r: my @primes := 2, 3, 5, -> $p { ($p+2, $p+4 ... &is-prime)[*-1] } ... *; sub is-prime($n) { $n %% none @primes ...^  * > sqrt $n }; .say for @primes[^5];  # Checking if this got fixed on r-j
17:09 camelia rakudo-jvm 8ae2de: OUTPUT«Cannot do aggregate operation on a type object␤  in method reify at gen/jvm/CORE.setting:7239␤  in method gimme at gen/jvm/CORE.setting:7649␤  in method exists_pos at gen/jvm/CORE.setting:7637␤  in method at_pos at gen/jvm/CORE.setting:7611␤  in m…»
17:09 camelia ..rakudo-parrot 8ae2de, rakudo-moar 8ae2de: OUTPUT«2␤3␤5␤7␤11␤»
17:09 japhb Dang, was hoping it got fixed with the gather-take stuff.
17:11 jnthn japhb: Does that one work on other backends?
17:15 diakopter ?
17:15 djanatyn joined #perl6
17:18 jnthn oh, I misread :)
17:18 jnthn duh :)
17:21 raydiak joined #perl6
17:22 ajr_ joined #perl6
17:23 TimToady (*-1).nth: $p+2, $p+4 ... &is-prime  # wonder if we should make this work
17:24 timotimo ooooh
17:24 TimToady it's a bit like ruby's throwing .times into the integer class, but I keep wanting a way to pull the subscript out front
17:25 Tene joined #perl6
17:25 TimToady and nth *-1, $p+2, $p+4 ... &is-prime doesn't parse right
17:26 ajr_ joined #perl6
17:27 timotimo actually, i wouldn't mind if it would have to be (*-1).&nth: ...
17:28 TimToady std: sub nth(\ix, *@list) { @list[ix] }; nth(*-1): 1,2,3...10
17:28 camelia std 09dda5b: OUTPUT«ok 00:01 132m␤»
17:28 TimToady n: sub nth(\ix, *@list) { @list[ix] }; say nth(*-1): 1,2,3...10
17:28 camelia niecza v24-109-g48a8de3: OUTPUT«[31m===[0mSORRY![31m===[0mâ�¤â�¤Interaction between semiargs and args is not understood at /tmp/puCu_6o1Wh line 1 (EOF):â�¤------> [32m { @list[ix] }; say nth(*-1): 1,2,3...10[33mâ��[31m<EOL>[0mâ�¤â�¤Unhandled exception: Check failedâ�¤â�¤  at…»
17:28 TimToady heh
17:28 timotimo o_O
17:28 timotimo oh well
17:29 TimToady std allows a switch from args inside parens to outside parens
17:29 TimToady (via colon)
17:29 djanatyn joined #perl6
17:30 TimToady std: map({.chr}): 1,2,3...10
17:30 camelia std 09dda5b: OUTPUT«ok 00:01 128m␤»
17:30 jnthn Hm, cute.
17:30 TimToady which would solve the problem of list infixes eating your first arg
17:31 TimToady would not imply a channel like <== might
17:32 TimToady r: say <== map {.chr} <== 65...69
17:32 camelia rakudo-parrot 8ae2de, rakudo-jvm 8ae2de, rakudo-moar 8ae2de: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/tmpfileâ�¤Unable to parse expression in quote words; couldn't find final '>'â�¤        â�¤at /tmp/tmpfile:1â�¤------> [32msay <== map {.chr} <== 65...69[33…»
17:33 djanatyn joined #perl6
17:33 TimToady r: say() <== map {.chr} <== 65...69
17:33 camelia rakudo-parrot 8ae2de, rakudo-jvm 8ae2de, rakudo-moar 8ae2de: OUTPUT«A B C D E␤»
17:33 jnthn That doesn't imply a channel today, fwiw.
17:34 jnthn Could be made to, but we might want to think through the cache conherency consequences for most problems...
17:35 * TimToady would like feeds to imply threading, even if we have to restrict them in some cases
17:36 dalek rakudo-star-daily: 0ba64a6 | coke++ | log/ (5 files):
17:36 dalek rakudo-star-daily: today (automated commit)
17:36 dalek rakudo-star-daily: review: https://github.com/coke/rakudo-star-daily/commit/0ba64a61ae
17:36 TimToady or some kind of concurrency resembling a pipe
17:36 * TimToady notes that Unix pipes batch by bufffer
17:36 jnthn TimToady: Yeah, my main concern is that people will say "oh, cute syntax" and write stuff that shoves a bunch of producer/consumer tasks off into the pool.
17:36 jnthn TimToady: But each task does a tiny amount of work
17:37 TimToady well, workers are supposed to be lightweight...
17:37 jnthn TimToady: So the work done is swamped by the inter-task communication
17:37 TimToady sure, people can be stupid :)
17:37 jnthn It's not that the workers aren't lightweight, more that memory isn't uniform...
17:38 TimToady but if that's a heavier construct, it increases the need for a way to tack a list onto a call, as above
17:38 jnthn But yeah, it's an education thing at some level.
17:38 TimToady .oO(who do we know that is in education...)
17:39 ggoebel1112 joined #perl6
17:39 djanatyn joined #perl6
17:40 jnthn I know *my* first thought when I saw feed syntax was "oh, that's much nicer than writing it backwards with sub calls". In C# I just chain method calls, but Perl 6's whitespace rules don't let me write that as cutely (.foo on each line) which, if others think like me, is another push towards feed syntax.
17:41 Tene joined #perl6
17:41 Tene joined #perl6
17:41 jnthn So I worry a little people might reach to it as a readability construct rather than a concurrency construct.
17:42 ajr_ joined #perl6
17:43 TimToady std: use MONKEY_TYPING; augment class Int { method nth(\n: *@a) { @a[n] } }; say nth 2: 0,10...100;
17:43 camelia std 09dda5b: OUTPUT«Compiled lib/MONKEY_TYPING.pm6â�¤[31m===[0mSORRY![31m===[0mâ�¤Undeclared routine:â�¤  'nth' used at line 1â�¤Check failedâ�¤FAILED 00:01 133mâ�¤Â»
17:43 TimToady hmm
17:44 TimToady std: print $*IN: "howdy"
17:44 camelia std 09dda5b: OUTPUT«ok 00:01 125m␤»
17:44 TimToady std: foo $*IN: "howdy"
17:44 camelia std 09dda5b: OUTPUT«ok 00:01 125m␤»
17:44 TimToady std: foo 2: "howdy"
17:44 camelia std 09dda5b: OUTPUT«ok 00:01 124m␤»
17:45 TimToady std: say foo 2: "howdy"
17:45 camelia std 09dda5b: OUTPUT«ok 00:01 124m␤»
17:45 * TimToady scratches head
17:45 TimToady oh, right
17:45 TimToady that : is not a precedence dropper compared to ...
17:45 [Sno] joined #perl6
17:48 * TimToady wonders if that can be a little more consistent with the other uses of :
17:53 Vlavv joined #perl6
17:55 TimToady std: use MONKEY_TYPING; augment class Int { method nth(\n: *@a) { @a[n] } }; say nth(2:): 0,10...100;
17:55 camelia std 09dda5b: OUTPUT«ok 00:01 134m␤»
17:55 TimToady works, but ugly
17:57 zby_home joined #perl6
17:58 TimToady std: use MONKEY_TYPING; augment class Int { method nth(\n: *@a) { @a[n] } }; say nth 2: list 0,10...100;
17:58 camelia std 09dda5b: OUTPUT«ok 00:01 133m␤»
17:58 TimToady slightly less ugly
17:58 TimToady nr: use MONKEY_TYPING; augment class Int { method nth(\n: *@a) { @a[n] } }; say nth 2: list 0,10...100;
17:58 camelia rakudo-parrot 8ae2de, rakudo-jvm 8ae2de, rakudo-moar 8ae2de: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/tmpfileâ�¤Two terms in a rowâ�¤at /tmp/tmpfile:1â�¤------> [32mthod nth(\n: *@a) { @a[n] } }; say nth 2[33mâ��[31m: list 0,10...100;[0mâ�¤    expecting a…»
17:58 camelia ..niecza v24-109-g48a8de3: OUTPUT«[31m===[0mSORRY![31m===[0mâ�¤â�¤Invocant handling is NYI at /tmp/tmpfile line 1:â�¤------> [32m { @a[n] } }; say nth 2: list 0,10...100[33mâ��[31m;[0mâ�¤â�¤Unhandled exception: Check failedâ�¤â�¤  at /home/p6eval/niecza/boot/lib/CORE.setting l…»
17:59 TimToady well, STD likes it...
18:00 airdisa joined #perl6
18:01 denisboyun joined #perl6
18:01 SamuraiJack joined #perl6
18:01 TimToady nr: sub nth(\n, *@a) { @a[n] }; say nth 2, list 0,10...100;
18:01 camelia rakudo-parrot 8ae2de, rakudo-jvm 8ae2de, rakudo-moar 8ae2de, niecza v24-109-g48a8de3: OUTPUT«20␤»
18:01 TimToady I guess that works on ordinary functions
18:02 TimToady nr: sub nth(\n, *@a) { @a[n] }; say nth *-1, list 0,10...100;
18:02 camelia rakudo-parrot 8ae2de, rakudo-jvm 8ae2de, rakudo-moar 8ae2de, niecza v24-109-g48a8de3: OUTPUT«100␤»
18:02 TimToady \o/
18:02 TimToady nr: sub nth(\n, *@a) { @a[n] }; say nth *-1, do 0,10...100;
18:02 camelia rakudo-parrot 8ae2de, rakudo-jvm 8ae2de, rakudo-moar 8ae2de, niecza v24-109-g48a8de3: OUTPUT«100␤»
18:03 TimToady even shorter
18:03 Vlavv_ joined #perl6
18:03 TimToady with a sharper drop
18:04 TimToady nr: sub nth(\n, *@a) { @a[n] }; say nth *-1, lazy 0,10...100;
18:04 camelia niecza v24-109-g48a8de3: OUTPUT«[31m===[0mSORRY![31m===[0mâ�¤â�¤Action method statement_prefix:lazy not yet implemented at /tmp/tmpfile line 1:â�¤------> [32m { @a[n] }; say nth *-1, lazy 0,10...100[33mâ��[31m;[0mâ�¤â�¤Unhandled exception: Check failedâ�¤â�¤  at /home/p6eval…»
18:04 camelia ..rakudo-parrot 8ae2de, rakudo-jvm 8ae2de, rakudo-moar 8ae2de: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/tmpfileâ�¤Undeclared routine:â�¤    lazy used at line 1â�¤â�¤Â»
18:05 djanatyn joined #perl6
18:06 djanatyn joined #perl6
18:07 FROGGS joined #perl6
18:10 TimToady r: say combinations(3,5)
18:10 camelia rakudo-parrot 8ae2de, rakudo-jvm 8ae2de, rakudo-moar 8ae2de: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/tmpfileâ�¤Undeclared routine:â�¤    combinations used at line 1â�¤â�¤Â»
18:10 TimToady on mine it says: "no ICU lib loaded"
18:11 TimToady why it wants ICU is a puzzle
18:11 jnthn ICU r using r-p :P
18:12 jnthn Quite possibly 'cus we translate into fixed width to parse then get a string from source to show in the error...
18:14 TimToady oh, the setting has it as: my sub combinations($n, $k) {
18:14 TimToady but it's in a subscope
18:22 grondilu combinations as a subroutine is only needed for internal use, but maybe we can change that and define it as our sub combinations
18:22 TimToady just need to move it outside the List class, I think; currently testing
18:23 FROGGS jnthn: so, you say that this patch is not totally wrong? https://gist.github.com/FROGGS/35233e86e1648ff732d9
18:24 TimToady if it's already there, and more efficient than (^5).combinations(3), why not make it visible
18:24 TimToady assuming (^5).combinations(3) even worked, which it doesn't
18:24 grondilu because it was not specced
18:24 jnthn FROGGS: oooh, I see it. Duh.
18:25 FROGGS oh oh, do telö
18:25 FROGGS tell*
18:25 jnthn FROGGS: Your patch is right except there's a line a few up that makes it wrong.
18:25 TimToady most list methods should show up in Any automatically
18:25 jnthn FROGGS: release($meth, $MVM_reg_obj);
18:25 FROGGS jnthn: yeah, I wondered if that is right, because $meth is used after that
18:25 jnthn FROGGS: That needs to happen way later
18:25 jnthn FROGGS: Right.
18:25 jnthn FROGGS: Move it and you're probably good.
18:26 FROGGS after nqp::push(@ins, call($meth, @flags, :result($meth), |@args)); I guess
18:26 jnthn Right
18:26 FROGGS cool
18:26 jnthn It was outside of the diff context so I missed it at first :(
18:26 FROGGS (gut feeling)++
18:28 jercos If you don't talk to your kids about $meth, who will?
18:29 FROGGS :P
18:33 jnthn :D
18:36 ajr joined #perl6
18:36 BullSherd joined #perl6
18:36 BullSherd left #perl6
18:41 zakalwe joined #perl6
18:41 zakalwe joined #perl6
18:44 FROGGS jnthn: I have no idea what to write in that commit msg :/
18:45 jnthn "Implement dynamic names in rxtype pass"?
18:50 dalek nqp: 7494c58 | (Tobias Leich)++ | src/vm/moar/QAST/QASTRegexCompilerMAST.nqp:
18:50 dalek nqp: implement dynamic names in rxtype pass, jnthn++
18:50 dalek nqp: review: https://github.com/perl6/nqp/commit/7494c58826
18:50 FROGGS :o)
18:56 jnthn FROGGS: That gets the thingy test passing?
18:59 FROGGS yes
18:59 FROGGS perl6-m -e 'sub circumfix:<<` `>>(*@args) { @args.join("-") }; say `3, 4, "f"`'
18:59 FROGGS 3-4-f
18:59 jnthn \o/
18:59 jnthn FROGGS++
18:59 FROGGS analysing the spectest out atm
19:00 moritz FROGGS++
19:02 pecastro joined #perl6
19:04 hoelzro hmm
19:04 hoelzro how does IO::Path ~~ Str work?
19:04 [Coke] MAGIC
19:04 hoelzro I take it that uses Any ACCEPTS?
19:04 jnthn hoelzro: Do you mean exactly that, or instances?
19:05 hoelzro jnthn: instances
19:05 jnthn hoelzro: A ~~ B is always turned into B.ACCEPTS(A)
19:05 hoelzro ah ha
19:05 jnthn So in this case, Str.ACCEPTS(...)
19:05 jnthn Which likely coerces what it gets to a Str
19:05 jnthn In this case, at least.
19:06 hoelzro the only ACCEPTS I see for Str is multi ACCEPTS(Str:D: $)
19:06 jnthn OK, then it's coercing I think
19:06 moritz r: say 4 ~~ "4"
19:06 jnthn r: say IO::Path.new('/o/m/g').Str.perl
19:06 camelia rakudo-parrot 8ae2de, rakudo-jvm 8ae2de, rakudo-moar 8ae2de: OUTPUT«True␤»
19:06 camelia rakudo-parrot 8ae2de, rakudo-jvm 8ae2de, rakudo-moar 8ae2de: OUTPUT«"/o/m/g"␤»
19:07 hoelzro r: say IO::Path.new('t/..') ~~ '..'
19:07 camelia rakudo-parrot 8ae2de, rakudo-jvm 8ae2de, rakudo-moar 8ae2de: OUTPUT«False␤»
19:07 hoelzro hmm
19:07 hoelzro something strange is afoot...
19:08 FROGGS IO::Path.new('t/..') ~~ '..' turn into '..' eq IO::Path.new('t/..').Str
19:08 FROGGS (via ACCEPTS)
19:08 hoelzro right
19:08 hoelzro but on my moar, that expression I tried is evaluated as True
19:08 hoelzro unless...
19:08 hoelzro r: say IO::Path.new('t/..') ~~ none('.', '..')
19:08 camelia rakudo-parrot 8ae2de, rakudo-jvm 8ae2de, rakudo-moar 8ae2de: OUTPUT«True␤»
19:09 spider-mario joined #perl6
19:09 hoelzro now I'm wondering how Parrot is passing
19:09 * hoelzro builds a parrot
19:10 [Coke] r: say "larger string" ~~ "rin"
19:10 camelia rakudo-parrot 8ae2de, rakudo-jvm 8ae2de, rakudo-moar 8ae2de: OUTPUT«False␤»
19:11 hoelzro [Coke]: how did your talk go?
19:11 [Coke] hoelzro: well. Could have used an hour, easy, for what I was showing.
19:11 [Coke] got slightly over 30m, with a lot of questions.
19:12 [Coke] but MarkFowler++ et al. seemed to be happy.
19:12 hoelzro slightly over? that's pretty good, considering, if you ask me!
19:12 hoelzro yeah, he showed up earlier
19:12 [Coke] apparently he used to run London.pm. Nice to have a seasoned PM pro on hand.
19:13 [Coke] Albany.pm is coming back from the dead. (yay)
19:13 hoelzro \o/
19:13 airdisa joined #perl6
19:14 dalek rakudo/nom: 80c1e76 | larry++ | src/core/List.pm:
19:14 dalek rakudo/nom: move combinations/permutations subs out of List
19:14 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/80c1e76b80
19:17 dalek rakudo/nom: 8d2fb6c | larry++ | src/core/Any.pm:
19:17 dalek rakudo/nom: combinations/permutations as list methods via Any
19:17 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/8d2fb6c4cc
19:23 [Coke] did larry just commit to -rakudo- ? O_o
19:23 moritz yes, and it wasn't the first time he did
19:24 psch joined #perl6
19:25 psch hi #perl6
19:25 jercos hi psch
19:26 TimToady no, done it before
19:27 TimToady don't remember what
19:28 * TimToady remembers jnthn++ yelling at me for leaving commented out code in
19:28 zakharyas joined #perl6
19:29 FROGGS hehe
19:29 jeffreykegler left #perl6
19:29 FROGGS that sounds like our rude jnthn++ :P
19:31 colomon TimToady++
19:35 dalek rakudo/nom: dc96308 | larry++ | src/core/List.pm:
19:35 dalek rakudo/nom: .permutations should also return arrays
19:35 dalek rakudo/nom:
19:35 dalek rakudo/nom: Someone fixed .combinations earlier but missed .permutations
19:35 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/dc96308ed3
19:35 moritz TimToady: that was a faster variant of combinations, iirc
19:36 TimToady oh, yeah, about 7 times faster iirc :)
19:50 pmichaud https://github.com/rakudo/rakudo/commits?author=TimToady
19:50 treehug88 joined #perl6
19:50 pmichaud The 2008 commits are pre-git, of course.  :)
19:52 pmichaud Also:  http://osrc.dfm.io/timtoady
19:52 pmichaud "Larry Wall is a top notch Perl coder who loves pushing code."
19:52 pmichaud Yeah, he's been pushing his code on us for over 25 years now.  It's like we're addicted or something.  :-P
19:52 jercos gotta get my code fix, man.
19:53 * pmichaud once again reflects on the double-meaning of the word "user".
19:53 pmichaud Perl has more users than any other language.  I mean, the top of Perl programs are liberally sprinkled with "use".
19:54 moritz ooh, since github now recognizes Perl 6 code, so does osrc.dfm.io
19:54 pmichaud "In particular, Larry seems to be a pretty serious Perl expert."
19:55 moritz he does now?
19:55 TimToady Bah, I'm rarely serious.
19:55 jercos Sirius? Didn't they buy XM?
19:56 perigrin TimToady: also you're seriously rare.
19:56 TimToady well done!
19:57 pmichaud Hmm, my profile doesn't seem to recognize Perl 6 code.
19:57 moritz mine only has a tiny fraction of Perl code marked as Perl 6
19:58 moritz so it's probably not reliable yet
19:58 colomon I thought from what hoelzro++ told me earlier it would only recognize p6 code with p6-specific filenames.
19:59 moritz does .nqp count?
19:59 pmichaud that's possible.
19:59 colomon (so far, with "use v6" recognition coming whenever github updates to his latest code -- probably months from now.)
19:59 colomon moritz: I don't know.
20:00 FROGGS "It seems like Larry is—or should be—friends with Tobias Leich. With this in mind, it's worth noting that Tobias is less of a pusher."
20:00 FROGGS :P
20:01 jnap joined #perl6
20:05 dalek specs: 0391fe1 | larry++ | S32-setting-library/Containers.pod:
20:05 dalek specs: doc sub forms of combinations, permutations
20:05 dalek specs: review: https://github.com/perl6/specs/commit/0391fe1f15
20:07 TimToady LHF for someone to write tests for the new capabilities
20:07 TimToady r: .say for <a b c>.permuations
20:07 camelia rakudo-moar 8ae2de: OUTPUT«Cannot find method 'permuations'␤  in block  at /tmp/tmpfile:1␤␤»
20:07 camelia ..rakudo-parrot 8ae2de, rakudo-jvm 8ae2de: OUTPUT«No such method 'permuations' for invocant of type 'Parcel'␤  in block  at /tmp/tmpfile:1␤␤»
20:07 TimToady r: .say for <a b c>.permutations
20:07 camelia rakudo-parrot 8ae2de, rakudo-jvm 8ae2de: OUTPUT«No such method 'permutations' for invocant of type 'Parcel'␤  in block  at /tmp/tmpfile:1␤␤»
20:07 camelia ..rakudo-moar 8ae2de: OUTPUT«Cannot find method 'permutations'␤  in block  at /tmp/tmpfile:1␤␤»
20:07 TimToady ENOTYET
20:08 pmichaud EAGAIN
20:08 FROGGS SPAGAIN
20:08 FROGGS r: my sub foo(:$n is required, *%h, *@a) { say 1 }; foo 1, x => 20, y => 300, 4000
20:08 camelia rakudo-moar 8ae2de: OUTPUT«1␤»
20:08 camelia ..rakudo-parrot 8ae2de, rakudo-jvm 8ae2de: OUTPUT«Required named parameter 'n' not passed␤  in sub foo at /tmp/tmpfile:1␤  in block  at /tmp/tmpfile:1␤␤»
20:08 FROGGS r: my sub foo(:$n!, *%h, *@a) { say 1 }; foo 1, x => 20, y => 300, 4000
20:08 camelia rakudo-parrot 8ae2de, rakudo-jvm 8ae2de, rakudo-moar 8ae2de: OUTPUT«Required named parameter 'n' not passed␤  in sub foo at /tmp/tmpfile:1␤  in block  at /tmp/tmpfile:1␤␤»
20:09 stevan_ joined #perl6
20:13 airdisa joined #perl6
20:14 pmichaud afk again
20:17 * TimToady senses that grondilu++ is working on http://rosettacode.org/wiki/Chinese_Remainder_Theorem
20:19 uvtc joined #perl6
20:19 treehug88 joined #perl6
20:22 treehug88 joined #perl6
20:22 jnthn FROGGS: Are you on that "is required" bug already?
20:22 jnthn FROGGS: If not, I think I know what it'll be...
20:22 FROGGS jnthn: I just scratch my head about it
20:22 FROGGS r: my sub foo(:$n is required, *%h, *@a) { say &?ROUTINE.signature.params[0].optional }; foo 1, x => 20, y => 300, 4000
20:22 camelia rakudo-parrot 8ae2de, rakudo-jvm 8ae2de: OUTPUT«Required named parameter 'n' not passed␤  in sub foo at /tmp/tmpfile:1␤  in block  at /tmp/tmpfile:1␤␤»
20:22 camelia ..rakudo-moar 8ae2de: OUTPUT«False␤»
20:23 FROGGS I mean, ! and 'is required' pretty much do the same
20:24 FROGGS I'd almost think it is about getstaticcode or some such
20:24 jnthn Nah
20:24 jnthn It's da lowerer
20:28 jnthn Trying a patching now
20:28 FROGGS I am looking at the failing Signature.gist
20:29 jnthn ok
20:29 TimToady r: .say for <a b c>.permutations
20:29 camelia rakudo-parrot 8ae2de, rakudo-jvm 8ae2de: OUTPUT«No such method 'permutations' for invocant of type 'Parcel'␤  in block  at /tmp/tmpfile:1␤␤»
20:29 camelia ..rakudo-moar 8ae2de: OUTPUT«Cannot find method 'permutations'␤  in block  at /tmp/tmpfile:1␤␤»
20:29 TimToady when does it recompile?
20:30 FROGGS ten past, every hour
20:30 FROGGS at least it was, some time ago
20:30 moritz 8  * * * *  perl ~/evalbot/build.pl rakudo    > /dev/null
20:30 FROGGS ohh, off by two -.-
20:31 TimToady does it take 22 minutes to recompile?
20:31 jnthn On the upside, Moar builds are fairly fast :P
20:31 jnthn TimToady: Possibly if it's doing all 3
20:31 beastd joined #perl6
20:31 moritz TimToady: depends on whether nqp needs to be recompiled
20:31 hoelzro colomon: GH uses a Bayesian classifier to figure out the file type
20:31 moritz r:  .say for <a b c>.permutations
20:31 camelia rakudo-parrot 8ae2de, rakudo-jvm 8ae2de: OUTPUT«No such method 'permutations' for invocant of type 'Parcel'␤  in block  at /tmp/tmpfile:1␤␤»
20:31 camelia ..rakudo-moar 8ae2de: OUTPUT«Cannot find method 'permutations'␤  in block  at /tmp/tmpfile:1␤␤»
20:32 hoelzro I have a PR with some enhancements
20:32 TimToady btw, that call coredumps perl6-m
20:32 hoelzro but it has yet to be acknowledged
20:32 moritz huh, it seems to be stuck
20:32 jnthn FROGGS: Think I made it work; which test was it, ooc?
20:32 FROGGS jnthn: t/spec/S06-signature/slurpy-params.rakudo.moar
20:33 jnthn FROGGS: ah,found it, np
20:33 jnthn oh
20:33 jnthn named-parameters passes after my fix
20:33 * jnthn checks that one too
20:33 jnthn hah, 2 for the price of 1
20:33 FROGGS *g*
20:33 FROGGS win!
20:34 TimToady lunch &
20:34 jnthn It seems that we fail some tests 'cus the signature gist patch no longer strips Any
20:34 jnthn Thing is
20:34 jnthn Any is not always the default
20:34 moritz TimToady: also note that the server has two rakudo build dirs, and uses each in turn (to minimize downtime between the rebuilds), so it has to rebuild nqp twice
20:34 colomon hoelzro: really?  I had no idea.
20:34 moritz and yes, it always does all three
20:35 grondilu p6:  say 10 div $_ for <1 2 5 7>;
20:35 grondilu ?
20:35 camelia rakudo-jvm dc9630: OUTPUT«(timeout)»
20:35 camelia ..niecza v24-109-g48a8de3: OUTPUT«10␤5␤2␤1␤»
20:35 camelia ..rakudo-parrot dc9630: OUTPUT«Cannot call 'infix:<div>'; none of these signatures match:␤:(Int:D \a, Int:D \b)␤:(int $a, int $b)␤  in any  at gen/parrot/BOOTSTRAP.nqp:1219␤  in any  at gen/parrot/BOOTSTRAP.nqp:1210␤  in sub infix:<div> at gen/parrot/CORE.setting:4013␤  in…»
20:35 camelia ..rakudo-moar dc9630: OUTPUT«Cannot call 'infix:<div>'; none of these signatures match:␤:(Int:D \a, Int:D \b)␤:(int $a, int $b)␤  in sub infix:<div> at src/gen/m-CORE.setting:4009␤  in block  at /tmp/tmpfile:1␤␤»
20:35 grondilu p6: say 10 div "3"
20:35 camelia rakudo-parrot dc9630: OUTPUT«Cannot call 'infix:<div>'; none of these signatures match:␤:(Int:D \a, Int:D \b)␤:(int $a, int $b)␤  in any  at gen/parrot/BOOTSTRAP.nqp:1219␤  in any  at gen/parrot/BOOTSTRAP.nqp:1210␤  in sub infix:<div> at gen/parrot/CORE.setting:4013␤  in…»
20:35 camelia ..niecza v24-109-g48a8de3: OUTPUT«3␤»
20:35 camelia ..rakudo-moar dc9630: OUTPUT«Cannot call 'infix:<div>'; none of these signatures match:␤:(Int:D \a, Int:D \b)␤:(int $a, int $b)␤  in sub infix:<div> at src/gen/m-CORE.setting:4009␤  in block  at /tmp/tmpfile:1␤␤»
20:36 camelia ..rakudo-jvm dc9630: OUTPUT«java.lang.NullPointerException␤  in sub infix:<=:=> at gen/jvm/CORE.setting:1322␤  in block  at gen/jvm/CORE.setting:16460␤  in method perl at gen/jvm/CORE.setting:9817␤  in method perl at gen/jvm/CORE.setting:1041␤  in method message at gen/jvm/C…»
20:36 moritz div explicitly wants two same-typed arguments
20:36 moritz and rakudo doesn't yet do the <...> magical auto-typing
20:36 grondilu ok
20:36 moritz it's a bug that niecza accepts 10 div "3"
20:38 lue Naturally, the recompile is done just when TimToady leaves :)
20:38 FROGGS r: .say for <a b c>.permutations # TimToady
20:38 camelia rakudo-moar dc9630: OUTPUT«Cannot call 'postcircumfix:<[ ]>'; none of these signatures match:␤:(Any \SELF, int $pos)␤:(Any \SELF, int $pos, :BIND($BIND)!)␤:(Any \SELF, int $pos, Any :SINK($SINK)!, *%other)␤:(Any \SELF, int $pos, Any :delete($delete)!, *%other)␤:(Any \SELF,…»
20:38 camelia ..rakudo-parrot dc9630, rakudo-jvm dc9630: OUTPUT«a b c␤a c b␤b a c␤b c a␤c a b␤c b a␤»
20:40 erkan joined #perl6
20:40 uvtc Hi, #perl6. Was looking around to see if any new tutorials were available. Is the perl6/book still usable? Or is it out-of-date and in need of updates at this point?
20:41 FROGGS uvtc: dunno about the book, but have you seen the advent calender?
20:41 hoelzro colomon: yeah, it's kinda annoying =/
20:41 uvtc Lemme hava look...
20:42 [Coke] the book hasn't been updated since the last time you looked.
20:42 lue uvtc: Last time I poked it, the book needs an overhaul in all the backend stuff. I imagine the book's contents are still somewhat useful though.
20:42 [Coke] jnthn++'s presentations are always good, but not really bookish.
20:42 Guest1338 joined #perl6
20:43 uvtc Ok, there's a nice 2013 P6 Advent Table of Contents page.
20:44 uvtc Will have a look at some of those articles.
20:46 stevan_ joined #perl6
20:48 PerlJam uvtc: What sort of things would you like to see a tutorial on?  (maybe someone could be convinced to write it)
20:48 uvtc Just built Rakudo/Moar from git clone. Starts fast! :)
20:49 jnap joined #perl6
20:50 uvtc PerlJam, one book I've found useful in the past (for Python) is the Python in a Nutshell book (by Alex Martelli). The first few chapters seem well organized and brief. Would like to see something like that for P6.
20:52 uvtc PerlJam, I mean a from-the-ground-up full-on language tutorial.
20:52 PerlJam .oO( Perl 6 in a Chrysalis? )
20:53 PerlJam uvtc: oh ... that's been on my todo list forever.
20:53 uvtc Perl 6 Unfurled :)
20:54 PerlJam uvtc: I think if you ask for a tutorial on a more focused topic area, it would be easier for someone to write than a full-blown Perl 6 tutorial
20:54 PerlJam (someone not just me :)
20:55 dalek rakudo/nom: 37f0403 | jnthn++ | src/Perl6/Actions.nqp:
20:55 dalek rakudo/nom: Fix lowering if 'is required'.
20:55 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/37f04039fb
20:55 PerlJam or it may already exist somewhere.
20:55 kaare_ joined #perl6
20:55 itz_ https://github.com/perl6/book ?
20:55 raiph joined #perl6
20:56 uvtc PerlJam, Ok. Focused smaller tutorials might have a better chance at being written.
20:56 PerlJam itz_: none of us has been inspired to work on that in a long while.
20:56 uvtc itz_, yes, that's what I was asking about a few moments ago. It appears to be a bit out of date.
20:56 itz_ uvtc: you could fork and modify!
20:57 uvtc Writing in general is difficult, time-consuming, personal, and often fairly thankless. :)
20:57 dalek rakudo/nom: 4aa8c5d | (Tobias Leich)++ | src/core/ (2 files):
20:57 dalek rakudo/nom: fix Parameter/Signature.gist
20:57 dalek rakudo/nom:
20:57 dalek rakudo/nom: We dont state thate a param is Any, because this is the default. This also
20:57 dalek rakudo/nom: fixes an explosion on jvm.
20:57 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/4aa8c5d690
20:58 PerlJam uvtc: thought itz_ has a good point.  You could contribute to the knowledge base yourself.  I, for one, would welcome new/different voices.
20:58 PerlJam s/thought/though/
20:59 PerlJam uvtc: "The Perl 6 Experience"  :)
20:59 dalek rakudo/nom: 24c544b | (Tobias Leich)++ | tools/build/NQP_REVISION:
20:59 dalek rakudo/nom: bump nqp rev to get latest fixes
20:59 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/24c544b994
21:01 uvtc PerlJam, personally, I don't find the approach taken by perl6/book to be as useful for me as one which starts at the bottom with the bits and pieces and works its way up.
21:01 moritz the book is dead anyway.
21:03 PerlJam merlyn and bdf still have a "Learning Perl 6" placeholder I think.  Maybe they could be convinced to actually write it.
21:03 itz_ merlyn was down on the idea when I spoke with him in a London pub recently
21:04 itz_ I did like the approach of the original Learning Perl as well the Nutshell guide mentioned
21:04 PerlJam yeah, he's been down on it every time I talk with about it too.  Something about book sales not being big enough  ;)
21:04 lue uvtc: what approach does work for you then, ooc?
21:05 uvtc lue, well, a pretty close description is the approach taken by chapter 4 of that Python in a Nutshell book I mentioned. It assumes you know some programming already, and then, says,
21:06 uvtc "here's how tokens work, identifiers, language keywords, operators, literals, delimiters, statements, etc."
21:07 uvtc lue, and works its way up from there (from the bottom up like that).
21:08 itz_ maybe the modern perl book (Creative Commons) could be modified
21:09 moritz no
21:09 lue uvtc: what do the first three chapters do, though? :)
21:10 uvtc lue, oh, lessee ... here's the link http://shop.oreilly.com/product/9780596100469.do .
21:11 lue Ah, fairly standard beginning-tutorial stuff I see.
21:12 FROGGS moritz: what is up with that book?
21:12 moritz FROGGS: it's dead.
21:12 uvtc , Jim.
21:12 uvtc :)
21:12 moritz just look at the timestamps of the last few commits
21:12 FROGGS moritz: in what way?
21:12 FROGGS I mean, it would be dead wood if it was printed
21:13 FROGGS ahhh
21:13 FROGGS okay
21:13 masak 'night, #perl6
21:13 moritz the last view times I've tried to contribute, I got starved by lonliness
21:13 FROGGS gnight masak
21:13 lue ♞ masak o/
21:13 moritz I've tried to port some of the perl6/book stuff into perl6/doc
21:14 denis_boyun joined #perl6
21:14 lue uvtc: Funny thing is, I can't stand most tutorials. I use them to get a basic handle on things, and then I do much better stumbling through whatever it is I'm trying to do in the first place, reference in hand :P
21:14 airdisa joined #perl6
21:14 itz_ the Python Nutshell book is very cheap on Google Books (at least in the UK)
21:15 [Coke] writing a book is hard, and most of the people here are writing the software.
21:15 uvtc lue, From one angle, I think I know what you mean. It's difficult to write without wasting your reader's time.
21:16 uvtc lue, that's why I like the nuts-and-bolts approach; just give the reader the details they need. They already know how to program, they just need to know how this language works so they can move one.
21:16 uvtc move on.
21:16 lue moritz: last time I tried, I found that I'd have to whack the generator into shape to do much.
21:16 itz_ I liked the cheatsheet
21:17 lue (IIRC correctly, it should be using free fonts, and also XeLaTeX for the Unicode support.)
21:17 dalek perl6-roast-data: 64466a4 | coke++ | / (6 files):
21:17 dalek perl6-roast-data: today (automated commit)
21:17 dalek perl6-roast-data: review: https://github.com/coke/perl6-roast-data/commit/64466a42a2
21:17 uvtc itz_, do you mean this? https://raw.github.com/perl6/mu/master/docs/Perl6/Cheatsheet/cheatsheet.txt
21:17 itz_ yes .. I think its in star as well
21:18 uvtc Ah. I got that link from http://perl6.org/documentation/ .
21:19 uvtc lue, I'd actually tried starting something like it in the past:  http://wiki.perl6.org/Mostly%20Harmless%20Docs/Main
21:21 uvtc But I found myself asking endless questions in this channel, and then ran out of steam.
21:24 geekosaur joined #perl6
21:25 treehug88 joined #perl6
21:26 stevan_ joined #perl6
21:26 dayangkun joined #perl6
21:31 jnthn FROGGS: You fixed the JVM null pointer exception causing thing, did you say?
21:31 FROGGS jnthn: I did
21:32 FROGGS started a spectest just now
21:33 jnthn \o/
21:33 treehug88 joined #perl6
21:33 jnthn I'm working on the thing that blows up JSON::Tiny
21:33 Mouq joined #perl6
21:35 FROGGS uhhh
21:35 FROGGS ++jnthn
21:35 FROGGS I'll look at t/spec/S02-types/isDEPRECATED.rakudo.jvm ..................... Failed 20/20 subtests
21:35 jnthn I think that's failed for a while. But yeah, do take a look
21:36 FROGGS Error while creating backtrace: No such method 'Int' for invocant of type 'Any'
21:36 FROGGS interesting
21:36 jnthn Oh...
21:36 FROGGS Int-eresting, even
21:36 jnthn I wonder if it leaves empty entries in the backtrace object when it doesn't figure out the line number...
21:36 FROGGS that is why I wanna take a look, it fails for too long now :o)
21:37 diakopter MIN_INT
21:37 FROGGS jnthn: I guessed something like that too
21:38 jnthn bah, I figured at least one reason why the JSON bug, and think I found a more general thing to improve...
21:38 FROGGS you say it as if it were a good and a bad thing at the same time
21:39 jnthn Well, one of those, "hm, I wonder what the history of this is" things :)
21:39 FROGGS ahh, yeah
21:40 FROGGS you will remember when half the tests assplode :o)
21:46 treehug88 joined #perl6
21:53 pdcawley joined #perl6
21:55 japhb uvtc: I love the name 'Perl 6 Unfurled'.  :-)
21:56 uvtc japhb, :)
21:58 lue For some reason, trying to generate perl6/book gets me failures in processing UTF-8. The prime suspect for this appears to be Perl 5...
22:00 FROGGS r: sub a is DEPRECATED { }; a()
22:00 camelia rakudo-moar 24c544: ( no output )
22:00 camelia ..rakudo-jvm 24c544: OUTPUT«No such method 'Int' for invocant of type 'Any'␤␤Error while creating backtrace: No such method 'Int' for invocant of type 'Any'␤  in block  at gen/jvm/CORE.setting:10679␤  in block  at gen/jvm/CORE.setting:7370␤  in method reify at gen/jvm/CORE.s…»
22:00 camelia ..rakudo-parrot 24c544: OUTPUT«Saw 1 call to deprecated code during execution.␤================================================================================␤Sub a (from GLOBAL) called at:␤  /tmp/tmpfile, line 1␤Please use something else instead.␤--------------------------…»
22:00 lue (Unless there's a "Carl Mäsak" I don't know about, that is ☺)
22:01 jnthn lue: I saw a guy wearing a badge like that at a conf once... :)
22:15 airdisa joined #perl6
22:18 FROGGS hmmm, jvm's stage parse is at 92s, which is way more than usual
22:18 jnthn o.O
22:22 dalek rakudo/nom: 387664f | (Tobias Leich)++ | src/core/Backtrace.pm:
22:22 dalek rakudo/nom: skip too if line is not defined
22:22 dalek rakudo/nom:
22:22 dalek rakudo/nom: The line number and file is not guaranteed to be set. Calling .Int on an undefined
22:22 dalek rakudo/nom: value would explode, so we move that to after the skip.
22:22 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/387664f38c
22:23 jnthn OK, with local patches, JSON tiny passed its tests (in 2s) on Moar.
22:23 FROGGS both sounds very nice
22:23 diakopter :)
22:23 * jnthn spectests
22:24 jnthn Darn, I've busted stuff :/
22:24 FROGGS that would mean that v5 might work
22:24 FROGGS >.<
22:24 jnthn oh, looks like a silly one
22:24 jnthn FROGGS: It may.
22:25 jnthn FROGGS: The thing I fixed is not the thing I expected to be fixing.
22:25 FROGGS hehe
22:25 jnthn FROGGS: And the other thing is still to do.
22:26 jnthn I guess JSON::Tiny working was the next Panda blocker...
22:26 FROGGS well, I am not even sure what in v5 is causing the fail, because I was unable to golf it down
22:34 dalek nqp: 080ce0b | jnthn++ | src/how/NQPClassHOW.nqp:
22:34 dalek nqp: Disable SC barrier around class cache flush.
22:34 dalek nqp: review: https://github.com/perl6/nqp/commit/080ce0bf20
22:35 treehug88 joined #perl6
22:37 dalek rakudo/nom: 8004501 | jnthn++ | src/ (2 files):
22:37 dalek rakudo/nom: Compute sig arity/count at compile time.
22:37 dalek rakudo/nom:
22:37 dalek rakudo/nom: Before, probably for some historical reasons that are now gone, we did
22:37 dalek rakudo/nom: this computation at runtime and cached it. This was bad for needing to
22:37 dalek rakudo/nom: do the "did we compute it yet" check, but also because it triggered SC
22:37 dalek rakudo/nom: write barriers, wasting time and bloating modules.
22:37 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/8004501ef1
22:37 dalek rakudo/nom: fd8316f | jnthn++ | tools/build/NQP_REVISION:
22:37 dalek rakudo/nom: Get latest NQP fixes.
22:37 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/fd8316f664
22:38 FROGGS joined #perl6
22:39 jnthn FROGGS: ^^ seems to fix JSON::Tiny tests
22:44 stevan_ joined #perl6
22:45 FROGGS joined #perl6
22:46 daxim joined #perl6
22:49 Sqirrel joined #perl6
22:52 FROGGS ===SORRY!===
22:52 FROGGS Object conflict detected during deserialization.
22:52 FROGGS v5 does not like it yet
22:57 jnthn Ah, then you probably have the Real Issue.
22:57 FROGGS (tm)
23:01 dmol joined #perl6
23:02 uvtc left #perl6
23:03 jnthn FROGGS: I hoped pre-compiling JSON::Tiny might reveal it, but no, that works.
23:03 FROGGS :/
23:03 jnthn ...and brings JSON::Tiny make test time down to 1s with parallel tests..
23:04 jnthn FROGGS: How might I reproduce the v5 thing?
23:04 jnthn FROGGS: Is it a lot of build hacking for me to try it?
23:05 FROGGS jnthn: no, clone v5, apply rakudo.patch, m-install in rakudo, and after configuring v5 do a m-install there
23:05 FROGGS git@github.com:rakudo-p5/v5.git
23:05 jnthn lemme try it
23:05 FROGGS note that its configure script is actually a perl 6 script
23:08 jnthn Any reason the final hunk of that patch isn't already in Rakudo?
23:08 jnthn (the postcircumfix bit)
23:08 FROGGS it makes it parse but these are still not callable
23:09 Sqirrel_ joined #perl6
23:09 FROGGS I can apply that if that is no problem
23:09 jnthn makefile(1) : fatal error U1033: syntax error : '=' unexpected
23:09 jnthn Stop.
23:09 FROGGS :(
23:09 FROGGS ahh, yeah
23:09 FROGGS damn
23:10 jnthn ugh
23:11 tadzik jnthn++ # PLPW appearance
23:11 jnthn I fixed a couple of things but the rest I dunno how to ddeal with
23:11 jnthn Oh, I guess I can hard code stuf
23:11 FROGGS jnthn: what is "the rest"?
23:11 jnthn M_NQPLIB = $(shell $(M_NQP) -e 'print(nqp::backendconfig<prefix> ~ "/languages/nqp");')
23:11 jnthn M_P6LIB  = $(shell $(M_PERL6) -e 'print %*CUSTOM_LIB<perl>')
23:12 FROGGS yes, I am working on that atm
23:14 jnthn I hacked it.
23:14 jnthn it's building
23:14 jnthn Where should it fail?
23:14 jnthn I see it building lots of stuff :)
23:15 jnthn I get to Compiling Config to mbc and then it can't find Perl5::Terms
23:16 FROGGS it should fail when hitting Perl5::Config after Perl5::Terms
23:16 jnthn ah, I see it
23:16 FROGGS it should find it...
23:16 airdisa joined #perl6
23:16 jnthn Yeah, the PERL6LIB=... perl6-m ... construct doesn't work out on Windows.
23:17 jnthn Hmm, it got way past Config here.
23:17 FROGGS Object conflict detected during deserialization.
23:17 FROGGS (Probable attempt to load two modules that cannot be loaded together).
23:17 FROGGS make: *** [lib/Perl5/Config.moarvm] Fehler 1
23:17 jnthn Into Module::BUILD at the moment...
23:18 FROGGS it fails at that point for me...
23:18 jnthn FROGGS: You did grab my NQP patch as well as the Rakudo one?
23:18 FROGGS ye
23:18 FROGGS yes
23:18 jnthn huh
23:18 jnthn Well, it got past Config here...
23:18 FROGGS I pulled again (all repos) and it only pulled the dirops patch
23:19 jnthn It completed...
23:19 FROGGS hmpf
23:19 FROGGS I typed "make m-install" in nqp, and it did sort of nothing
23:20 jnthn FROGGS: How can I try out v5 now I make m-instlal'd it? :)
23:20 FROGGS perl6-m -e "use v5; say $^O"
23:21 jnthn ooh
23:21 jnthn C:\consulting\perl6\v5>perl6-m -e "use v5; foreach my $x (1,2,3) { say $x; }"
23:21 jnthn 1
23:21 jnthn 2
23:21 jnthn 3
23:21 jnthn C:\consulting\perl6\v5>perl6-m -e "use v5; say $^O"
23:21 jnthn MSWin32
23:21 jnthn Cool!
23:21 FROGGS :o)
23:21 FROGGS it does what it should... mostly
23:21 jnthn FROGGS: Lemme know if you can't get it to work for you still...
23:22 FROGGS I will
23:23 jnthn cute...a chunk of Perl 5 on MoarVM. :D
23:23 FROGGS *g*
23:23 FROGGS sigil-variance ftw!!
23:24 FROGGS I like to mix both worlds like: use v5; sub foo { use v6.0.0; say @_.perl }; foo(1, 2, 3)
23:24 jnthn yes :)
23:24 FROGGS you can even do: say 1->WHAT
23:24 jnthn Geez...
23:25 jnthn FROGGS++ # v5
23:25 FROGGS thanks
23:28 sftp joined #perl6
23:29 diakopter m: { use v5; say 1->WHAT }
23:29 camelia rakudo-moar 24c544: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/VCJxLeMekg�Unsupported use of -> as postfix; in Perl 6 please use either . to call a method, or whitespace to delimit a pointy block�at /tmp/VCJxLeMekg:1�------> [32m{ use v5; say 1->[33m�…»
23:32 lue Just so I know, how long does it usually take to process a CLA (sent via email)?
23:33 FROGGS lue: I am not sure that you will get any feedback
23:34 lue ah, so I'll just have to bug [Coke] in a week or so to check The List™, right? :)
23:34 FROGGS lue: for a rakudo commit bit?
23:34 lue yes.
23:35 FROGGS ohh, I'd say you would need at least jnthn's blessing too
23:35 FROGGS and I also asked pmichaud and moritz for their +1/-1
23:36 lue FROGGS: well sure, I meant bugging [Coke] more for checking that the CLA got processed than being the sole step in getting access.
23:37 FROGGS ahh, yeah
23:44 FROGGS it still fails for me...
23:45 FROGGS I pull in all repos again, and built all afresh
23:45 FROGGS I even deleted the installed v5 files
23:47 jnthn FROGGS: Huh. That's really weird. You got install paths right, finding the right perl6-m, etc?
23:48 FROGGS I only have one perl6-m
23:48 FROGGS really
23:48 FROGGS which perl6-m
23:48 FROGGS /home/froggs/dev/nqp/install/bin/perl6-m
23:48 FROGGS and only one nqp, one rakudo etc
23:48 FROGGS PERL6LIB=lib perl6-m -MPerl5 -e 'use v6.0.0; sub EXPORT(|) { use Perl5::Terms }'
23:48 FROGGS ===SORRY!===
23:48 FROGGS Object conflict detected during deserialization.
23:48 jnthn odd
23:48 jnthn FROGGS: Did you make clean in your v5 too?
23:49 FROGGS yes
23:49 jnthn r: 1.omg
23:50 FROGGS and there is no *.moarvm left after a m-clean, I checked
23:50 camelia rakudo-moar fd8316: OUTPUT«Cannot find method 'omg'␤  in block  at /tmp/tmpfile:1␤␤»
23:50 camelia ..rakudo-parrot fd8316, rakudo-jvm fd8316: OUTPUT«No such method 'omg' for invocant of type 'Int'␤  in block  at /tmp/tmpfile:1␤␤»
23:50 FROGGS btw, a v5 block get Perl5::Terms imported implicitly
23:50 FROGGS I guess that is part of the problem
23:52 jnthn Yeah, I'm just confused why you get the error and I don't...
23:52 FROGGS yeah, that problem should not be os related at all
23:53 jnthn Right
23:53 FROGGS humm, precomp is the problem
23:53 jnthn Oh?
23:53 FROGGS this passes after a m-clean: PERL6LIB=lib perl6-m -MPerl5 -e 'use v6.0.0; sub EXPORT(|) { use Perl5::Terms }'
23:53 FROGGS takes 30s or so but works
23:53 FROGGS and failed before
23:58 jnthn That one runs in like a second for me...
23:58 FROGGS hmmm, after compiling Perl5::Terms I can't even do: perl6-m -e 'use v5'
23:59 FROGGS so it has nothing to do with Config.pm, but entirely with loading that Terms.pm
23:59 araujo joined #perl6

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

Perl 6 | Reference Documentation | Rakudo