Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2014-03-20

Perl 6 | Reference Documentation | Rakudo

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

All times shown according to UTC.

Time Nick Message
00:20 lizmat someone in my family had her internet disconnected: turned out someone in another street at the same number hadn't paid the bill
00:20 lizmat and the drone responsible for disconnecting people, simply had taken the wrong street
00:21 lizmat How Brazil is that for you...
00:25 japhb__ moritz: Regarding http://irclog.perlgeek.de/perl6/2014-03-19#i_8463201 , see http://en.wikipedia.org/wiki/Transdichotomous_model .  So brrt's intuitive sense of integer arithmetic as constant-time is true under the transdichotomous model ... which is to say, native machine words of problem-appropriate size.
00:26 japhb__ lizmat: Yeah, that's pretty damn common too.  :-(
00:29 rurban1 joined #perl6
00:33 dayangkun joined #perl6
00:39 lizmat sleep&
01:00 rurban1 joined #perl6
01:04 hoverboard joined #perl6
01:10 BenGoldberg r: sub infix:<oof> is assoc('list') { $^a + $^b }; say [Roof] 1..4
01:10 camelia rakudo-parrot 01d16c: OUTPUT«Too many positional parameters passed; got 4 but expected 2␤  in block  at gen/parrot/CORE.setting:17009␤  in sub  at gen/parrot/CORE.setting:17151␤  in block  at /tmp/tmpfile:1␤␤»
01:10 camelia ..rakudo-moar 01d16c: OUTPUT«Too many positional parameters passed; got 4 but expected 2␤  in block  at src/gen/m-CORE.setting:16808␤  in sub  at src/gen/m-CORE.setting:16950␤  in block  at /tmp/tmpfile:1␤␤»
01:10 camelia ..rakudo-jvm 01d16c: OUTPUT«Too many positional parameters passed; got 4 but expected 2␤  in block  at gen/jvm/CORE.setting:16754␤  in sub  at gen/jvm/CORE.setting:16896␤  in block  at /tmp/tmpfile:1␤␤»
01:11 BenGoldberg r: sub infix:<oof> is assoc('list') { [+] @_ }; say [Roof] 1..4
01:11 camelia rakudo-parrot 01d16c: OUTPUT«Too many positional parameters passed; got 4 but expected 2␤  in block  at gen/parrot/CORE.setting:17009␤  in sub  at gen/parrot/CORE.setting:17151␤  in block  at /tmp/tmpfile:1␤␤»
01:11 camelia ..rakudo-moar 01d16c: OUTPUT«Too many positional parameters passed; got 4 but expected 2␤  in block  at src/gen/m-CORE.setting:16808␤  in sub  at src/gen/m-CORE.setting:16950␤  in block  at /tmp/tmpfile:1␤␤»
01:11 camelia ..rakudo-jvm 01d16c: OUTPUT«Too many positional parameters passed; got 4 but expected 2␤  in block  at gen/jvm/CORE.setting:16754␤  in sub  at gen/jvm/CORE.setting:16896␤  in block  at /tmp/tmpfile:1␤␤»
01:47 iolympian joined #perl6
01:48 iolympian left #perl6
01:48 telex joined #perl6
02:01 rurban1 joined #perl6
02:03 rurban2 joined #perl6
02:04 colomon joined #perl6
02:27 xragnar_ joined #perl6
02:32 skids .oO(short form "-o <file>" and "-t <level>" options on rakudo eat the arg, but behave like they are not there)
02:52 klapperl joined #perl6
02:53 xinming_ joined #perl6
02:57 Sqirrel joined #perl6
03:03 rurban1 joined #perl6
03:12 redhands joined #perl6
03:13 redhands "a day may come when the efforts of perl6 developers fail, when we forsake our friends and break all bonds of fellowship, but it is not this day"
03:27 sbauer left #perl6
03:37 fgomez joined #perl6
04:03 jnap joined #perl6
04:37 aindilis joined #perl6
04:50 bjz_ joined #perl6
05:03 kaare_ joined #perl6
05:04 jnap joined #perl6
05:06 sunnavy joined #perl6
05:10 woosley joined #perl6
05:24 hoverboard joined #perl6
05:49 kaleem joined #perl6
06:04 jnap joined #perl6
06:06 xenoterracide joined #perl6
06:23 kaare_ joined #perl6
06:28 [Sno] joined #perl6
06:39 kurahaupo joined #perl6
06:55 denis_boyun joined #perl6
07:00 xenoterracide joined #perl6
07:05 jnap joined #perl6
07:11 darutoko joined #perl6
07:17 aborazmeh joined #perl6
07:31 bjz joined #perl6
07:50 kivutar joined #perl6
07:57 FROGGS joined #perl6
07:58 denis_boyun joined #perl6
08:06 jnap joined #perl6
08:10 zakharyas joined #perl6
08:19 IllvilJa joined #perl6
08:31 nwc10 jnthn: plan C involved a pringles can and a step ladder?
08:35 dmol joined #perl6
08:36 dayangkun joined #perl6
08:47 slavik joined #perl6
09:03 arnsholt joined #perl6
09:07 jnap joined #perl6
09:15 slavik joined #perl6
09:18 fhelmberger joined #perl6
09:25 SamuraiJack_ joined #perl6
09:30 kurahaupo joined #perl6
09:35 sqirrel_ joined #perl6
09:52 zakharyas joined #perl6
09:55 salv0 joined #perl6
10:01 dakkar joined #perl6
10:05 iolympian1 joined #perl6
10:07 jnap joined #perl6
10:27 jnthn Morning o/
10:27 jnthn Quiet night
10:27 nwc10 quiet morning too
10:27 nwc10 everywhere
10:28 jnthn nwc10: No, it involved waiting for midnight and hoping whatever magical phone company action happens at that time would fix things.
10:28 nwc10 aha. A pumpkin turned back into a router, or something?
10:28 jnthn Who knows... :)
10:28 jnthn Anyways, it works.
10:28 cosimo joined #perl6
10:29 jnthn Still need to get my fast connection fixed, but this'll do.
10:29 masak ohaj, #perl6
10:29 arnsholt o/
10:30 vendethiel \o
10:30 arnsholt Do we have any docs/similar on how to use the evalserver stuff?
10:31 mls joined #perl6
10:34 IllvilJa Just searched for 'rakudo' in 'Ubuntu Software Center' and look and behold, there was a package!
10:34 tadzik :)
10:35 tadzik col
10:35 tadzik it's 2013.03 in my ubuntu
10:35 IllvilJa Last time I checked (actually years ago but still), there was no rakudo, at least not in Redhat or in Gentoo.  One had to download and install it outside of the package/software manager included in the system.
10:36 nwc10 This is perl6 version 2013.12 built on parrot 5.9.0 revision 0
10:36 IllvilJa That means it is easier to start writing FOSS software in Perl6 and just tell users "install rakudo and get rolling!"
10:36 nwc10 that's on this rather new desktop
10:36 nwc10 arnsholt: did I miss some key line on IRC about what your hopefully-faster branch does differently to be faster?
10:39 arnsholt nwc10: Probably didn't say explicitly
10:39 nwc10 aha. That would explain a lot :-)
10:40 arnsholt For historical reasons (there was a process boundary between the two steps initially), the JAST tree (JVM specific syntax tree) in NQP is dumped to string before being converted to bytecode
10:40 FROGGS ohh dear
10:40 arnsholt Of course, now that the process boundary is gone it makes no sense to dump the tree to strings just to reparse it back to create the bytecode
10:40 FROGGS and I thought that this is a parrotism :o)
10:40 arnsholt The Parrot thing is a bit different, actually, but sort of similar yeah
10:41 FROGGS yeah
10:41 arnsholt Reason is that in the earliest days of the JVM backend, NQP/Parrot did the parsing and generated the JAST tree, while a new NQP/JVM process reparsed the dumped JAST and made the bytecode
10:42 arnsholt So back then it had to be like that. Of course, when the process disappeared, the easiest thing to do was to just glue the two bits together =)
10:42 * FROGGS feels kinda strange to paste output of rakudo to a #perl-* channel, but somehow a good kind of strange
10:42 arnsholt nwc10: Anyways, feel free to give the branch a go
10:43 arnsholt I suspect it'll fail if you try to run Rakudo's testsuite, but I'm not entirely sure what's going on there
10:43 FROGGS arnsholt: does it fail faster? *g*
10:43 FROGGS err, do you see a recognizable improvement for stuff that works?
10:43 arnsholt Haven't looked, really
10:44 arnsholt But it'll mostly be an improvement for Rakudo's setting compile, I think
10:44 arnsholt It's a compile-time thing, not a run-time thing
10:45 FROGGS yes, I see
10:51 tadzik so, LWP+SSL as a gsoc project, yay or nay?
10:51 tadzik (I know the answer, just want to spread my excitement :P)
10:51 FROGGS yay!
10:51 masak yay!
10:51 FROGGS because I need it!
10:52 masak what was the thing we needed for Twitter again?
10:52 tadzik OAuth?
10:52 masak oh, yes.
10:52 tadzik I hate it
10:52 risou_awy joined #perl6
10:52 masak noted :)
10:52 tadzik I don't even want to mentor it :P
10:52 masak haha
10:52 FROGGS hehe
10:52 FROGGS I will help on the LWP+SSL thing as good as I can
10:54 tadzik now I'm thinking about maybe doing the game engine thing as a project myself
10:55 FROGGS tadzik: that is not one of the easiest topics, and I think it is very easy to get this wrong
10:55 tadzik I was planning to do it for myself anyway, but I usually work a lot better when under pressure and deadlines
10:55 tadzik true :|
10:55 FROGGS yeah, me too
10:56 tadzik just in case: would anyone be willing to mentor it?
10:56 aindilis joined #perl6
10:57 bbkr joined #perl6
11:01 FROGGS I'd like to, but masak might be a better option
11:02 nwc10 there are a lot of branches in the NQP repository. How many are dead?
11:07 sorear joined #perl6
11:08 FROGGS nwc10: none of mine (mine is only loop_labels, which is intended to be worked on some day)
11:08 jnap joined #perl6
11:09 nwc10 b
11:11 sqirrel_ joined #perl6
11:24 sergot joined #perl6
11:25 nwc10 arnsholt: most strangely, it fails all but the first Rakudo test run.
11:25 nwc10 So /usr/local/bin/perl t/harness --jvm  t/01-sanity/01-literals.t  t/01-sanity/02-op-math.t
11:25 nwc10 first passes, second fails
11:25 nwc10 /usr/local/bin/perl t/harness --jvm  t/01-sanity/02-op-math.t  t/01-sanity/03-op-logic.t
11:26 nwc10 first passes, second fails, despite the fact that the test t/01-sanity/02-op-math.t failed previously
11:28 jnthn What's it fail with ooc?
11:28 jnthn It's not that the JAST tree nodes are being over-shared, I guess?
11:34 arnsholt That'd be weird. Pretty sure it doesn't keep any references to the JAST being compiled
11:35 jnthn arnsholt: I more meant the types
11:35 arnsholt Oh, that can be it, yeah
11:36 arnsholt It saves away references to the typeobjects on the first compile
11:36 jnthn That'd do it.
11:36 jnthn They're different for different compiles
11:36 jnthn Everything gets its own isolated deserialize, even on eval server, since anything can potentially be augmented, mutated, etc.
11:37 arnsholt Yeah, I didn't think about the eval server
11:37 jnthn It's not only that
11:37 jnthn Oh, it is actually
11:37 jnthn :)
11:37 jnthn Was gonna say something about eval and threads, but of course the threads share it.
11:39 arnsholt Heh
11:40 arnsholt I'll give that a whack tonight, probably
11:40 jnthn \o/
11:40 jnthn I'm hunting a perl6-debug-m SEGV japhb__ found, and then I'll work on cutting a Moar release in time for today's Rakudo release.
11:41 tadzik awesome
11:45 AW3i joined #perl6
12:00 rindolf joined #perl6
12:03 dalek rakudo-debugger: 93e4b51 | jnthn++ | lib/Debugger/UI/CommandLine.pm:
12:03 dalek rakudo-debugger: Fix broken &fail detection in thrown.
12:03 dalek rakudo-debugger: review: https://github.com/jnthn/rakudo-debugger/commit/93e4b517e1
12:09 jnap joined #perl6
12:21 bjz joined #perl6
12:22 bjz_ joined #perl6
12:46 iolympian joined #perl6
12:46 iolympian left #perl6
12:47 dayangkun joined #perl6
13:06 kbaker joined #perl6
13:10 jnap joined #perl6
13:12 denis_boyun joined #perl6
13:17 kaare__ joined #perl6
13:17 jnap joined #perl6
13:20 Lox joined #perl6
13:22 Lox Is the panda build packaged with rakudo known to be a little bit... squirrely?
13:24 tadzik squirrely? :o
13:24 Lox yeah, squirrely
13:24 guru joined #perl6
13:24 Lox as in when i try to run it, it pitches a fit and dies
13:24 tadzik hah
13:24 Lox and it's not a very happy, peaceful death
13:24 tadzik can you paste me the output?
13:24 jnap1 joined #perl6
13:25 Lox it's more like a violent action movie villain death where you know he's probably not really dead
13:25 Lox and yeah sure
13:25 Lox http://prntscr.com/32h7hs
13:27 tadzik hwoa :)
13:27 tadzik yeah, something's seriously wrong
13:27 masak hm, star-2014-01...
13:27 tadzik I think it's the one I released
13:27 tadzik and I'm sure panda worked in it
13:27 masak guess not enough things have changed in Panda since then...
13:28 tadzik https://github.com/tadzik/panda/blob/2014.01/bin/panda#L30 h,m
13:29 tadzik ah, it's here: https://github.com/tadzik/panda/blob/2014.01/lib/Panda/Ecosystem.pm#L33
13:30 tadzik Lox: does 'panda update' do something sensible, and/or help?
13:30 Lox it pitches the same fit
13:31 Lox the exact same fit, actually
13:32 arnsholt joined #perl6
13:33 xenoterracide joined #perl6
13:35 Lox i feel like it's probably an issue in my rakudo install, since i get the same error even if i install panda separately from the git repo (which also pitches a fit, but instead with the File::Find install)
13:48 xenoterracide joined #perl6
13:55 btyler joined #perl6
13:58 Lox joined #perl6
13:59 Lox managed to pull my power strip out of the wall accidentally, so if you had any revelations in the last 5 minutes I didn't see them
13:59 nwc10 Lox: sadly not. But as the topic says, irclog: http://irc.perl6.org
13:59 FROGGS Lox: no, there was no conversation
14:00 Lox bleh, missed that.  thanks
14:03 FROGGS Lox: can you try to a) slurp a file and see if that works, and b) use JSON::Tiny to decode json? (aka call from-json ...)
14:04 Lox give me a sec, i'm rebuilding rakudo from scratch just to see if that helps
14:04 Lox because hey, why not
14:04 FROGGS sure :o)
14:04 xenoterracide joined #perl6
14:14 masak Lox++ FROGGS++ # debugging
14:15 dalek rakudo-star-daily: d2e60b4 | coke++ | log/ (4 files):
14:15 dalek rakudo-star-daily: today (automated commit)
14:15 dalek rakudo-star-daily: review: https://github.com/coke/rakudo-star-daily/commit/d2e60b4d4d
14:15 dalek rakudo-star-daily: a5f3008 | coke++ | log/ (5 files):
14:15 dalek rakudo-star-daily: today (automated commit)
14:15 dalek rakudo-star-daily: review: https://github.com/coke/rakudo-star-daily/commit/a5f30083f5
14:15 dalek perl6-roast-data: 4d0daaf | coke++ | / (4 files):
14:15 dalek perl6-roast-data: today (automated commit)
14:15 dalek perl6-roast-data: review: https://github.com/coke/perl6-roast-data/commit/4d0daaf92e
14:15 dalek perl6-roast-data: a959935 | coke++ | / (6 files):
14:15 dalek perl6-roast-data: today (automated commit)
14:15 dalek perl6-roast-data: review: https://github.com/coke/perl6-roast-data/commit/a9599351e5
14:16 [Coke] no rakudo backend is testing clean.
14:16 FROGGS :/
14:17 [Coke] looks like rakudo.moar was at 0% yesterday.
14:17 FROGGS yeah
14:18 [Coke] build failure, looks like.
14:18 thou joined #perl6
14:19 FROGGS I had no issues
14:20 dalek Perlito: 2feb601 | (Flavio S. Glock)++ | / (3 files):
14:20 dalek Perlito: Perlito5 - grammar - 1st expression can be empty in C-style for
14:20 dalek Perlito: review: https://github.com/fglock/Perlito/commit/2feb601bce
14:21 logie_ joined #perl6
14:23 rindolf joined #perl6
14:25 hoelzro ahoy #perl6
14:25 moritz \o hoelzro, *
14:28 exouniverse joined #perl6
14:30 dayangkun joined #perl6
14:31 dayangkun joined #perl6
14:34 dalek Perlito: ac8d035 | (Flavio S. Glock)++ | / (3 files):
14:34 dalek Perlito: Perlito5 - js runtime - fix test if subroutine exists, when the namespace was not created yet
14:34 dalek Perlito: review: https://github.com/fglock/Perlito/commit/ac8d0358e7
14:36 Lox apparently rebuilding everything from scratch fixed it
14:36 Lox how... odd
14:36 Lox i'm still curious what would've caused that issue in the first place
14:37 FROGGS Lox: do you had a previous installation of rakduo/parrot dangling around?
14:37 FROGGS rakudo*
14:37 Lox i don't think so, i think i got rid of everything
14:38 FROGGS odd
14:43 colomon joined #perl6
14:45 arnsholt joined #perl6
14:53 exouniverse left #perl6
14:58 dakkar joined #perl6
15:03 raiph joined #perl6
15:12 Rotwang joined #perl6
15:31 colomon Compress::Zlib?  \o/
15:31 kaare__ joined #perl6
15:51 virtualsue joined #perl6
15:55 xinming_ joined #perl6
15:58 prammer joined #perl6
16:05 kivutar joined #perl6
16:06 FROGGS what?
16:06 tadzik yeah, we have that :)
16:08 jnthn tadzik: When are you planning to work on the release?
16:09 tadzik jnthn: very late evening, probably
16:09 tadzik like, borderline midnight
16:09 tadzik may be sooner, probably won't be later
16:09 tadzik jnthn: is moar already released?
16:11 jnthn tadzik: No, I'm gonna do that shortly, but there's a bugfix that affects getting a decent Moar Star that I'd like to try and get in.
16:11 jnthn tadzik: Sounds like I've got a couple of hours to take a shot at it? :)
16:16 masak who created Compress::Zlib? are they on this channel? I have a pull request.
16:16 retupmoca masak: right here
16:17 masak retupmoca: see https://github.com/retupmoca/P6-Compress-Zlib/pull/1
16:17 retupmoca masak: merged. I've been really bad about those swap files lately
16:18 masak retupmoca: it's probably because you add file with your eyes shut :P
16:18 masak anyway, retupmoca++ # for Compress::Zlib
16:18 masak retupmoca++ # for accepting pull requests
16:18 masak files*
16:18 cosimo you guys seen this yet? http://hacklang.org/
16:19 cosimo ( http://www.wired.com/wiredenterprise/2014/03/facebook-hack/ )
16:19 cosimo i'm curious to know your thoughts :)
16:22 masak cosimo: ooh, a TypeScript-for-PHP? :)
16:24 cosimo looks like a php wrapping of some kind, yes
16:24 kurahaupo joined #perl6
16:25 Lox excuse my cluelessness, but what are you getting at with lines like this?
16:25 Lox (10:14:34) «masak» Lox++ FROGGS++ # debugging
16:25 Lox i'm assuming that it means something to one of the bots
16:26 kurahaupo Lox: social bonus points, of a sort
16:26 kurahaupo Lox++
16:26 kurahaupo :)
16:26 Lox I guess that means I'm cool or something, then
16:27 Lox Charming
16:27 moritz Lox: currently no bot tracks that karma, but at some point I'll probably recover it from the logs
16:27 Lox oh, karma, that makes sense
16:27 Lox I'll go back to lurking and learning, then
16:34 FROGGS joined #perl6
16:42 [Coke] .seen allison
16:42 yoleaux I haven't seen allison around.
16:44 rurban isnt it /seen
16:45 [Coke] not an irssi command.
16:46 treehug88 joined #perl6
16:48 TimToady sigh, power failure this morning, so dunno how long I'll be on (laptop battery not in good shape)
16:50 [Coke] are we pretty much dedicated to a monthly star release at this point?
16:50 [Coke] I know in the past we had said ad hoc.
16:54 jnthn [Coke]: Given the current velocity, I think monthly will continue to make sense for a while.
16:54 [Coke] Anyone mind if I create perl6.github.com ?
16:54 moritz [Coke]: what will you put on it?
16:54 [Coke] jnthn: fair enough. when plotting out milestones, wanted to have monthlies for the compiler, will add them for star as well.
16:55 moritz I don't mind, just want to avoid fragementation of the websites; but if there's a good reason, go for it
16:55 [Coke] moritz: the project stuff I was suggesting came to mind. though I can do that in perl6.org, I suppose. nevermind.
16:55 [Coke] moritz++
16:56 [Coke] how can I run perl6.org repo locally?
16:57 moritz [Coke]: you need App::Mowyw from CPAN to build it
16:57 moritz [Coke]: and there's an app.psgi in the repo which you can use as a dev webserver
16:58 isBEKaml joined #perl6
16:58 [Coke] moritz++
16:58 hoverboard joined #perl6
16:59 [Coke] if I were to track roadmap information in a text friendly format, any issues with json?
17:03 [Coke] perl app.psgi -> Can't return outside a subroutine at app.psgi line 8.
17:06 moritz [Coke]: plackup app.psgi
17:09 moritz [Coke]: no issues with json
17:09 moritz the "features" page also uses json, iirc
17:10 [Coke] how do i populate online/ ?
17:11 arnsholt jnthn: I'm gonna assume attribute hints are stable over different incarnations of the various JAST::Node classes. That sound reasonable?
17:11 Rotwang joined #perl6
17:12 moritz [Coke]: run mowyw
17:13 jnthn arnsholt: yes.
17:16 arnsholt Excellent
17:17 TimToady out of battery &
17:21 PerlJam .oO( I didn't know that TimToady was battery operated )
17:22 isBEKaml .oO(I think he just sprang out of a battery and tried to get back in)
17:24 dalek perl6.org: 382c11c | coke++ | README.md:
17:24 dalek perl6.org: Document how to run the site locally
17:24 dalek perl6.org: review: https://github.com/perl6/perl6.org/commit/382c11c89b
17:25 moritz [Coke]++ # faster than me
17:26 Timbus joined #perl6
17:26 moritz stats for perl6.org since 2009 show an IE6 percentage of 4.4%
17:27 timotimo hello my friends
17:27 timotimo what have i missed today so far?
17:27 [Coke] if my stuff doesn't work in IE6, I'll happily provide phone support getting those people to upgrade.
17:27 [Coke] because that is less painful.
17:27 isBEKaml moritz: Well, I used to browse with IE6 not too long ago. :-)
17:27 moritz isBEKaml: but you don't anymore?
17:28 moritz [Coke]: I'm thinking about dropping the IE6 requirement from README.md
17:28 isBEKaml moritz: Workplace enforcement. Thankfully, they were all upgraded since.
17:28 timotimo oh god i dropped right into a discussion on IE6?
17:29 [Coke] moritz: wow, I must have an IE6 blindspot.
17:29 dalek perl6.org: 8877076 | moritz++ | README.md:
17:29 dalek perl6.org: bump IE version requirement to IE8
17:29 dalek perl6.org: review: https://github.com/perl6/perl6.org/commit/8877076751
17:32 kurahaupo_mobile joined #perl6
17:35 molaf joined #perl6
17:39 kaleem joined #perl6
17:46 denis_boyun joined #perl6
17:53 guru joined #perl6
17:59 SamuraiJack_ joined #perl6
18:01 jnap joined #perl6
18:02 dalek rakudo/nom: 0488f54 | jnthn++ | / (6 files):
18:02 dalek rakudo/nom: Fix sub calls from pre-compiled roles in Moar.
18:02 dalek rakudo/nom:
18:02 dalek rakudo/nom: In some cases, the fixup to the role outer would not be done. On the
18:02 dalek rakudo/nom: Parrot and JVM backends, things "magically" work out due to keeping
18:02 dalek rakudo/nom: track of a prior invocation. Moar's stricter closure model (which the
18:02 dalek rakudo/nom: JVM backend should also move to) lacks the "magic", and so highlighted
18:02 dalek rakudo/nom: the reliance on it. This fixes things for on MoarVM, but leaves alone
18:02 dalek rakudo/nom: JVM and Parrot's current approach given we're right up against the
18:02 dalek rakudo/nom: monthly release; after the release, they can all switch to the same,
18:02 dalek rakudo/nom: newer, model.
18:02 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/0488f541c9
18:02 jnthn That deals with the "pre-compiling modules that makes native calls doesn't work" issue
18:03 timotimo that sounds good!
18:03 jnthn And boy was it fiddly...
18:03 timotimo yeah, it does look fiddly :(
18:04 retupmoca jnthn++
18:04 retupmoca building now - hopefully Compress::Zlib will panda install on moar now
18:05 [Coke] jnthn++
18:05 jnthn Hopefully!
18:05 jnthn My Windows message box example worked out.
18:05 timotimo will we get a win32 api module? :P
18:05 jnthn Some day, I imagine :P
18:06 timotimo can we get a GUI based on MFC for Perl 6?
18:07 timotimo that'll be a hit with the hip masses!
18:07 jnthn ...MFC? :P
18:07 jnthn The 1990s called :)
18:07 isBEKaml Is MFC still popular? I thought it died in a fit. :D
18:08 timotimo haha
18:08 timotimo i don't think anyone anywhere still likes that
18:08 timotimo what was bad about it, btw?
18:08 ajr joined #perl6
18:08 jnthn Mebbe it's still used, but WPF and WinForms on .Net are the way I see most folks building their Windows GUI apps.
18:09 jnthn timotimo: I dunno if it was bad at the time it was the thing to use, it's just that there's a bunch of more pleasant options now.
18:09 timotimo fair enough
18:19 TimToady power back on \o/
18:20 FROGGS jnthn: I have heard that MS kinda switched back to MFC for windows 8... is that not correct?
18:21 TimToady what, they don't use OLE automation anymore?
18:21 FROGGS I just know that a software company here is pretty pissed because they changed everything to use wpf with a great effort and then got slapped like that my MS
18:23 FROGGS hmmm, can't find anything about that in the internets
18:23 ajr_ Depending on Microsoft must be a  miserable existence; at best you'll get assimilated, at worst abandoned.
18:24 TimToady otoh, my Linux Mint just told me I have to upgrade now
18:24 jnthn FROGGS: Microsoft politics are messy.
18:24 jnthn FROGGS: WPF came from one division, Windows is built by another.
18:26 retupmoca Compress::Zlib::Raw still won't install :(
18:26 retupmoca None of the parametric role variants for 'IntTypedCArray' matched the arguments supplied.
18:26 retupmoca Cannot call ''; none of these signatures match:
18:27 retupmoca my golfed version from the other day works fine now though
18:30 jnthn Ah. That must be a totally separate problem.
18:31 retupmoca probably
18:31 retupmoca I'll try and golf it tonight
18:33 grondilu joined #perl6
18:33 grondilu r: say 1 +> -1
18:34 camelia rakudo-parrot 01d16c, rakudo-jvm 01d16c, rakudo-moar 01d16c: OUTPUT«2␤»
18:34 grondilu n: say 1 +> -1
18:34 camelia niecza v24-109-g48a8de3: OUTPUT«0␤»
18:34 * grondilu suspects that's the reason he fails to translate http://rosettacode.org/wiki/Elementary_cellular_automaton#C
18:36 grondilu though I'm not sure, this bitwise wizardry is quite tough to grasp
18:38 k1lldash9 joined #perl6
18:38 grondilu my attempt so far:  http://paste.siduction.org/20140320183809
18:38 pecastro joined #perl6
18:39 jnthn retupmoca: Oddly, just putting a CArray[int] usage into my example code here doesn't fail...
18:40 grondilu hum...  that paste link does not work :/  so here's the gist https://gist.github.com/grondilu/9670877
18:41 retupmoca jnthn: I replicate it in my repl by compiling Compress::Zlib::Raw; and then doing "use Compress::Zlib::Raw; use NativeCall; my $x = CArray[int8].new;"
18:41 vendethiel grondilu: you're trying to go faster ?
18:41 jnthn retupmoca: I...think Compress::ZLib::Raw is missing from the Github repo?
18:42 retupmoca jnthn: but that only fails when Compress::Zlib::Raw is compiled
18:42 [Sno] joined #perl6
18:42 retupmoca if the use is loading the .pm6 it works fine
18:42 retupmoca jnthn: missing where?
18:42 jnthn retupmoca: There's no Raw.pm
18:43 retupmoca jnthn: https://github.com/retupmoca/P6-Compress-Zlib-Raw
18:43 retupmoca jnthn: I split ::Raw into a different repo
18:43 jnthn Oh...it's a separate module
18:44 unclefester2 joined #perl6
18:46 unclefester2 Trying to build jvm rakudo.  In step .../nqp-j --target=jar ... /World.nqp - Getting: Error while compiling op lexotic: Method 'jast' not found for invocant of class 'NQPMu'
18:50 jnthn Ouch...wonder if that's the thing I just changed.
18:50 jnthn oh fuck, yes.
18:50 jnthn *sigh*
18:50 FROGGS jnthn: I get that too
18:51 FROGGS I was just wondering what I am doing wrong :o)
18:51 nwc10 new release?
18:52 jnthn No, it's not a Moar issue
18:52 jnthn And Rakudo one isn't cut
18:52 nwc10 yay
18:52 nwc10 yay again
18:52 tadzik <cat looking over the couch saing "soon".png>
18:54 unclefester2 Thanks all.  I'll either try to roll back a few days or wait until it's fixed ...
18:55 jnthn Yeah, I'm working on the fix. Sorry.
18:56 unclefester2 left #perl6
18:57 kurahaupo_mobile joined #perl6
19:03 grondilu vendethiel: yes
19:06 grondilu it's fast as it is on this gist, but it's not right:  the pattern should be seen on the right side as well.
19:07 grondilu set the initial state to 2**31 instead of 1 to see the problem more clearly if you want.
19:10 cognominal with latest git rakudo compiled with parrot , trying to install HTTP::Client gives  No STable at index 1
19:10 cognominal build stage failed for HTTP::Client: Failed building lib/HTTP/Client/Request.pm6
19:12 cognominal ... on a 64bit mac
19:12 FROGGS star: use HTTP::Client
19:12 camelia star 2013-09: OUTPUT«===SORRY!===␤Could not find HTTP::Client in any of: /home/p6eval/star/lib/parrot/5.5.0/languages/perl6/site/lib, /home/p6eval/star/lib/parrot/5.5.0/languages/perl6/vendor/lib, /home/p6eval/star/lib/parrot/5.5.0/languages/perl6/lib, /home/p6eval/.perl6/2013.09/…»
19:12 FROGGS cognominal: and it was working before?
19:12 cognominal never tried
19:13 cognominal will try to install other packages
19:15 cognominal can I use panda with moarvm?
19:16 nwc10 cognominal: I believe that the answer is "panda is supposed to work with Moarvm"
19:16 nwc10 but you would be an early adopter :-)
19:16 retupmoca cognominal: it's been working for me
19:16 nwc10 it needs to work by next week
19:16 cognominal is it faster than with parrot?
19:17 retupmoca I'm not sure, it's been a long time since I've used parrot
19:17 retupmoca I think it is
19:17 cognominal install the Web package works (with parrot).
19:19 FROGGS panda does work with moar, problem is that it delegates to your perl6 binary, which is the parrot one on my box
19:19 nwc10 FROGGS: well fix that then! :-)
19:19 nwc10 (The perl 6 binary)
19:20 nwc10 but also fix panda, to delegate to $^X
19:20 nwc10 r: say $^X
19:20 camelia rakudo-parrot 01d16c, rakudo-jvm 01d16c, rakudo-moar 01d16c: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/tmpfile�Unsupported use of $^X variable; in Perl 6 please use $*EXECUTABLE_NAME�at /tmp/tmpfile:1�------> [32msay $^X[33m�[31m<EOL>[0m�»
19:20 FROGGS nwc10: done so in the "eleven" branches :o)
19:20 nwc10 yay, non-LTA message
19:20 nwc10 prn: say $*EXECUTABLE_NAME
19:20 cognominal it is very exciting to see moarvm leapfroggin parrot.
19:20 nwc10 oh, that doean't work
19:20 nwc10 rn: say $*EXECUTABLE_NAME
19:21 camelia rakudo-jvm 01d16c: OUTPUT«perl6-j␤»
19:21 camelia ..rakudo-parrot 01d16c: OUTPUT«./rakudo-inst/bin/perl6-p␤»
19:21 camelia ..rakudo-moar 01d16c: OUTPUT«perl6-m␤»
19:21 camelia ..niecza v24-109-g48a8de3: OUTPUT«/home/p6eval/niecza/run/Niecza.exe␤»
19:21 FROGGS P: say $*EXECUTABLE_NAME
19:21 camelia pugs: OUTPUT«␤»
19:21 FROGGS heh
19:26 moritz p6: say $*EXECUTABLE_NAME.^name
19:26 camelia rakudo-parrot 01d16c, rakudo-jvm 01d16c, rakudo-moar 01d16c, niecza v24-109-g48a8de3: OUTPUT«Str␤»
19:29 dalek rakudo/nom: 6d62ea4 | jnthn++ | src/ (4 files):
19:29 dalek rakudo/nom: Unbust JVM/Parrot builds.
19:29 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/6d62ea48c7
19:32 cosimo joined #perl6
19:35 tadzik release tiem \o/
19:42 dalek rakudo/nom: 1d29330 | tadzik++ | docs/ (2 files):
19:42 dalek rakudo/nom: Add a new release announcement
19:42 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/1d293300dd
19:42 tadzik patches and proofreading welcome
19:50 TimToady grondilu: I already added a version that is 10 times faster or so
19:50 TimToady so I could remove all the comments badmouthing P6's speed :)
19:52 cognominal I also get   "No STable at index 1" for LWP::Simple which is a  rakudo star module. Something fishy here.
19:54 cognominal seems a serialization problem :
19:54 cognominal nqp/MoarVM/src/6model/sc.c:174:            "No STable at index %d", idx);
19:54 cognominal nqp/src/vm/parrot/6model/serialization_context.c:95:            "No STable at index %d", idx);
19:58 telex joined #perl6
20:00 moritz tadzik: announcement looks good
20:00 colomon \o/ release day!
20:01 tadzik hooray, release day
20:05 masak yay
20:05 * masak chips in and reads l'announcement
20:08 jnthn tadzik: Maybe the TimToady++ Unicode improvements deserve a mention?
20:08 masak announce looks good.
20:08 tadzik oh, serly :)
20:08 tadzik surely
20:08 * masak .oO( ser Larry Wall )
20:09 dalek nqp: 21a89d7 | tadzik++ | VERSION:
20:09 dalek nqp: Bump version to 2014.03
20:09 dalek nqp: review: https://github.com/perl6/nqp/commit/21a89d778c
20:10 jnthn masak: How cheesy :P
20:10 dalek rakudo/nom: 35b2bff | tadzik++ | tools/build/NQP_REVISION:
20:10 dalek rakudo/nom: [release] Bump NQP revision
20:10 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/35b2bff60b
20:10 dalek rakudo/nom: c413905 | tadzik++ | VERSION:
20:10 dalek rakudo/nom: [release] Bump VERSION
20:10 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/c413905f61
20:10 dalek rakudo/nom: b576526 | tadzik++ | docs/announce/2014.03.md:
20:10 dalek rakudo/nom: Add things to release annoucement
20:10 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/b5765267b4
20:10 tadzik so much karma
20:10 jnthn so appreciation
20:12 PerlJam This channel makes me think of "Guys and Dolls" way more often than I should  :)
20:12 * masak queues up "Guys and Dolls"
20:15 masak cosimo: reading more about Hack has only reinforced my feeling that JavaScript:TypeScript::PHP:Hack.
20:15 masak cosimo: the fact that it's written in OCaml is interesting, too. I suppose that's how they are able to compile it so fast.
20:16 masak it all makes a lot of sense. doesn't make me like PHP more, but probably makes life loads easier for the Facebook folks.
20:18 tadzik parrot build is the most time-consuming step of a release process :o
20:19 PerlJam masak: that analogy seems spot-on to me.
20:19 PerlJam (ergo, you're not only reinforcing it with yourself, but with me as well :)
20:19 masak PerlJam: TypeScript, on the other hand, makes me increasingly excited. it just makes so much sense.
20:20 masak PerlJam: I've resolved to try and write my next big JavaScript thing in TypeScript, just to try it out.
20:20 PerlJam nice.
20:20 PerlJam Be sure and blog about it so we can gain some vicarious knowledge :)
20:21 tadzik prepare for 3-second-long compilation times :)
20:21 tadzik it's like rakudo pre-nom
20:21 tadzik so still fun :P
20:22 arnsholt masak: Did you see my link to my Python thing the other day, BTW?
20:22 FROGGS t/spec/S05-mass/charsets.rakudo.parrot                        (Wstat: 0 Tests: 16 Failed: 0)
20:22 FROGGS TODO passed:   8
20:22 FROGGS t/spec/S32-exceptions/misc.rakudo.parrot                      (Wstat: 0 Tests: 709 Failed: 2)
20:22 FROGGS Failed tests:  10, 647
20:31 jnthn FROGGS: Hm, mebbe fudge/ticket those latter 2?
20:31 FROGGS jnthn: maybe I can fix them, I'm about to have a look
20:31 jnthn ok
20:32 FROGGS not ok 647 - &uc is a suggestion
20:32 FROGGS #      got: 'uc'
20:32 FROGGS # expected: '&uc'
20:32 FROGGS not ok 10 - 'pack "A1", "mÄ"' died
20:32 FROGGS let's see
20:34 FROGGS m: say pack "A1", "mÄ" # why is that supposed to die?
20:34 camelia rakudo-moar 1d2933: OUTPUT«Buf:0x<6d>␤»
20:35 masak arnsholt: yes. been looking through the commits so far.
20:35 FROGGS m: say pack "A2", "mÄ"
20:35 camelia rakudo-moar 1d2933: OUTPUT«non-ASCII character 'Ã' while processing an 'A' template in pack␤  in block  at src/gen/m-CORE.setting:5410␤  in block  at src/gen/m-CORE.setting:5400␤  in method reify at src/gen/m-CORE.setting:7559␤  in method reify at src/gen/m-CORE.setting:742…»
20:36 arnsholt masak: Cool. If inspiration strikes, I'd be happy to commitbit you =)
20:39 dalek roast: 4070a73 | (Tobias Leich)++ | S32-exceptions/misc.t:
20:39 dalek roast: we need to cover both chars to hit Ä
20:39 dalek roast: review: https://github.com/perl6/roast/commit/4070a73272
20:42 masak arnsholt: let me try to deserve that commitbit. I'll hit you with some PRs when you least expect it ;)
20:42 Goodbox joined #perl6
20:44 masak Goodbox: greetings, my good box!
20:44 dalek roast: 97fa523 | (Tobias Leich)++ | S32-exceptions/misc.t:
20:44 dalek roast: &instead is only suggested if there was an & in the unkown sub
20:44 dalek roast: review: https://github.com/perl6/roast/commit/97fa5234e0
20:44 arnsholt masak: Deal!
20:48 dalek roast: 2d11bf4 | (Tobias Leich)++ | S32-exceptions/misc.t:
20:48 dalek roast: fix query for routine_suggestion
20:48 dalek roast: review: https://github.com/perl6/roast/commit/2d11bf4ef1
20:49 Goodbox hey :p (Yes I choosen strange nick). Just taking a walk on this corner of freenode. I recently interested to (R)?ex and Perl6, firstly because I found mysefl managing a node, and because I just love computing =)
20:49 FROGGS Goodbox: welcome :o)
20:50 FROGGS m: say 'Ä'.encode
20:50 camelia rakudo-moar 1d2933: OUTPUT«Buf:0x<c3 84>␤»
20:50 Goodbox Thank's ! :)
20:52 * masak hands Goodbox back that apostrophe
20:53 masak Goodbox: Perl 6 is well worth becoming interested in!
20:53 FROGGS m: say 'mÄ'.ords
20:53 camelia rakudo-moar 1d2933: OUTPUT«109 196␤»
20:53 masak Goodbox: we like to think of it as, if not a good first language, then at least a good last language. :)
20:55 FROGGS tadzik: I have a rakudopatch in the queue, please hold back that release :o)
20:55 tadzik nooooooo
20:55 Goodbox Dont hesitate to correct my bad english, Yes I found of right computing despite I'm not writing perl everyday ^^ However for a apprentice sysadmin like me, this is essential to know what about :)
20:55 moritz \o/ parrot release notes contain a quote from HPMoR
20:56 masak Goodbox: you're doing fine. and we're happy to have you here.
20:56 masak moritz: yes, Util++
20:56 masak Goodbox: the beginner experience here is strange and bewildering. some of us are bots, not humans. don't hesitate to ask about stuff.
20:57 * masak .oO( on IRC nobody knows you're a bot )
20:58 dalek rakudo/nom: 0caa198 | (Tobias Leich)++ | src/core/Buf.pm:
20:58 dalek rakudo/nom: complain about chars not bytes
20:58 dalek rakudo/nom:
20:58 dalek rakudo/nom: When why want to cry about non-ASCII characters we should use .ords instead of using
20:58 dalek rakudo/nom: the list of bytes.
20:58 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/0caa1989da
20:58 FROGGS tadzik: done, now I am rerunning spectests
20:59 tadzik ok, they're your responsibility now :P
20:59 tadzik (I'm serious)
20:59 FROGGS sure
21:00 Goodbox I was sometimes reading articles from "Les mongeurs de Perl" which were published in linuxmag's. About my personnal preferencies, i find python really snicky (and so slow), (But far from me the idea to start a troll)
21:02 masak Goodbox: people have all sorts of preferences. Perl 6 is for people who like expressive power, lexical scoping, and extensibility.
21:02 cognominal GoodBox, the mongueurs hangout is #perlfr on irc.perl.org :)
21:03 cognominal and we have "les journées Perl" in June :  http://journeesperl.fr/fpw2014/
21:03 Goodbox Thank's cognominal ! I guess, knowing me, having more florescence using perl families languages ;)
21:04 Goodbox cognominal: Nice! Plan to go to the RMLL's ?
21:04 cognominal I don't know.
21:05 cognominal where are you? In France?
21:05 grondilu cognominal: will there be talks about Perl 6?
21:05 Goodbox (Yes, I first join english communities beacause I have to prictice :p) cognominal I'm from the east of France and I plan to go at Montpellier with some french librist's :)
21:06 cognominal grondilu, it is up to you :) and people in this channel. I am not sure to attend :(
21:07 cognominal I wish I would be sure.
21:08 grondilu well, there aren't many french people in this channel, are there?  You and me are the only one I know, IIRC.
21:08 cognominal elbeho told me  fpw is almost at the same time as YAPC::NA  :(
21:11 cognominal grondilu, I am working on articles about Perl 6 for the french linuxmag. If you are interested to collaborate... Witth rakudo star supporting moarvm, it will be a good time to publsh
21:13 grondilu cognominal: I suggest you wait until MoarVM supports conc.
21:13 grondilu then you'll have even more pretty cool stuff to talk about
21:13 cognominal conc?
21:13 nwc10 concurrency
21:13 grondilu concurrency
21:13 nwc10 I think that jnthn hopes to have it in for April
21:14 nwc10 how frequently does the magazine come out? and what are its publication deadline dates?
21:15 cognominal We need to give an article at the end of the month to get it publish for the next month.
21:15 japhb__ jnthn: I've recently started loading down r-j enough with threads to result in some instability (errors coming up when they don't at lower loads).  However, I've yet to golf it down to A) something easy to reproduce, and B) prove that it isn't something external to r-j that is failing (since the program in question spawns many dozens of subprocess pipes).  Do we have any existing threading load-tests for Rakudo?
21:15 grondilu as cann have been observed on perlmonks, whenever you talk about MoarVM, people go "meh, yet another dead end" or something.  So I think you really need to have something spectacular to show off before bothering them.
21:16 nwc10 cognominal: how many articles do you think you will be writing?
21:16 nwc10 sorry if I seem to be all questions
21:16 jnthn japhb__: No, we don't. BUT I do know that there can be issues with regards to not having switched over to the thread-safer coderef stuff in JVM.
21:17 brrt joined #perl6
21:17 jnthn japhb__: You may be hitting that. I'd love to say "try it on Moar" but ENOTYET.
21:17 jnthn japhb__: Since Moar has the right model.
21:17 japhb__ Ah, OK.  Was this the stuff that you said in your recent commit you'll need to do after the release is cut?
21:17 * masak finds much of the debate over "crumbling foundations" truly interesting in https://news.ycombinator.com/item?id=7436401
21:17 japhb__ jnthn: Yep, understood.
21:17 cognominal I have one which use rosalind as example to talk about strings and a little about OO. I have others that have bitrotten
21:18 masak best way to describe the whole situation -- that I've found thus far -- is Kent Beck's phrase "clean code that works". you can have either "clean code" or "that works", or both, and much discusssion will happen between the two camps that have found one but not the other.
21:19 nwc10 cognominal: ah OK, these are general Perl 6 articles? Rather than "new news" articles?
21:19 jnthn japhb__: No, the thing I fixed there was a case of "Moar caught us cheating".
21:19 * japhb__ suddenly wonders if he's managing to spawn *so* many subprocess pipes that he's actually starting to run into FD limits or some such and the errors are just LTA.
21:19 japhb__ jnthn: ;-D
21:20 brrt masak - it's a good discussion isn't it
21:20 masak moritz: http://i.imgur.com/F0t5T21.png :)
21:20 * grondilu opens his eyes wide at the |args syntax on http://rosettacode.org/wiki/Elementary_cellular_automaton#Perl_6
21:21 masak brrt: for a long time, I had big problems with "clean code" vs "that works", because it seemed that both camps have a point. well, it turns out they do, and you can have both, and it's not obvious that you can, and that's why there are long discussions about this.
21:21 FROGGS grondilu: about the "yet another dead end": when you state what improvements moarvm brought us, then nobody can still think it is a dead end :o)
21:21 lue masak: distressing is the fact that 50% of all mentions of 2048 were misspelled :)
21:22 lue (in that png you shared)
21:22 masak lue: I suspect the screenshot is a fake :)
21:24 lue meanwhile, xkcd mentioned 2048, and I keep coming back to it (worse, I see potential mathematical questions, like "what is the largest tile you can have?" etc. :P)
21:25 masak lue: what *is* the largest tile you can have?
21:26 lue masak: my guess is ∞, like Hanoi. (You can solve a three-peg hanoi with any amount of disks; the shortest solution is just 2**n steps, IIRC)
21:27 tadzik argh
21:27 tadzik jvm is failing a lot of stuff
21:27 moritz lue: no, you can't
21:28 tadzik https://gist.github.com/tadzik/9674264
21:28 masak lue: no, there has to be an upper limit.
21:28 masak lue: because there's finite room, and it runs out.
21:28 moritz lue: you are limited to 2 ** ($n + 2), where $n is number of positions
21:28 lue moritz: really? I recall that a three-peg hanoi could handle any amount of disks. I agree that's possibly not the case for 2048 though.
21:28 moritz hanoi, yes. 2048, no
21:29 jnthn tadzik: I can explain/take blame for one:
21:29 jnthn t/spec/S17-concurrency/lock.t                              (Wstat: 0 Tests: 6 Failed: 0) Parse errors: Bad plan.  You planned 7 tests but ran 6.
21:29 jnthn tadzik: I shoulda pushed that to a branch in roast, as the code that makes it pass is also in a branch.
21:30 FROGGS tadzik: jvm always failed tests
21:30 lue oh, right. In an absolutely perfect game, the largest would be 2**16 I think (imagine a snake of 15 blocks, each double the last, and a 2 appearing in the empty spot next to the existing 2)
21:30 FROGGS tadzik: seems like this month's low was at 5
21:31 tadzik oh
21:31 tadzik so, should I just carry on with the release>
21:32 FROGGS tadzik: in case I still can count we will end up with 9 failed tests (taking lock.t not into account)
21:33 jnthn We should really clean these up, but it's no worse than in previous releases.
21:33 FROGGS damnit, spectesting for something other than moar takes ages >.<
21:33 tadzik you tell me
21:34 tadzik my laptop was at load 8 for an hour spectesting this
21:34 FROGGS *g*
21:34 tadzik and make release fails with no error message
21:34 tadzik https://gist.github.com/tadzik/9674390
21:35 jnthn o.O
21:35 FROGGS O.o
21:36 colomon joined #perl6
21:36 BenGoldberg joined #perl6
21:37 FROGGS I guess `make manifest` does fail too?
21:38 japhb__ Something fun to show off serious concurrency + NativeCall: Something that draws a fractal by recursively spawning tasks to take smaller and smaller pieces of it until they hit one pixel, calculate it, *DRAW IT*, and then finish off.  Because seeing that in motion would be very fun.
21:38 tadzik FROGGS: yes
21:39 tadzik I dunno why though
21:39 FROGGS tadzik: is the MANIFEST file locked by some process or so?
21:39 tadzik the MANIFEST file is not empty
21:39 FROGGS I cannot imagine what can fail in `find t/spec -type f | grep -v '\.git' >>MANIFEST`
21:39 FROGGS and t/spec is a regular directory?
21:39 tadzik um, no
21:39 tadzik it's a symlink
21:40 tadzik should probably be find t/spec/
21:40 FROGGS maybe that confuddles it?
21:40 tadzik ah, now it works :)
21:40 FROGGS yay :o)
21:42 tadzik http://feather.perl6.nl/~tjs/rakudo-2014.03.tar.gz
21:42 tadzik please test
21:43 raiph joined #perl6
21:43 FROGGS tadzik: btw, p-spectest is clean now, will test your tarball now
21:44 tadzik awesmoe, thank you
21:46 BenGoldberg japhb__ Why not just have a top-level task spawn one sub-task per pixel directly?  I don't see the benefit of recursion, here.
21:47 BenGoldberg There would still be concurrency (all those subtasks calculating pixels), and there would still be NativeCall (drawing those pixels).
21:48 * sjn gets an error when trying to build rakudo-2014.03 with a parrot background
21:48 sjn backend*
21:48 japhb__ BenGoldberg: I left out some details.  I imagined information flowing between the layers via e.g. channels and such, and showing that multi-level task fanout doesn't result in one portion of the fractal blocking on some other portion.  (e.g. The first N tasks get assigned real threads, but then all other pixels start plodding through getting thread resources as if queued)
21:48 grondilu r: sub infix:<o>(&f, &g) { -> |args { f(g(|args)) } }; say (&atan o &tan)(pi/4)
21:49 [Coke] sjn: deets?
21:49 camelia rakudo-jvm 0caa19: OUTPUT«0.7853981633974483␤»
21:49 camelia ..rakudo-parrot 0caa19, rakudo-moar 0caa19: OUTPUT«0.785398163397448␤»
21:49 colomon BenGoldberg, japhb: I wouldn't do either -- each task doing one scanline of the fractal probably makes the most sense.  at least for something like the Mandelbrot set
21:49 [Coke] r: so?
21:49 camelia rakudo-parrot 0caa19, rakudo-jvm 0caa19, rakudo-moar 0caa19: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/tmpfileâ�¤Bogus statementâ�¤at /tmp/tmpfile:1â�¤------> [32mso?[33mâ��[31m<EOL>[0mâ�¤    expecting any of:â�¤        statement listâ�¤        prefix or t…»
21:49 sjn perl Configure.pl --gen-parrot --backends=parrot
21:49 sjn sh: 1: /home/sjn/src/rakudo-2014.03/install/bin/parrot: not found
21:50 [Coke] yup, that's fine, usually. it's looking for an existing parrot to use int eh build location, and noting it didn't.
21:50 sjn how about quieting the warning, if it's fine?
21:50 japhb__ BenGoldberg: My inspiration was a visual, thread-based variation on this: http://perl.plover.com/obfuscated/
21:51 japhb__ (But not needing to be obfuscated per se.  I'm not trying to create a sig JAPH.)
21:51 [Coke] sjn: seems reasonable. did your build stop there, or did it continue? (also, you might want --gen-nqp also)
21:51 sjn (it's the first line you get when running the first build instruction in the README file)
21:51 sjn it seems to continue fine
21:52 [Coke] I agree, it's scary.
21:52 FROGGS sjn: yes, we should hide that to not frighten anybody
21:54 tadzik agreed
21:55 japhb__ colomon: Oh, I know what makes *sense* -- I had a misspent youth doing renderings of various sorts -- but this is about crazy-that-works.  :-)
21:55 colomon oh, crazy, carry on, then.  :)
21:57 kivutar joined #perl6
22:02 dalek perl6-roast-data: 7567a7d | coke++ | perl6_pass_rates.csv:
22:02 dalek perl6-roast-data: today (automated commit)
22:02 dalek perl6-roast-data: review: https://github.com/coke/perl6-roast-data/commit/7567a7da05
22:03 [Coke] ... that's weird. usually at least pugs changes some of the failure modes.
22:05 * sjn is easily frightened
22:06 brrt joined #perl6
22:07 cosimo joined #perl6
22:07 * sjn get some more compilation warnings
22:07 sjn pastebin?
22:07 sjn no bot with pastebin suggestions..
22:09 FROGGS there are a lot of warning for the parrot backend, and just one or two when building moarvm
22:13 FROGGS tadzik: Rakudo has been built and installed.
22:14 tadzik awesome
22:15 FROGGS nice, $dayjob is done too
22:17 FROGGS tadzik: are you going to patch Make-Common?
22:17 FROGGS MAKEFILE-Commin.in or what it is called
22:18 tadzik FROGGS: probably, yes :)
22:18 FROGGS tadzik++
22:21 brrt joined #perl6
22:23 tadzik but tomorrow
22:23 tadzik for today, I'm more than done :)
22:23 skids joined #perl6
22:23 tadzik thanks for release assistance :)
22:23 FROGGS tadzik: does your tarball include my rakudo patch?
22:23 tadzik FROGGS: yes, it got it
22:23 FROGGS hmmm
22:23 tadzik s:2nd/it/in/
22:23 FROGGS t/spec/S32-exceptions/misc.rakudo.moar                      (Wstat: 0 Tests: 709 Failed: 8)
22:23 FROGGS Failed tests:  10, 554, 556, 558, 560, 562, 598, 647
22:24 FROGGS test 10 and 647 should pass on moar too
22:24 tadzik huh
22:24 FROGGS perhaps you did not pull roast?
22:24 tadzik I think I did
22:25 FROGGS the rakudo patch is in there, yeah
22:25 tadzik maybe I forgot :o
22:25 tadzik doesn't 'make spectest
22:25 tadzik ' pull roast though?
22:25 FROGGS yeah
22:25 FROGGS throws_like 'pack "A1", "mÄ"',     X::Buf::Pack::NonASCII, char => 'Ä';
22:26 FROGGS that should read A2 for example
22:26 FROGGS tadzik: no, since it is not a git repo anymore
22:26 tadzik ah, right ;|
22:26 tadzik damn
22:26 tadzik but no, impossible
22:27 tadzik I did make spectest on the rakudo repo before doing the release
22:27 tadzik that pulled roast alright
22:27 brrt left #perl6
22:27 FROGGS pull again?
22:27 FROGGS I dunno
22:28 tadzik the roast patch got there too late :/
22:28 tadzik it's not in
22:28 FROGGS /o\
22:29 tadzik oh noes wat do
22:29 FROGGS pull and rebundle
22:29 tadzik and sneak-reupload :/
22:29 FROGGS we cannot ship failing tests for our "stable" backend
22:29 tadzik 2014.03.01 wuld be the right thing to do
22:30 tadzik I'll do that, but I think I'll spare the release announcement
22:34 dalek rakudo/nom: 1c1a085 | tadzik++ | VERSION:
22:34 dalek rakudo/nom: [release] Bump VERSION
22:34 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/1c1a08595c
22:37 tadzik ok it's there
22:37 tadzik I'll update all the wikis tomorrowm since it's almost tommorow anyway :P
22:41 Halldor joined #perl6
22:44 Mouq joined #perl6
22:45 vky joined #perl6
22:46 Mouq tadzik++ FROGGS++
22:47 jnthn tadzik++ # release
22:48 jnthn FROGGS++ # making sure release hadn't failing tests
22:49 * timotimo is looking forward to MoarVM-2048
22:51 timotimo tadzik++ FROGGS++ # writing this seems to be hip today
22:51 FROGGS hehe
22:53 jnthn Time for some rest
22:54 jnthn I've have few to no tuits until Monday evening.
22:54 jnthn 'night o/
22:54 lizmat gnight jnthn!
22:55 FROGGS gnight, I am going to bed also
22:55 Mouq o/ jnthn
22:55 Mouq o/ FROGGS
22:56 Mouq .seen smls
22:56 yoleaux I saw smls 25 Feb 2014 16:46Z in #perl6: <smls> :P
22:56 Mouq .seen raydiak
22:56 yoleaux I saw raydiak 7 Mar 2014 20:49Z in #perl6: <raydiak> isBEKaml: agreed, I can only imagine someone didn't want to add another dependency just for that one use
22:58 denis_boyun joined #perl6
23:07 dalek doc: 93563f5 | Mouq++ | lib/modules.pod:
23:07 dalek doc: lib/modules.pod Formatting tweaks
23:07 dalek doc: review: https://github.com/perl6/doc/commit/93563f5a93
23:26 Mouq http://doc.perl6.org/language/modules
23:26 Mouq That's all copied from the wiki, and should probably face some editing, especially as we revamp S11/22
23:27 Mouq But there ya' go
23:27 timotimo better than having lost it, thanks!
23:28 * timotimo changes modules.perl6.org to match the new url
23:30 timotimo there we go.
23:30 dalek modules.perl6.org: f9c8436 | (Timo Paulssen)++ | web/index.tmpl:
23:30 dalek modules.perl6.org: link to where the modules creation guide lives now
23:30 dalek modules.perl6.org: review: https://github.com/perl6/modules.perl6.org/commit/f9c843676d
23:31 Mouq \o/
23:31 Mouq timotimo: Do you have the power to actually update modules.perl6.org though?
23:31 timotimo i think it updates automatically regularly
23:31 timotimo (last update 2014-03-20 23:17:26 GMT).
23:31 Mouq It updates the modules list
23:32 Mouq But I made a change a while ago that hasn't shown up
23:32 Mouq Well, maybe I'm wrong
23:32 timotimo oh
23:33 timotimo maybe moritz will have to do it for us?
23:33 Mouq But I don't see what I'd expect the change to be from https://github.com/perl6/modules.perl6.org/commit/ef1e109
23:33 Mouq You mean moritz++? :9
23:34 timotimo i do see that change show up
23:36 * Mouq curls http://modules.perl6.org/ to check if his browser's just weird
23:36 Mouq Mmm, still points to the ecosystem wiki
23:37 Mouq You sure you're not reading the commit backwards?
23:37 Mouq :P
23:37 timotimo oh damn
23:37 timotimo i was
23:38 Mouq timotimo: Heh, I do that too :)
23:43 araujo joined #perl6

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

Perl 6 | Reference Documentation | Rakudo