Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2013-11-25

Perl 6 | Reference Documentation | Rakudo

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

All times shown according to UTC.

Time Nick Message
00:02 Ben_Goldberg joined #perl6
00:03 denis_boyun joined #perl6
00:04 arnsholt_ joined #perl6
00:12 Psyche^ joined #perl6
00:17 timotimo berekuk: i just ran a chrome developer tools audit thingie on questhub
00:20 timotimo there's a crapton of unused CSS rules (which is probably just bootstrap being what it is), but you could let the browser cache a lot of things
00:20 timotimo also interesting: Optimize the order of styles and scripts (4) To ensure CSS files are downloaded in parallel, always include external CSS before external JavaScript.
00:21 timotimo and it claims that button.css and button.min.js would save about two thirds of their size by getting gzip-compressed by your web server
00:21 timotimo just tiny things to make your site snappier :)
00:21 dalek joined #perl6
00:22 bloonix joined #perl6
00:36 [Coke] looks like niecza is failing a bunch of tests.
00:37 [Coke] as is rakudo.jvm
00:39 [Coke] preflex: tell moritz https://github.com/moritz/Math-Model/pull/2 is the only  deprecation warning left in the R* trial builds.
00:39 preflex Consider it noted.
00:39 raiph joined #perl6
00:45 [Coke] one more failing test on lwp-simple
00:46 [Coke] er, one *remaining* failing test. looks like most of the failures have been fixed.
00:48 dayangkun joined #perl6
00:54 BenGoldberg preflex: seen yoleaux
00:54 preflex yoleaux was last seen on #perl6 30 days, 5 hours, 32 minutes and 35 seconds ago, saying: 18:23Z <diakopter> Mouq: awesome work on stdsigspace Mouq++
01:04 MikeFair joined #perl6
01:06 cognominal__ joined #perl6
01:07 pecastro joined #perl6
01:07 salv0 joined #perl6
01:08 moritz joined #perl6
01:09 rurban1 joined #perl6
01:10 rurban2 joined #perl6
01:26 colomon [Coke]: dang it, merged your Math::Model pull request and *then* figured out it doesn't work.  :\
01:30 colomon r: class A { has %!a; method b() { %!a{"hello"}:exists; };  }; A.new.b
01:30 camelia rakudo-parrot 60c8c1, rakudo-jvm 60c8c1: ( no output )
01:30 colomon r: class A { has %!a; method b() { 1 || %!a{"hello"}:exists; };  }; A.new.b
01:30 camelia rakudo-jvm 60c8c1: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/rqb98vBvJlâ�¤You can't adverb thatâ�¤at /tmp/rqb98vBvJl:1â�¤------> [32ma; method b() { 1 || %!a{"hello"}:exists[33mâ��[31m; };  }; A.new.b[0mâ�¤    expecting any of:â�¤        pair valueâ�¤Â»
01:30 camelia ..rakudo-parrot 60c8c1: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/ftMf8XqZ2xâ�¤You can't adverb thatâ�¤at /tmp/ftMf8XqZ2x:1â�¤------> [32ma; method b() { 1 || %!a{"hello"}:exists[33mâ��[31m; };  }; A.new.b[0mâ�¤    expecting any of:â�¤        pair valueâ�¤Â»
01:31 colomon rakudobug?
01:32 colomon anyway, I've added parens around it in Math::Model and everything is happy again.
01:33 colomon moritz: ^^
01:37 * colomon feels good making even a tiny little commit somewhere.
01:50 Entonian joined #perl6
02:00 raiph joined #perl6
02:05 rurban1 joined #perl6
02:08 prevost joined #perl6
02:32 [Coke] colomon: thank you for fixing it. I think I didn't test it because it had prereqs. sorry about that
02:32 colomon no worries
02:33 colomon I made sure to set up panda first so the prereqs would be easy.
02:34 colomon and now I have an up-to-date rakudo and panda and whatnot
02:35 colomon Actually merging the pull request was the easiest way I could figure out to test it.  I know there must be something better.
02:48 kurahaupo_ joined #perl6
02:50 kurahaupo_ joined #perl6
02:50 kurahaupo_ left #perl6
02:59 btyler joined #perl6
03:16 FROGGS joined #perl6
03:27 BenGoldberg p56: say 'à';
03:27 camelia p5-to-p6 : OUTPUT«say(chr(195) ~ chr(160))␤»
03:34 TimToady um, no
03:36 telex joined #perl6
03:46 kaleem joined #perl6
04:04 adu joined #perl6
04:04 adu std: $*LANG
04:04 camelia std 8370f3f: OUTPUT«ok 00:01 121m␤»
04:05 preflex joined #perl6
04:12 TimToady std: $*NONESUCH
04:12 camelia std 8370f3f: OUTPUT«ok 00:01 121m␤»
04:20 ssutch joined #perl6
04:41 raydiak 'Foo::Bar'.eval.new; # is there a better way than eval to get a type object from a string containing a class name?
04:42 TimToady ::($string) should work
04:43 raydiak beautiful...thanks!
04:53 BenGoldberg r: my $foo = 'Str'; say ::($foo)
04:53 camelia rakudo-parrot 60c8c1, rakudo-jvm 60c8c1: OUTPUT«(Str)␤»
04:54 BenGoldberg r: my $foo = 'Asdf::qwerty'; say ::($foo)
04:54 camelia rakudo-jvm 60c8c1: OUTPUT«No such symbol 'Asdf::qwerty'␤  in block  at /tmp/nhZwoY8yd7:1␤  in any eval at gen/jvm/stage2/NQPHLL.nqp:1086␤  in any evalfiles at gen/jvm/stage2/NQPHLL.nqp:1292␤  in any command_eval at gen/jvm/stage2/NQPHLL.nqp:1196␤  in any command_eval at sr…»
04:54 camelia ..rakudo-parrot 60c8c1: OUTPUT«No such symbol 'Asdf::qwerty'␤  in method gist at gen/parrot/CORE.setting:12004␤  in method gist at gen/parrot/CORE.setting:1014␤  in sub say at gen/parrot/CORE.setting:12904␤  in block  at /tmp/Q13hx1O5Z4:1␤  in any  at /tmp/Q13hx1O5Z4:1␤  i…»
04:56 TimToady r: package Asdf::qwerty {}; my $foo = 'Asdf::qwerty'; say ::($foo)
04:56 camelia rakudo-parrot 60c8c1, rakudo-jvm 60c8c1: OUTPUT«(Asdf::qwerty)␤»
05:01 preflex_ joined #perl6
05:04 adu joined #perl6
05:15 FROGGS joined #perl6
05:21 Rotwang joined #perl6
05:26 FROGGS joined #perl6
05:26 ssutch joined #perl6
05:58 SamuraiJack__ joined #perl6
06:05 darutoko joined #perl6
06:13 FROGGS joined #perl6
06:33 dayangkun joined #perl6
06:47 TimToady http://rosettacode.org/wiki/Metered_concurrency#Perl_6  <-- here's a pretty little thing
06:50 FROGGS joined #perl6
06:51 TimToady (though the site is very slow right now for some reason)
06:52 FROGGS preflex: tell jnthn The cursor-mem-opts does not break any nqp/r-p/r-j tests but there is no performance boost visible except in moar
06:52 preflex Consider it noted.
06:52 FROGGS morning
06:53 kaleem joined #perl6
06:57 rindolf joined #perl6
07:02 TimToady jnthn: would be nice to be able to get a count of queued messages from a channel, then my Semaphor class could report its current usage
07:04 diakopter well, it'd be an estimate
07:04 TimToady sure, there's races
07:04 diakopter an upper bound
07:04 diakopter er.
07:04 diakopter not a bound. :)
07:05 TimToady but for my Semaphore class, $max - $.tickets.count would be the semaphor counter
07:05 diakopter DO. LORD. OH DO. LORD. OH DO.
07:07 TimToady hmm, my spelling isn't consistent...
07:08 hummeleB1 joined #perl6
07:10 brrt joined #perl6
07:36 * lizmat wonders whether such a Semaphore class shouldn't be specced
07:36 lizmat decommute&
07:37 lizmat joined #perl6
07:41 TimToady we don't want to encourage people to suboptimal primitives though...
07:41 jnthn If we built a construct that sat in a loop continually acquiring the semaphore, we could call it semafor...
07:42 jnthn uh, I mean...morning o/
07:42 TimToady er, o/
07:42 * jnthn takes the first few sips of coffee :)
07:42 * diakopter dunks head under coffee
07:43 brrt morning all
07:44 TimToady well, it'll take another 15 minutes here before asteroids become likely again
07:44 brrt whats so suboptimal about a semaphor :-) - early morning brain ca't find a good reason
07:44 TimToady composibility?
07:45 jnthn TimToady: spec what you want the method to be called
07:45 TimToady *sab*
07:45 woolfy left #perl6
07:45 TimToady leave me with the difficult bikesheds, do ya?
07:46 diakopter only the underwater antigrav ones
07:46 * TimToady is tempted to call it .backlog :
07:46 TimToady :)(
07:46 jnthn Well, I can say I'm not sure we should make channels numify to it :P
07:46 TimToady geh, typer is broke
07:46 TimToady no, it's not supposed to be used to write race conditions on purpose :)
07:46 TimToady but it'd be nice to know if our channel is keeping up, at any rate
07:47 TimToady and how close we are to a max buffer size, if we implement one
07:48 FROGGS joined #perl6
07:48 jnthn Yes, I also want to have a way to limit the buffer size.
07:49 jnthn Right now it's unbouned, which gives no back-preasure...
07:49 * TimToady hasn't really thought of anything better than .count or .elems...maybe .pending?
07:51 TimToady if a channel tracked a max, then my Semaphore class wouldn't have to remeber it to return .roomleft, or whatever the .max - .pending is
07:52 TimToady .elems would make sense to someone who thinks of a channel like an array that's getting pushed
07:53 jnthn well, underneath a channel is just a lock-free queue...
07:53 TimToady does the JVM keep track of the size of it for you?
07:53 jnthn but just 'cus it *is* that doesn't mean we want people to think of it like that...
07:53 jnthn TimToady: Believe so...lemme check...
07:54 jnthn TimToady: yes, but with Big Warning:
07:55 TimToady .oO(Don't use this to write race conditions!)
07:56 jnthn "Beware that, unlike in most collections, this method is NOT a constant-time operation. Because of the asynchronous nature of these queues, determining the current number of elements requires an O(n) traversal. Additionally, if elements are added or removed during execution of this method, the returned result may be inaccurate. Thus, this method is typically not very useful in concurrent applications."
07:56 jnthn sorry, some guy wandered in with a thermometer to see how warm the room was... :)
07:56 TimToady oh, so it's actually going out and counting, hmm
07:57 jnthn Yeah
07:57 jnthn We can maintain it by keeping an atomic integer in the channel too...
07:57 TimToady so if we want it, we do our out atomic ++/--, and call it .approx so we don't have to make it atomic along with the queue ops
07:57 jnthn But then of course everything pays for it whether we ever want the count or not
07:58 TimToady *our own
07:58 TimToady right
07:58 jnthn yes, the atmoic I mentioned is so ++ and -- themselves on it are not gonna lose count
07:58 TimToady but we have to track it if we ever want to enforce a limit
07:59 jnthn Well, the JVM provides other queue types that can be bounded
07:59 jnthn BUT it needs to knwo the size up-front then.
08:00 jnthn So we could implement our bounded one in those terms
08:00 TimToady do Go's channels provide a count?
08:01 TimToady yes, len(channel), apparently
08:02 jnthn yes, and it seems they default to zero-length buffer
08:02 jnthn (meaning that a send always blocks unless a receiver is already there waiting)
08:03 adu o hi jnthn
08:03 TimToady oh, well, that wouldn't work for the way I use it :)
08:03 jnthn well, you'd have to specify a size if you want async is all...
08:03 moritz \o
08:03 preflex moritz: you have 1 new message. '/msg preflex messages' to read it.
08:03 jnthn There's (perhaps wisely) no unbounded.
08:03 jnthn I do wonder if unbounded is a good default
08:03 jnthn bbi5
08:03 TimToady but apparently they have "buffered channels" too
08:05 adu TimToady: afaik, len() only works on buffered channels
08:06 TimToady I think unbounded is a good default if they don't specify a bound :)
08:07 adu TimToady: I'm a big fan of golang
08:07 TimToady though one might want to discourage some threads in the scheduler if they're known to be feeding large channels...
08:08 TimToady adu: it certainly fixes a few of the things that are notably wrong about C  :)
08:09 adu like switch/case with default fallthrough and explicit break
08:09 zakharyas joined #perl6
08:10 TimToady well, was thinking more along the lines of distinguishing pointers from slices of arrays
08:10 adu I much prefer the default break with explicit fallthrough
08:10 adu oh, well, that too
08:14 * jnthn back
08:14 jnthn for a little bit at least :)
08:14 jnthn o/ adu
08:14 adu jnthn: great job on jvm!
08:15 TimToady adu: maybe you should write the Go backend :)
08:15 jnthn adu: Well, it's not just me that's worked on it. :) But happy it's appreciated. :)
08:15 adu TimToady: I wouldn't know where to start
08:16 adu I know the internals of Go quite well (I tried to write a call/cc implementation using 6c, the intermediate compiler between 6g and the assembler), but I'm still quite a newbie at Perl6 internals
08:16 fhelmberger joined #perl6
08:17 TimToady I'm sure jnthn++ would have reams of advice on where to start; my advice comes more in the realm of where to finish.  :)
08:18 adu :)
08:19 moritz "start with nqp" :-)
08:19 adu ok
08:19 moritz it seems that most ports have started by implementing 6model in the host language
08:19 dmol joined #perl6
08:20 moritz then cross compilation from an existing NQP host language to the new target language
08:21 fhelmberger joined #perl6
08:21 * TimToady --> zzz
08:22 jnthn Yeah...
08:22 jnthn Well, really for JVM I started with code-gen for the most basic QAST nodes
08:23 jnthn But you need to start porting 6model pretty soon to handle anything non-trivial :)
08:23 adu why soon?
08:23 jnthn As in, pretty soon after you start porting anything
08:23 jnthn 'cus you can't do anything with objects until you do so
08:27 masak mornin', #perl6
08:28 FROGGS o/
08:28 jnthn ok, teaching time...
08:28 jnthn bbl
08:28 xinming joined #perl6
08:30 jlaire left #perl6
08:33 dalek rakudo/pr/229: 7ee27fe | Arcterus++ | / (36 files):
08:33 dalek rakudo/pr/229: Merge branch 'nom' into sized-arrays
08:33 dalek rakudo/pr/229: review: https://github.com/rakudo/rakudo/commit/7ee27fe222
08:33 dalek rakudo/pr/229: 558080d | Arcterus++ | src/Perl6/Actions.nqp:
08:33 dalek rakudo/pr/229: Fix our arrays
08:33 dalek rakudo/pr/229: review: https://github.com/rakudo/rakudo/commit/558080d92a
08:33 dalek rakudo/pr/229: 48ca3c6 | Arcterus++ | src/core/Array.pm:
08:33 dalek rakudo/pr/229: Fixed typed array regression
08:33 dalek rakudo/pr/229: review: https://github.com/rakudo/rakudo/commit/48ca3c6676
08:33 dalek rakudo/pr/229: 53633ba | Arcterus++ | src/core/Array.pm:
08:33 dalek rakudo/pr/229: Fix .reverse and .rotate regressions
08:33 dalek rakudo/pr/229: review: https://github.com/rakudo/rakudo/commit/53633ba40c
08:33 dalek rakudo/pr/229: 3baa649 | (Tobias Leich)++ | / (3 files):
08:33 dalek rakudo/pr/229: Merge branch 'nom' of github.com:rakudo/rakudo into pr/229
08:33 dalek rakudo/pr/229: review: https://github.com/rakudo/rakudo/commit/3baa6490e6
08:43 salv0 joined #perl6
08:48 dmol joined #perl6
08:53 [Sno] joined #perl6
09:22 masak TimToady++ # http://rosettacode.org/wiki/Metered_concurrency#Perl_6
09:41 ssutch joined #perl6
09:49 daxim joined #perl6
09:55 denisboyun joined #perl6
09:59 aborazmeh joined #perl6
10:15 lizmat joined #perl6
10:31 dakkar joined #perl6
10:40 tgt joined #perl6
10:42 fridim_ joined #perl6
11:36 rindolf joined #perl6
11:39 cognominal__ joined #perl6
11:42 masak why do most languages return only one value, even though they accept 0 or more parameters? why does this feel completely natural?
11:43 timotimo because most languages don't have parameter flattening
11:43 timotimo is my first guess
11:44 hoelzro I'm guessing it has to do with complexity in doing such a thing in C
11:44 hoelzro and other languages following its example
11:45 masak clearly we should have used Forth more as a role model, and C less ;)
11:54 PacoAir joined #perl6
11:58 colomon Forth++
11:59 hoelzro oh man, I need to add rosetta code examples to my Perl 6 "TO READ" list
12:20 masak p: enum Choice <Nay Aye>; say Nay ~~ Choice; say Nay ~~ Aye; say Choice ~~ Choice
12:20 camelia rakudo-parrot 60c8c1: OUTPUT«True␤False␤True␤»
12:20 masak \o/
12:36 xenoterracide joined #perl6
12:51 benabik joined #perl6
12:51 dakkar joined #perl6
12:56 denis_boyun joined #perl6
13:07 berekuk joined #perl6
13:29 dayangkun joined #perl6
13:41 markov joined #perl6
13:52 stevan_ joined #perl6
13:54 raiph joined #perl6
13:57 lowpro30 joined #perl6
13:57 colomon joined #perl6
13:59 kaare_ joined #perl6
14:01 sqirrel joined #perl6
14:02 xinming joined #perl6
14:06 xinming joined #perl6
14:10 lowpro30_ joined #perl6
14:13 kaare_ joined #perl6
14:23 atroxaper joined #perl6
14:23 Psyche^ joined #perl6
14:26 denis_boyun__ joined #perl6
14:27 ajr joined #perl6
14:27 atroxaper Hello #perl6 !
14:27 moritz hello atroxaper, *
14:28 atroxaper hello moritz
14:28 Psyche^_ joined #perl6
14:29 nwc10 good UGT, #perl6
14:30 atroxaper nwc10 o/
14:31 denis_boyun joined #perl6
14:33 Psyche^ joined #perl6
14:38 Psyche^_ joined #perl6
14:48 timotimo o/
14:52 jnap joined #perl6
14:58 raiph joined #perl6
15:01 pdurbin left #perl6
15:01 nnunley joined #perl6
15:08 mls joined #perl6
15:13 jnap1 joined #perl6
15:14 thou joined #perl6
15:20 denis_boyun__ joined #perl6
15:25 Psyche^ joined #perl6
15:28 jnap joined #perl6
15:31 btyler joined #perl6
15:33 lowpro30 joined #perl6
15:39 Psyche^ joined #perl6
15:42 hoelzro all(@x Xeqv @y) checks if every element of @x is equivalent the corresponding element of @y, right?
15:43 FROGGS all @x to all @y
15:43 moritz hoelzro: that would be Zeqv
15:44 hoelzro ok, hang on
15:44 * hoelzro is new to this metaoperator stuff
15:44 FROGGS rp: say [1, 2, 3] Xeqv [1, 2, 7]
15:44 camelia rakudo-parrot 60c8c1: OUTPUT«True False False False True False False False False␤»
15:44 FROGGS rp: say [1, 2, 3] Zeqv [1, 2, 7]
15:44 hoelzro Xeqv would check @x[0] against all of @y
15:44 camelia rakudo-parrot 60c8c1: OUTPUT«True True False␤»
15:44 hoelzro along with @x[1], @x[2], etc, right?
15:44 FROGGS correct
15:44 FROGGS see the putput
15:45 hoelzro but Z would just check @x[i] eqv @x[i]
15:45 moritz puput!
15:45 FROGGS :o)
15:45 moritz *putput
15:45 hoelzro ok, I'm just trying to put it into words =)
15:45 FROGGS hoelzro: right
15:45 * moritz typo'd the typo quote
15:45 FROGGS ha!
15:45 * hoelzro picked "@x Xeqv @y" from S03
15:45 hoelzro odd example
15:45 hoelzro but that's a really cool feature
15:45 hoelzro all(@x Zeqv @y)
15:46 Psyche^_ joined #perl6
15:46 moritz though note that it stops comparing at the shorter array
15:47 moritz @x eqv @y  is basically @x == @y && all(@x Zeqv @y)
15:47 hoelzro so if +@y < @x, it only compares the first +@y elements
15:47 moritz aye
15:48 hoelzro good to know
15:48 FROGGS like a zipper stops at the shorter side
15:48 FROGGS (unless you rip it off)
15:49 timotimo i have a zipper that won't start any more :(
15:56 FROGGS eww :(
15:57 raiph joined #perl6
15:59 kaare_ joined #perl6
15:59 Psyche^ joined #perl6
16:16 jnap joined #perl6
16:17 benabik joined #perl6
16:26 jnap joined #perl6
16:27 jnap joined #perl6
16:28 ajr joined #perl6
16:37 BenGoldberg joined #perl6
16:39 FROGGS[mobile] joined #perl6
16:46 kaleem joined #perl6
16:47 jnap joined #perl6
16:50 colomon joined #perl6
16:52 atroxaper Is there right? https://github.com/perl6/Pod-To-HTML/blob/master/lib/Pod/To/HTML.pm#L17
16:52 atroxaper I mean is it right that we need a good pod tool?
16:53 atroxaper I just want to practice with Perl6. And want to do something useful.
16:55 moritz yes, that's a good starting point
16:56 TimToady Itch, meet Scratch.  Scratch, meet itch.
16:57 atroxaper moritz: ok. I'm in! )
16:57 BenGoldberg http://rosettacode.org/wiki/Reports:Tasks_not_implemented_in_Perl6
16:58 atroxaper moritz: do you know where that pod2html module are used?
17:00 moritz atroxaper: perl6/doc, and by perl6 --doc=html
17:00 atroxaper moritz: ok. I'll see there. Thank you.
17:03 atroxaper see you #perl6
17:03 BenGoldberg Ignore that link I posted, it's actually http://rosettacode.org/wiki/Reports:Tasks_not_implemented_in_Perl_6
17:07 xinming_ joined #perl6
17:08 SamuraiJack__ joined #perl6
17:19 benabik joined #perl6
17:22 xenoterracide joined #perl6
17:24 Rotwang joined #perl6
17:24 Rotwang joined #perl6
17:48 dmol joined #perl6
17:54 lizmat joined #perl6
17:55 ssutch joined #perl6
18:00 * lizmat got to the AmsterdamX meeting in time
18:05 Suhosin joined #perl6
18:05 araujo_ joined #perl6
18:06 lowpro30_ joined #perl6
18:09 lowpro30 joined #perl6
18:10 slavik joined #perl6
18:12 Entonian joined #perl6
18:14 tgt joined #perl6
18:15 FROGGS joined #perl6
18:18 colomon joined #perl6
18:19 ajr_ joined #perl6
18:19 araujo joined #perl6
18:22 grep0r joined #perl6
18:24 TimToady afk 6 hours or so, see ya in SoCal &
18:24 colomon \o
18:25 masak o/
18:32 [Sno] joined #perl6
18:36 benabik joined #perl6
18:37 raiph joined #perl6
18:47 Esylt joined #perl6
18:48 spider-mario joined #perl6
18:48 PZt joined #perl6
18:53 kurahaupo joined #perl6
18:55 jeffreykegler joined #perl6
18:58 jnap joined #perl6
19:03 sqirrel joined #perl6
19:09 Rotwang joined #perl6
19:11 BenGoldberg joined #perl6
19:14 beastd joined #perl6
19:16 grondilu joined #perl6
19:16 grondilu r: say [+] rand x 5
19:16 camelia rakudo-jvm 60c8c1: OUTPUT«Cannot convert string to number: trailing characters after number in '0.459950550734632540⏏.459950550734632540.459950550734632540.459950550734632540.45995055073463254' (indicated by ⏏)␤  in block  at /tmp/SoAYfjfkLa:1␤  in any eval at gen/jvm/stage2…»
19:16 camelia ..rakudo-parrot 60c8c1: OUTPUT«Cannot convert string to number: trailing characters after number in '0.6182410727228390⏏.6182410727228390.6182410727228390.6182410727228390.618241072722839' (indicated by ⏏)␤  in method gist at gen/parrot/CORE.setting:12004␤  in method gist at g…»
19:16 grondilu n: say [+] rand x 5
19:16 camelia niecza v24-98-g473bd20: OUTPUT«0.4748114237910190.4748114237910190.4748114237910190.4748114237910190.474811423791019␤»
19:16 jnap joined #perl6
19:16 * grondilu is confused
19:17 timotimo you need xx
19:17 grondilu oh
19:17 timotimo x is a string operator
19:17 BenGoldberg rn: say [+] (rand xx 5)
19:17 * grondilu facepalms himself
19:17 timotimo r: say (rand x 5).perl
19:17 camelia rakudo-jvm 60c8c1: OUTPUT«2.569878910978993␤»
19:17 camelia ..niecza v24-98-g473bd20: OUTPUT«2.0931222229698312␤»
19:17 camelia ..rakudo-parrot 60c8c1: OUTPUT«2.09440449332543␤»
19:17 camelia rakudo-jvm 60c8c1: OUTPUT«"0.253500739322205850.253500739322205850.253500739322205850.253500739322205850.25350073932220585"␤»
19:17 camelia ..rakudo-parrot 60c8c1: OUTPUT«"0.4545349572758540.4545349572758540.4545349572758540.4545349572758540.454534957275854"␤»
19:17 BenGoldberg rn: say (rand xx 5).[+]
19:17 * grondilu hasn't been writing perl6 code for a while, thus this silly mistake
19:17 camelia niecza v24-98-g473bd20: OUTPUT«[31m===[0mSORRY![31m===[0m��Prefix requires an argument at /tmp/BBDyNA5S49 line 1:�------> [32msay (rand xx 5).[+[33m�[31m][0m��Parse failed��»
19:17 camelia ..rakudo-jvm 60c8c1: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/ueRvz1azY3�Preceding context expects a term, but found infix + instead�at /tmp/ueRvz1azY3:1�------> [32msay (rand xx 5).[+[33m�[31m][0m�»
19:17 camelia ..rakudo-parrot 60c8c1: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/Zl02BC8a_H�Preceding context expects a term, but found infix + instead�at /tmp/Zl02BC8a_H:1�------> [32msay (rand xx 5).[+[33m�[31m][0m�»
19:18 BenGoldberg rn: rand xx 5 >> [+] >> say
19:18 camelia rakudo-jvm 60c8c1: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/MT8lifSOW0â�¤Unsupported use of >> to do right shift;  in Perl 6 please use +> or ~>â�¤at /tmp/MT8lifSOW0:1â�¤------> [32mrand xx 5 >>[33mâ��[31m [+] >> say[0mâ�¤Â»
19:18 camelia ..rakudo-parrot 60c8c1: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/sD7HpHQiRpâ�¤Unsupported use of >> to do right shift;  in Perl 6 please use +> or ~>â�¤at /tmp/sD7HpHQiRp:1â�¤------> [32mrand xx 5 >>[33mâ��[31m [+] >> say[0mâ�¤Â»
19:18 camelia ..niecza v24-98-g473bd20: OUTPUT«[31m===[0mSORRY![31m===[0m��Unsupported use of >> to do right shift; in Perl 6 please use +> or ~> at /tmp/quxDHoTkgA line 1:�------> [32mrand xx 5 >>[33m�[31m [+] >> say[0m��Unsupported use of >> to do right shift; in Perl 6 please…»
19:23 geekosaur if you wanted a hyper, no spaces
19:23 geekosaur >>op>>
19:23 colomon but >>op>> say probably will not do what you want
19:24 colomon nor will >>[op]>>, I think
19:24 Rotwang joined #perl6
19:25 grondilu So, MoarVM.  Any progress?
19:27 timotimo today has been mostly debugging a tricky failure or gc bug or something
19:28 timotimo a day ago, jnthn fixed a non-optimal memory usage pattern in the grammar engine that caused about 1/3rd of the time for stage parse in rakudo's setting to disappear
19:28 timotimo FROGGS has measured the stage parse time for parrot and moarvm to be almost exactly the same with an optimized moarvm build on windows, but on linux it would crash due to the mystery bug
19:29 timotimo but we do make it all the way through stage parse by now, which is really nice
19:29 timotimo i don't know when you last checked in
19:29 grondilu ok, thanks for the heads up
19:29 FROGGS timotimo: stage parse using an optimized moar takes 176s on my linux, where parrot takes 120s
19:29 * grondilu is glad to see that the team is still working hard and hasn't given up :-)
19:29 timotimo oh, oops
19:30 timotimo i'm not working hard :\
19:30 colomon timotimo: I didn't see a noticeable improvement overnight in the parrot stage parse on my smoke test.
19:30 FROGGS that is by compiling BOOSTRAP.nqp using an unoptimized moar (aka cheating)
19:30 PZt joined #perl6
19:30 grondilu talking about memory, today I compiled rakudo-star-2013.11 and I noticed it used much more RAM.  I had to activate swap (I have 1Go RAM).
19:31 FROGGS hmmm, that is unexpected I'd say
19:31 slavik \o/
19:31 FROGGS the memory usage should drop
19:31 FROGGS hi slavik
19:31 slavik hi all
19:31 slavik yesterday was awesome ^_^
19:32 FROGGS slavik: in what sense?
19:32 slavik starcraft and booze :)
19:32 Rotwang joined #perl6
19:32 FROGGS :o)
19:32 slavik I also got to meet Mike Morhaime
19:32 slavik co-founder/ceo of blizzard
19:33 slavik so, it was an awesome day :)
19:33 FROGGS yeah
19:33 FROGGS that is kinda awesome :o)
19:33 slavik now, I am reading rakudo* updates, so today is going to be a very good, day, too.
19:33 slavik well ... I have yet to meet TimToady :)
19:34 FROGGS yes you have
19:34 timotimo colomon: yeah, that improvement only helped on moar.
19:34 FROGGS he's a genius (and he is a very nice person)
19:35 colomon timotimo: ah
19:35 slavik FROGGS, I mean in person. I have not made it out to any of the YAPC :(
19:35 colomon timotimo: didn't realize moarakudo was that far along
19:36 FROGGS slavik: yes, I mean in person too
19:36 slavik wait, "yes you have", was that an answer to my question?
19:36 slavik or rather, statement
19:37 FROGGS err, "you have to"
19:37 FROGGS I mean, you need to meet him :o)
19:38 slavik haha
19:38 slavik agreed
19:39 Rotwang hi
19:39 FROGGS hi Rotwang
19:41 * PerlJam wakes up enough to remind everyone about the advent calendar
19:42 * lizmat further decommutes
19:52 arnsholt The test failures on NQP/JVM, do they happen to everyone, or only a subset of people?
19:53 moritz seems to be everyone
19:53 arnsholt Oh, good (sort of)
19:53 arnsholt In that case, any objections to marking them as TODO?
19:58 moritz if you also open tickets, no problem (IMHO)
19:59 arnsholt Good point
20:02 rurban1 joined #perl6
20:06 arnsholt jnthn: printfh on JVM returning String is just wrong, right?
20:06 lue Anyone know why panda can't compile Shell::Command for me during bootstrap?
20:07 jnthn arnsholt: Just about everything returns something in case it shows up in rvalue context...
20:07 preflex jnthn: you have 3 new messages. '/msg preflex messages' to read them.
20:08 arnsholt jnthn: Yeah, but it looks like the tests expect an integer (number of bytes written)
20:08 FROGGS nqp: my $fh := nqp::getstdout; say(nqp::printfh($fh, "olá"));
20:08 camelia nqp-moarvm, nqp-parrot: OUTPUT«olá4␤»
20:08 camelia ..nqp-jvm: OUTPUT«oláolá␤»
20:08 arnsholt Yeah, what FROGGS said. Parrot returns bytes, whereas JVM returns the string printed
20:09 FROGGS (actually bytes written)++
20:09 jnthn arnsholt: Oh...
20:09 jnthn The Parrot answer is probably usefuller :P
20:09 arnsholt Yeah. Also, kills a failing test
20:10 FROGGS ohh, nice catch
20:10 arnsholt The JVM test failures mean I can't make test install
20:10 FROGGS :/
20:10 arnsholt And now I just had to do something about it =)
20:11 FROGGS I think I owe you a beer then :o)
20:11 FROGGS damn, I don't wanna hack on S11+panda :(
20:12 arnsholt If it's any consolation, this is instead of NativeCall hackery =)
20:15 lue the weird thing is, I was able to manually compile Command.pm, so it must be a problem with how panda's doing it.
20:16 moritz lue: did you check the shell exit status?
20:16 lue no, let me run rebootstrap again
20:17 xinming joined #perl6
20:20 FROGGS O.o
20:20 FROGGS timotimo: that is you! http://www.youtube.com/watch?v=SMPMFfqAeS8
20:21 diakopter on a plane!
20:23 btyler joined #perl6
20:23 arnsholt "2 TODO tests unexpectedly succeeded" =D
20:24 arnsholt Looks like the return value was a problem elsewhere as well \o/
20:25 FROGGS cool!
20:25 FROGGS arnsholt++
20:25 jnthn arnsholt++ # nice
20:25 lue moritz: I get a $? of zero :/ (also, panda now refuses to compile File::Find somehow. I'll clear out the panda repo and ~/.perl6 and try again)
20:26 FROGGS lue: can paste then what happens?
20:26 lue FROGGS: sure, I will.
20:27 jnap joined #perl6
20:27 moritz lue: I surely hope ~/.perl6 isn't used anymore by contemporary panda
20:28 lue moritz: that's where it installed to. I also happen to like ~/.perl6 :)
20:28 FROGGS moritz: it is, but suffixed by the compiler version
20:28 dalek nqp: 0713cbb | (Arne Skjærholt)++ | src/vm/jvm/ (5 files):
20:28 dalek nqp: Make nqp::{write,print,say}fh on JVM return number of bytes written.
20:28 dalek nqp:
20:28 dalek nqp: This brings NQP/JVM in line with the Parrot and Moar backends, and fixes three
20:28 dalek nqp: failing tests in t/nqp/19-file-ops.t
20:28 dalek nqp: review: https://github.com/perl6/nqp/commit/0713cbbe11
20:28 FROGGS rp: say %*CUSTOM_LIB<home>
20:28 camelia rakudo-parrot 60c8c1: OUTPUT«/home/p6eval/.perl6/2013.11-15-g60c8c1a␤»
20:28 lue [ and "HEAD is now at 41a73f3 Merge pull request #60 from FROGGS/master" is where I am, so it seems to be a pretty new panda :) ]
20:28 moritz shouldn't install to <site> by default?
20:28 moritz *it
20:29 FROGGS it tries that, and falls back to home is site is not writeable
20:29 moritz ah.
20:31 lue And since my site is in /usr/local, I'd need to use sudo all the time to install modules.
20:31 * moritz thinks he even implemented that, once upon a time
20:34 lue And now Shell::Command went perfectly. I'm almost willing to blame rebootstrap specific shenanigans.
20:34 FROGGS that might be a good bet
20:35 lue (because before IIRC I would always run rebootstrap, because re- has the helpful "no need for rebootstrap, going to normal bootstrap" behavior, meaning more laziness on my part :D)
20:36 eternaleye joined #perl6
20:38 lue roar. Now JSON::Tiny is the failer: https://gist.github.com/lue/1522bddcd64e62669748
20:39 lue moritz: is there some sort of --verbose command for (re)bootstrap that I can use to see what commands exactly are being used?
20:40 FROGGS hmmm, panda should print the command it has invoked in case it failed
20:40 colomon joined #perl6
20:41 lue FROGGS: nope, it just moves on. The only reason panda tries stuff again is because the next item on the list "requires" it, and thus it must be compiled. Which will fail as before.
20:41 FROGGS no, that was a mumbled feature request
20:42 lue ah :)
20:44 Rotwang did anyone notice that make install -j<more than one> fails with 2013.11?
20:44 Rotwang or is it just me
20:46 lue and of course I can compile the offending JSON::Tiny file just fine. I looked, and it seems panda's running the exact same command I am. blergh :/
20:46 Rotwang $ /usr/bin/perl5.16.3 -MExtUtils::Command -e cp  does_not_exist does/not/exist/as/well
20:46 Rotwang $ echo $?
20:46 Rotwang 0
20:47 Rotwang so basically there are two problems, 1) makefile is bad, 2) ExtUtils::Command cp is proken
20:47 Rotwang broken*
20:48 xenoterracide joined #perl6
20:49 Rotwang I guess I should file a bug report then
20:50 lue Rotwang: yes :) Then we'd know what specific error occurs upon use of -j
20:51 Rotwang lue: so is it enough that I send email to a rakudobug@perl.org ?
20:52 lue Rotwang: yes.
20:53 Rotwang I also get "Possible cross-site request forgery" while trying to access: http://rakudo.org/rt/open-all
20:53 Rotwang just informing
20:54 * lue doesn't know why, but he feels as though he's been unable to use panda for quite a long while now :(
20:55 lue Rotwang: go to rt.perl.org instead.
20:55 Rotwang lue: I know, just letting you guys know, so someone who maintains rakudo.org may fix it someday
20:56 lue OK. I just wasn't sure if you knew or not.
21:01 xinming joined #perl6
21:04 FROGGS joined #perl6
21:07 lue Trying again still fails JSON::Tiny :(
21:09 FROGGS timotimo++ # http://www.youtube.com/watch?v=SMPMFfqAeS8
21:09 FROGGS lue: so you can install it manually but not as a dep?
21:10 lue FROGGS: no, whenever I say "I can compile it manually", I mean I run "perl6 --target=pir ..."
21:10 FROGGS and that gives exit code 0 ?
21:11 lue FROGGS: yep. Nothing I've run yet has given me anything else. (Although I have not tried a manual use of 'panda' yet, admittedly)
21:12 * lue is running rebootstrap to see if the failure point changes
21:13 FROGGS I'll install rakudo@parrot and try to install JSON::Tiny
21:14 lue File::Find is what fails on rebootstrap :/
21:14 FROGGS any good error message?
21:15 lue FROGGS: nope. I'm going to regain a partial installation via bootstrap, and then try manually installing (via panda) JSON::Tiny
21:15 lue (assuming JSON::Tiny is where it will fail this time, that is.)
21:21 FROGGS build stage failed for JSON::Tiny: Failed building lib/JSON/Tiny/Actions.pm
21:21 FROGGS :o(
21:23 lue FROGGS: I still say it's panda's fault though, on account of the moving failure point.
21:26 FROGGS umm, using bootstrap.pl instead of reboostrap.pl did it for me
21:27 lue o.o
21:27 FROGGS Compiling lib/JSON/Tiny/Actions.pm to pir
21:27 FROGGS Compiling lib/JSON/Tiny/Grammar.pm to pir
21:27 FROGGS Compiling lib/JSON/Tiny.pm to pir
21:27 FROGGS ==> Testing JSON::Tiny
21:27 masak 'night, #perl6
21:27 FROGGS gnight masak
21:28 lue *gurgle*
21:30 ajr_ Is there some reason that 2013.11 changed the name of perl6 to perl6-p? It broke my download-compile-and set-links script.
21:31 stevan_ joined #perl6
21:31 jnthn I think you should got a perl6 that points to it also?
21:31 ajr_ BTW if such a script would be useful to anyone else, I'd be happy to share it
21:31 lue ajr_ it still installs as perl6 for me (in addition to perl6-p)
21:32 FROGGS ajr_: if you configure for jvm and parrot, it will install the first mentioned as perl6
21:32 lue buh buh buh buh buh... running bin/panda manually works! But.. but... gaaaah
21:32 FROGGS (in addition to perl6-j and perl6-p)
21:33 lue so now I guess I manually copy over bin/panda to the location...? (or the entirety of bin/ ?)
21:33 rjbs Shouldn't perl6-p just return true if the input is perl 6?
21:34 FROGGS rjbs: it should exit with zero if it executed it successfully
21:34 ajr_ my config line is perl Configure.pl --gen-parrot --gen-nqp
21:34 ajr_
21:34 lue FROGGS: what does a newly-installed panda have in its bin directory? just panda or a bunch of other files too?
21:34 rjbs FROGGS: (Lisp joke.)
21:34 FROGGS ajr_: that should give you a perl6 binary... how do you invoke make?
21:35 FROGGS rjbs: I am not a Lisper :o)
21:35 FROGGS lue: others too
21:35 rjbs FROGGS: functions ending in p are predicates
21:35 arnsholt rjbs: Not sure I would've gotten that one without the Lisp joke hint, but now I'm smiling =)
21:35 ajr_ make on the next line, then make install after that
21:35 FROGGS ajr_: that is al right
21:36 FROGGS ajr_: maybe it is a problem when --backends is not passed...
21:37 ajr_ The same script works on 2013.10
21:37 lue oh right, there's still installing Panda... :|
21:37 FROGGS ajr_: I am trying it on my box...
21:40 FROGGS ajr_: after configuring, does it print that it found both nqp-j and nqp-p?
21:40 FROGGS ajr_: that happened on my box, and since it checks for nqp-j first, this is the implicit candidate for perl6
21:41 FROGGS perl Configure.pl --gen-parrot --gen-nqp --prefix=../nqp/install
21:41 FROGGS Found /home/froggs/dev/rakudo/../nqp/install/bin/nqp-j (backend jvm)
21:41 FROGGS Found /home/froggs/dev/rakudo/../nqp/install/bin/nqp-p (backend parrot)
21:41 FROGGS like that
21:43 pecastro joined #perl6
21:46 timotimo FROGGS: did you like the talk? >_<
21:46 FROGGS yes, I did :o)
21:46 ajr_ I don't have the output of the 2013.11 job at hand now. The script is at http://pastebin.com/180Znipd
21:47 lue ==> Fetching panda
21:47 lue Start of substr out of range. Is: 2, should be in 0..1
21:48 lue I... I just don't know.
21:48 FROGGS ajr_: you must have an nqp in path somewhere, otherwise you would get:
21:48 FROGGS perl Configure.pl --gen-parrot --gen-nqp
21:48 FROGGS No suitable nqp executables found! Please specify some --backends, or a --prefix that contains nqp-{p,j} executables
21:48 ajr_ What's the change between 2013.10 and 2013.11 that makes that necessary?
21:49 timotimo FROGGS: i only finished the slides a minute before the talk started :|
21:50 FROGGS timotimo: that is how it is meant to be :P
21:50 lue Does anyone know why the "fetch" step failed? ... Oh, I can guess.
21:50 FROGGS lue: that must be: panda/lib/Panda/Fetcher.pm:28:        my $d = $_.directory.substr($from.chars);
21:50 jnthn timotimo: That means you're on the path to being a pro conf speaker! :)
21:50 dalek nqp: 44d4c1d | (Arne Skjærholt)++ | t/nqp/19-file-ops.t:
21:50 dalek nqp: Mark broken nqp::readlinefh tests as TODO.
21:50 dalek nqp:
21:50 dalek nqp: Now t/nqp/19-file-ops.t passes cleanly through the test harness.
21:50 dalek nqp: review: https://github.com/perl6/nqp/commit/44d4c1dcbd
21:50 FROGGS ajr_: the multi backend refactoring
21:51 FROGGS ajr_: just pass --backends=jvm,parrot or only one of the two
21:52 lue FROGGS: I get the feeling what happened is that ./ is a failure, but ../panda/ isn't. I'll let you know in a minute if it helps.
21:52 FROGGS arnsholt: I had thought the condition should go after the todo(), no?
21:52 timotimo jnthn: i spent way more time making the framework for the presentation, though
21:53 FROGGS lue: "./" sounds about right (being the thing that kills it)
21:53 arnsholt FROGGS: Whoops! That's a pasto, indeed!
21:54 FROGGS :o)
21:54 ajr_ Does that mean that http://rakudo.org/how-to-get-rakudo/ is no longer correct?
21:55 * lue still would like to know how a fetch step takes so long with local directories
21:55 FROGGS ajr_: ohh yes, that should mention the --backends option...
21:55 FROGGS moritz: are you able to fix http://rakudo.org/how-to-get-rakudo/ about --bakends?
21:57 FROGGS lue: it takes about six seconds on my box to parse the ecosystem's projects.json (which is 35kB)
21:57 lue bah, still on the 'fetch' step. [ In case you can't tell, I'm getting a little annoyed with panda about now. I just want one stinking module... ]
21:58 FROGGS :/
21:58 FROGGS I'd wipe site/ at some point... and start afresh (in case you install to site)
21:59 raiph joined #perl6
21:59 lue .oO(One of these days, I'll go into panda, and finally stop it from being so dang awkward to use all the time...)
22:00 lue FROGGS: I don't think starting this over again will help matters :)
22:00 dalek nqp: 2a89f4e | (Arne Skjærholt)++ | t/jvm/02-pipes.t:
22:00 dalek nqp: Relax an overly zealous string check a bit.
22:00 dalek nqp:
22:00 dalek nqp: This fixes a spurious failure in t/jvm/02-pipes.t.
22:00 dalek nqp: review: https://github.com/perl6/nqp/commit/2a89f4ee28
22:00 dalek nqp: 7f7d915 | (Arne Skjærholt)++ | t/nqp/19-file-ops.t:
22:00 dalek nqp: Fix braino. FROGGS++ for spotting it.
22:00 dalek nqp: review: https://github.com/perl6/nqp/commit/7f7d9150cf
22:00 lue >:(    ==> Fetching panda
22:00 lue Start of substr out of range. Is: 9, should be in 0..8
22:02 lue (this is getting utterly idiotic by now)
22:03 Rotwang just created my first rakudobug evar \:D/
22:04 lue \o/ Rotwang++ # and may there be many more you file :)
22:04 * timotimo also watches his talk now
22:04 timotimo i didn't do half bad i think
22:04 timotimo except the text really isn't readable on the video recording
22:04 timotimo maybe i should have recorded the screen as well and built a composite video and upload it to youtube
22:04 timotimo or with youtube-doubler
22:07 * lue will go and consider either fixing panda or building an alternative to it at some point in the future. *grumble grumble* &
22:08 FROGGS ahh
22:15 FROGGS rakudo$ perl6-p -MFile::Find -e 'my $dir = "tools"; say find(:$dir).list>>.directory>>.substr($dir.chars)'
22:15 FROGGS /lib /distro /build /build /build /build /build /build /build /build /build /util /lib/NQP
22:15 FROGGS rakudo$ perl6-p -MFile::Find -e 'my $dir = "./tools"; say find(:$dir).list>>.directory>>.substr($dir.chars)'
22:15 FROGGS Start of substr out of range. Is: 7, should be in 0..5
22:15 FROGGS lue: ^^ here is your bug
22:16 FROGGS that is because listing a directory sort of normalizes their path
22:16 japhb rjbs: re: http://irclog.perlgeek.de/perl6/2013-11-25#i_7913782 , I once explained Lisp predicates to a non-programmer friend of mine.  She began inviting me lunch by IM'ing 'foodp'.  My reaction pretty much amounted to \o/
22:17 rjbs foodp is like classic hacker jargon, isn't it?
22:17 rjbs we used to say that a lot
22:17 japhb She somehow got it from first principles.
22:17 rjbs proving that lisp is truly intuitive
22:17 japhb Which clearly indicates her head was screwed on properly, methinks.
22:17 rjbs Lisp:  You can even understand it when hungry.
22:18 jnthn .oO( Lis? )
22:18 japhb .oO( Logo: LISP for people who hate all the parens )
22:18 huf itth not that hard to underthand
22:19 colomon joined #perl6
22:23 FROGGS lue: patch: https://github.com/tadzik/panda/pull/62
22:23 FROGGS tadzik: ^^
22:30 timotimo # We need to cleanup the path, because the returned elems are too.   -  are too what?
22:30 timotimo ah. are cleaned up, too
22:30 FROGGS err, yes :/
22:34 bbkr_ has anyone used Storm - http://storm-project.net/ before? Perl 6 is mature enough to enter world of distributed computation and I'm thinking about porting IO::Storm bindings.
22:36 jnap joined #perl6
22:39 japhb WRITE ALL THE BINDINGS   o/
22:41 * japhb is working this one layer down -- writing the wire-protocol code (in my case, for protocol buffers)
22:41 lue Start of substr out of range. Is: 6, should be in 0..5
22:41 lue FROGGS: I assume I need your patch to make this work at all, correct?
22:41 lue (that happened one dir above panda btw)
22:42 timotimo ah, storm is doing protobuf? that sounds good
22:44 dalek rakudo/moar-support: eb14f79 | jonathan++ | src/ (2 files):
22:44 dalek rakudo/moar-support: Stub p6trialbind on Moar (always returns "unsure")
22:44 dalek rakudo/moar-support: review: https://github.com/rakudo/rakudo/commit/eb14f79bc7
22:46 rurban1 joined #perl6
22:46 Tene japhb: What are you using protobuf for?
22:46 japhb timotimo, Incorrect interpretation.  I meant that I am addressing the general "bind all the things" problem by writing wire-protocol code, while others are writing service-level code.
22:46 timotimo ah
22:46 japhb Tene, nothing quite yet, since I can write them a lot better than I can read them at the moment.  ;-)
22:46 Tene Heh heh heh.
22:47 * lue gives up on panda for the time being and just downloads/installs the stupid module he wanted himself.
22:47 Tene capnproto looks like an appealing protobuf successor, although I've never use it yet.
22:47 japhb Eventually, we should be able to integrate with anything that speaks protobuf (which is basically everything Google, and a lot of other projects that adopted it besides)
22:49 rurban2 joined #perl6
22:50 japhb I've seen a number of protobuf "replacements".  My general view (not specific to Cap'n Proto, since this is new to me), is: 1) as much hype as substance, and 2) Generally are optimized for a subtly different problem space than protobufs.  Now mind you, this one claims to be written by the author of protobuf 2, so it ought to be interesting to read his docs.
22:51 japhb FROGGS: You shouldn't need to say .IO.path (unless that's a Niecza compatibility thing)
22:53 * lue wishes we wouldn't call them protobufs. Sounds too much like a Perl 6 thing. :)
22:53 japhb You suggest instead?
22:54 rurban2 the google protobuf library?
22:54 japhb rurban2: Precise, but wordy.
22:55 japhb (And actually a little inaccurate, since I was speaking about the protocol itself, not the library that implements it at runtime.)
22:56 lue probuf, prtbfr, protbfs, etc. Just not protobuf, because until *just now* I always thought it was some obscure low-level Perl6 type :)
22:57 rurban2 the protobuf serialization format
22:57 lue PBSF :)
22:58 lue (although it could conceivably be misconstrued as referring to rakudo's code serialization...)
22:59 japhb rurban2: Definitely more accurate.  Still wordy, and lue's PBSF is too short, methinks.
23:00 japhb I don't like referring to them as "Google protobufs" because that implies that they can *only* be used with Google services (a common misperception I didn't want to continue).  On the other hand, they *were* invinted here.
23:00 japhb *invented
23:00 timotimo spacex is apparently about to re-start the 13 minutes count-down for their geosync satellite launch if i understand correctly
23:00 lue protlbufs perhaps
23:01 fridim__ joined #perl6
23:01 japhb heh
23:01 lue or protobuf -> p6f . Oh wait, nevermind, that'll be confusing when the FORTRAN backend comes for Rakudo...
23:02 japhb .oO( ".proto bufs" )
23:03 japhb PBs?
23:03 japhb Kinda jargony
23:04 timotimo peanutbuffer and jelly?
23:04 lue Come to think of it, I like how protlbufs sounds in my head...
23:04 japhb .oO( It just always comes down to language, doesn't it?
23:05 japhb "peanutbuffer"  :-D
23:05 rurban2 google is the creator, and it's hosted on code.google.com
23:05 japhb Very nice, timotimo
23:05 japhb rurban2: Where else would you expect them to host it?
23:05 timotimo codeplex
23:06 japhb :-P
23:06 japhb (and LOL)
23:07 rurban2 well, someone else could have taken over (happened before) and then it would be hosted elsewhere. but it's still at google
23:07 pecastro_ joined #perl6
23:09 lue .oO("protocolbuffer" would shorten to p14r ...)
23:09 japhb Sure.  It's the public fork of an active internal project, as opposed to being tossed over the wall and "abandoned to the community".  Hence, hosted at the mothership.  Still, I get your point rurban2.  Any idea of a better way to indicate that while Google shepherds it, in actual use it's not Google-specific?
23:10 geekosaur which someone will manage to typo as p1r
23:11 lue geekosaur: i18n and l10n seem to manage :)
23:11 japhb (Just in case someone is wondering, I have *no* inside knowledge of the future roadmap for Protocol Buffers; I'm specifically doing my implementation based entirely on the external docs.)
23:12 rurban2 they just put out libprotobuf8 this year
23:14 rurban2 but the format itself seems to be stable since a few years
23:15 lue Net::IRC's eight-ball isn't very nice...
23:35 ajr_ rakudo* 2013.11 is still generating perl6-p instead of perl6, even with perl Configure.pl --gen-parrot --gen-nqp --backends=parrot
23:36 colomon joined #perl6
23:36 timotimo doesn't it have to be --backend=parrot?
23:36 prevost joined #perl6
23:37 ajr_ see FROGGS at 16:56:58
23:37 timotimo 16:56 < FROGGS> eww :(
23:40 btyler joined #perl6
23:40 baest joined #perl6
23:41 ajr_ SpaceX have had to halt the countdown at T-00:03:40, if anybody's following that.
23:42 Suhosin joined #perl6
23:43 kurahaupo_ joined #perl6
23:44 lizmat joined #perl6
23:45 * lizmat has finally decommuted, but is very tired
23:45 timotimo yeah, they've stopped it for today and will continue on thursday
23:46 colomon any word on what happened?
23:47 timotimo flight computer didn't boot i think
23:47 timotimo how do i bind an op for moarvm again? :|
23:49 timotimo i know there's some tool to do it, but i forget how ...
23:50 Celelaptop joined #perl6
23:51 * timotimo has to build a perl6 first
23:51 d4l3k_ joined #perl6
23:51 * timotimo has a backup in the benchmark components
23:51 Vlavv` joined #perl6
23:52 masak joined #perl6
23:53 * timotimo is not sure where to put the mapping from "MVM_OP_replace" to "MVM_str_replace"
23:53 timotimo actually MVM_string_replace
23:55 timotimo huh, there was a commented out op for it already
23:55 Celelaptop joined #perl6
23:56 woolfy joined #perl6
23:56 pecastro joined #perl6
23:58 * timotimo has a little stubby but probably working implementation
23:59 rurban1 joined #perl6
23:59 stevan_ joined #perl6
23:59 * lizmat goes to get some sleep
23:59 timotimo sleepwalking?

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

Perl 6 | Reference Documentation | Rakudo