Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2013-11-29

Perl 6 | Reference Documentation | Rakudo

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

All times shown according to UTC.

Time Nick Message
00:00 rurban1 joined #perl6
00:18 cognominal__ joined #perl6
00:27 rafl joined #perl6
00:33 rafl joined #perl6
00:38 rurban1 joined #perl6
00:40 * [Coke] waves from post-thanksgiving food coma.
00:45 rafl joined #perl6
00:49 prevost joined #perl6
00:53 rafl joined #perl6
01:00 hypolin joined #perl6
01:15 * colomon is watching end of MST3K marathon
01:32 fridim__ joined #perl6
01:41 jnap joined #perl6
01:48 rurban1 joined #perl6
01:50 araujo joined #perl6
01:53 araujo joined #perl6
02:11 raydiak joined #perl6
02:39 BenGoldberg joined #perl6
03:13 colomon joined #perl6
03:16 rurban1 joined #perl6
03:55 Maddingu1 joined #perl6
04:07 gfldex joined #perl6
04:09 aindilis joined #perl6
04:16 colomon joined #perl6
04:21 cxreg so regarding concurrency, is it the plan to take the heretical "what we did on jvm" and make that the official spec?
04:28 woosley joined #perl6
04:33 TimToady it already is
04:33 TimToady and I've already installed at least one rosettacode entrie that relies on it :)
04:34 TimToady 'entrie' sound tres chic
04:36 TimToady cxreg: see http://rosettacode.org/wiki​/Metered_concurrency#Perl_6
04:36 cxreg neat.  TimToady++
04:37 cxreg i really like the look of the whole thing.
04:37 cxreg jnthn++
04:37 TimToady dining philosophers is a bit more...problematic...
04:37 cxreg it's like the good parts of node.js without the horrible parts
04:37 cxreg now, we just need a complement of useful libraries.  to the batcave!
04:39 cxreg we have deadlocking issues?
04:39 TimToady depends on which primitives you use, and how you use them :)
04:40 cxreg ok, fair enough
04:40 cxreg enuf rope and such
04:40 TimToady but D.P. is specifically designed to be difficult
04:41 * TimToady is thinking of solving it by making whoever just finished dinner grab the single large lollipop, which takes two hands to hold, and keeps you from getting hungry again until someone else grabs it :)
04:43 TimToady but we don't have forks or lollipops as primitives; philosophers can easily be started, however :)
04:46 cxreg nci for rakudo on jvm is still zavolaj right?
04:46 * TimToady suspects his lollipop solution is equivalent to using a "waiter"
04:46 TimToady it's still in the works, as I understand it
04:46 cxreg Waiter Model > Actor Model? :)
04:47 TimToady using a metaphorical waiter who refuses to serve a philosopher that might cause deadlock
04:48 TimToady other solutions are of the form, make some subset of the philosophers left-handed
04:49 TimToady or introduce a retry on failure to get both forks
04:51 * TimToady suspects the Pike solution can still deadlock, if we take its description at face value: this solution avoids deadlocks by adding a 20% chance that a philosopher drops the fork if he can't pick up both.
04:53 TimToady seems quite likely that all five philosophers would choose the 80% chance
04:54 TimToady http://rosettacode.org/wiki/Dining_philosophers is the entry in question, btw
04:57 preflex_ joined #perl6
05:03 lue .oO(This situation is easily avoided by reminding the waitstaff that everyone requires two forks and really it would be simpler to just have 10 forks right there.)
05:08 lue Is there no Perl 6 way for http://rosettacode.org/wiki/Mutex yet, or has our resident RosettaCode writer TimToady just not gotten to it yet? :)
05:15 TimToady we have more than one such writer
05:16 TimToady in a sense, a Mutex is just a degenerate case of the Semaphore we already used
05:27 TimToady but S17 does define locks, which are essentially Mutexes
05:40 BenGoldberg joined #perl6
06:02 berekuk joined #perl6
06:11 ssutch joined #perl6
06:25 kaleem joined #perl6
06:29 [Sno] joined #perl6
06:29 cxreg traits on jvm seem to cause an infinite stream of java.lang.NullPointerException
06:29 cxreg roles, that is
06:39 Woodi o/ today :)
06:42 Woodi I think philosophers should be gentle enough to just ask each other if they need something :)
06:45 Woodi but probably circular structures always will have some specific problems... probably hierarchical work distribution can help for teams. but if we have GC subsystem we can have transactional subsystem and others probably too...
06:46 Woodi ETOMANYPROBABILITY...
06:51 Woodi about naming things: there is (probably) forgotten naming scheme used by OSI for their famous OSI network framework - tree like: root have no label, lvl1 nedes are [a-z]/, lvl2 is 100, 200, later in each branch recomendations have numbers 105, 115, etc
06:56 SamuraiJack_ joined #perl6
06:56 Woodi moust known example is branch 'X' with mail system recomendation X.400 used by some rich folks :) and X.500 - "The Directory": ambitious one directory for all and everything: DNS, employers list, configurations. LDAP is TCP/IP descendant of it. and our public key infrastructure is just X.509 recomendation :)
06:57 Woodi maybe it can help somehow...
07:01 xenoterracide joined #perl6
07:02 xinming joined #perl6
07:06 raydiak preflex: tell moritz raydiak@cyberuniverses.com is ready to receive mail, now
07:06 preflex Consider it noted.
07:10 rurban1 joined #perl6
07:13 darutoko joined #perl6
07:28 rindolf joined #perl6
07:37 nnunley_ joined #perl6
07:45 dmol joined #perl6
07:58 dakkar joined #perl6
08:08 TimToady by the way, IO on jvm seems to have a lock that hangs if you try to print stuff from multiple threads
08:08 mathw Oh. Well that's fun.
08:08 TimToady worked around by setting up a channel for output...
08:09 moritz raydiak: invitation sent.
08:09 preflex moritz: you have 1 new message. '/msg preflex messages' to read it.
08:11 xinming joined #perl6
08:19 TimToady http://rosettacode.org/wiki​/Dining_philosophers#Perl_6
08:20 hoelzro TimToady++ # Perl 6 examples
08:33 kaleem joined #perl6
08:33 Rotwang joined #perl6
08:49 jnthn TimToady: How'da thought that you can fix concurrency problems using a concurrency control mechanism ;)
08:49 jnthn TimToady: Please file a ticket on the IO lock bug, though. It should not hang.
08:50 jnthn TimToady: In fact, there's already code in there that's meant to make it safe.
08:50 jnthn So something's up with it
08:53 fhelmberger joined #perl6
08:58 TimToady just for lue++ http://rosettacode.org/wiki/Mutex#Perl_6
09:02 jnthn Short code is short :)
09:19 kst` joined #perl6
09:20 mathw I like that a lot
09:21 mathw Good, comprehensible terminology.
09:21 moritz j: Lock.new.protect: { say 42 }
09:21 camelia rakudo-jvm e5fd34: OUTPUT«42␤»
09:21 Ulti_ joined #perl6
09:22 retupmoc1 joined #perl6
09:22 japhb__ joined #perl6
09:24 mathw Unfortunately after that glimpse of joy, and a couple of happy hours writing Clojure yesterday, I now have to return to wrangling SQL Server Reporting Services, for which I do at least actually get paid...
09:25 cooper- joined #perl6
09:26 noxgirl joined #perl6
09:27 Alina-malina joined #perl6
09:27 cooper joined #perl6
09:27 moritz mathw: it sounds like you should really find another job :-)
09:28 gfldex joined #perl6
09:28 mathw moritz: I've got recruiters queuing up. Unfortunately it's all .NET stuff.
09:33 djanatyn joined #perl6
09:33 felher joined #perl6
09:34 roboticus joined #perl6
09:35 markov joined #perl6
09:35 cxreg joined #perl6
09:50 nwc10 mathw: where in the world are you?
09:52 mathw nwc10: Nottingham, England
09:53 nwc10 ah OK. I'm aware of everyone trying to recruit for Perl jobs in London
09:53 mathw There are many interesting jobs I could go for in London
09:53 mathw Unfortunately that does involve moving to London
09:53 lizmat TimToady: Lock.new.protect: { your-ad-here() } would be shorter
09:54 nwc10 it seems that there is a #nottingham.pm over on irc.perl.org, but I presume that you're aware of that and have asked them
09:54 lizmat and I also wonder whether that couldn't be optimized to a special, disposable/reusable lock
09:54 mathw lizmat: if nobody else can see the lock, is it worth having?
09:54 mathw nwc10: yeah I know them
09:55 lizmat in a single threaded program, it could become a noop
09:55 mathw nwc10: my job's still better than most of theirs
09:55 mathw working for a university does have its perks
09:57 nwc10 clearly one of these London based firms with vacant seats needs to set up a Nottingham satellite office to solve this.
09:59 grondilu joined #perl6
10:01 rurban1 joined #perl6
10:01 grondilu TimToady: about http://rosettacode.org/wiki​/Dining_philosophers#Perl_6.  Why not 'class Fork is Lock { method grab($who, $which) { remark "$who grabbing $which fork"; self.lock }' instead of 'class Fork { has $!lock = Lock.new }' ?
10:02 grondilu in a nutshell, inheritence seems more appropriate than encapsulation, here.
10:03 lizmat commute to LPW continues&
10:03 woolfy left #perl6
10:04 mathw nwc10: well I'm hoping that the just-approved initiative to transform the IT department here is actually going to work. But then we'll still be a Microsoft dev shop. Hmm.
10:04 * mathw wonders if he can make everyone fall in love with F#. Or at least start using C# effectively.
10:05 nwc10 or get working on the .NET backend for NQP? :-)
10:06 masak antenoon, #perl6
10:06 moritz an, masak
10:07 masak grondilu: I don't see the significant advantage that inheritance confers in that situation.
10:07 woosley left #perl6
10:07 moritz is a Fork a Lock?
10:07 masak grondilu: seems to me you're only shuffling the pieces a bit, and then looking in admiration at 'self.' instead of '$!lock.'
10:08 masak moritz: with all respect, I think that's not the question to ask.
10:08 mathw nwc10: I'd then have to convince them of the value of Perl 6, which would be no easier.
10:08 masak moritz: because the answer to that depends on the model you choose, and that's what the discussion is all about.
10:09 masak moritz: grondilu proposes that a Fork be a Lock instead of haz a Lock. I'm not saying it's (esthetically or morally) *wrong* for it to be so, I'm just saying it's not significantly better AFAICS.
10:09 FROGGS joined #perl6
10:10 masak I tend to vastly prefer composition to inheritance. but in this case, I don't see the difference.
10:10 masak in fact, at some level or other, the inheritance will make a composition.
10:12 grondilu masak: the code is shorter
10:12 grondilu and conceptually simpler, imho
10:14 masak grondilu: and you say this, knowing full well about the recent move away from inheritance? :)
10:14 masak ...not just in Perl 6 or Perl, but in the whole OO industry.
10:14 moritz masak: maybe the better question to ask is: do you ever want to pass a Fork object to a routine that expects a Lock?
10:15 grondilu masak: I did not know about it, actually.  Good to know.
10:15 moritz maybe Lock should be a role :-)
10:15 masak moritz: yes, that is a better question to ask.
10:16 masak grondilu: let me see if I can find you a link.
10:16 grondilu Lock as a role would make sense, imho
10:16 masak grondilu: maybe start here? http://stackoverflow.com/questions/490​02/prefer-composition-over-inheritance
10:17 grondilu ok, thanks
10:18 kivutar joined #perl6
10:22 grondilu_ joined #perl6
10:35 [particle]1 joined #perl6
10:40 xinming joined #perl6
10:45 mathw the substition principle is the test I always use for inheritance. "Is it ever valid to pass a Subclass as a Superclass?" is a great question to ask when considering inheritance.
10:46 masak I can honestly not think of a situation where I need to pass in a Lock, and I go "hey, I have this Fork, let's pass that in!"
10:46 masak :P
10:46 mathw I'm not even sure what a Fork is
10:46 mathw So I can't weigh in on that one
10:46 masak Fork as in Dining Philosophers...
10:46 mathw ah
10:46 mathw hmmm
10:46 mathw I suppose you *could*, but it's a bit of a conceptual twist
10:47 mathw what you're really saying is that "This Fork has an internal Lock, I'll repurpose it", surely
10:47 masak again, it simply doesn't seem to confer any advantage over "a Fork haz a Lock".
10:48 mathw also with inheritance: it's far less necessary if you're in a language where you can apply 'class' or 'role' or 'interface' or somesuch to types that you're not defining yourself (i.e. Haskell typeclasses, Lisp multimethods...)
10:49 mathw always find it a shame when I have an object in C# that conforms to an interface I want to be able to pass around, but because its class definition doesn't say that I'm stuffed.
10:50 moritz that's what golang does differently :-)
10:52 masak mathw: I guess... the Adapter pattern helps you out there, though? retroactively providing something with an interface?
10:58 mathw masak: yeah, but it's a cludge
10:58 mathw like most patterns, it substitutes for something your language doesn't do
10:58 masak r: role Foo {}; class Foo {} # most definitely LTA error message -- via lizmat++
10:58 camelia rakudo-parrot e5fd34, rakudo-jvm e5fd34: OUTPUT«===SORRY!===␤No such method 'is_composed' for invocant of type 'Perl6::Metamodel::ParametricRoleGroupHOW'␤»
10:58 * masak submits rakudobug
10:59 mathw and as a Haskell programmer, not being able to just define a new typeclass and make everything I need be an instance of it pains me
10:59 moritz oh, I think I know what's going on
10:59 moritz because roles are multi, there's no redeclaration error
10:59 moritz though there should be one, because the second defintion is not a role
11:00 mathw woops
11:00 mathw I rather like that error message though, it's like C++.
11:00 jnthn I think it's actually detected that it's looking at a re-decl and is then trying to establish if the thing is a stub or not
11:01 jnthn Since is_composed is the way to detect that
11:01 jnthn But yeah, needs to check if the object *can* is_composed before actually calling it.
11:05 SamuraiJack__ joined #perl6
11:11 jnthn I suspect Lock.new.protect(...) is completely useless. :)
11:12 masak yay
11:13 masak I mean, "HOORAY!". http://yellow5.com/pokey/archive/
11:13 mathw jnthn: unless Lock.new.protect accesses, implicitly, the global interpreter lock
11:13 mathw we've got a global interpreter lock, right? All the best languages have one, I heard...
11:13 masak mathw: wouldn't that be fun!
11:13 jnthn wtf is a GIL :P
11:14 mathw "A design decision you will later come to regret"
11:14 mathw official definition
11:14 mathw ask the OED
11:15 masak the OED has a *TLA* for "GIL"? :P
11:22 jnthn :)
11:31 moritz OED is "original equipment destroyer", right? :-)
11:42 rindolf joined #perl6
12:09 tgt joined #perl6
12:25 Woodi I was trying to install nqp from R* 2013.11, backend moar, do I need to install it from somewhere or it is included in distribution tgz ?
12:26 grondilu joined #perl6
12:27 zakharyas joined #perl6
12:28 moritz Woodi: R* only supports the parrot backend at the moment.
12:28 moritz Woodi: nqp's Configure.pl happens to support --gen-moar
12:29 moritz (which will use git to fetch MoarVM)
12:32 Woodi ok, thanx
12:37 jnthn And note it doesnt' do anything useful yet :P
12:38 moritz well, it does lots of things that NQP on other backends does too
12:39 jnthn Oh, NQP on Moar is very capable, yes :)
12:39 jnthn I more meant Rakudo on Moar.
12:39 Woodi but R* should include m-clean Makefile target anyway :)
12:40 Woodi yes, I was hoping for some more Perl6 on moarvm :)
12:41 FROGGS Woodi: you need to wait a few hours/days :o)
12:42 moritz not for a R* on moar
12:42 FROGGS no, but for a perl6-m binary that can say "hello world"
12:42 FROGGS R* will take some months to happen
12:43 Woodi probably I do not know much (or 0 or even negative amount) about cross-compilation or what it is to bring language to live but having precompiled parts (stage0) is not ideal for me...
12:43 FROGGS my guess is that it will be summer by then, but nobody can predict it
12:44 FROGGS Woodi: why?
12:45 jnthn Woodi: It's not a cross-comp thing, it's a bootstrapping thing.
12:45 dalek rakudo/moar-support: e0d6ad3 | (Tobias Leich)++ | tools/build/Makefile-Moar.in:
12:45 dalek rakudo/moar-support: put correct libpath in perl6-m runner
12:45 dalek rakudo/moar-support: review: https://github.com/rakudo/rakudo/commit/e0d6ad32dc
12:45 Woodi I usually use C code to compile thing. I *think* that compiling gcc requires just C code in tarball...
12:45 jnthn Woodi: It also requires a C compiler :)
12:46 jnthn NQP is written in NQP, so we need an NQP compiler. That's what stage0 is.
12:46 Woodi yes, but gcc.tgz do not include it... so negative amount of knovledge ?
12:46 moritz Woodi: you can't compile gcc without a C compiler if the stage0 isn't included
12:46 FROGGS and stage0 is not platform dependant, so there is no problem with that
12:47 Woodi so how you make first nqp compiler ? that method should be distributed, IMO...
12:47 FROGGS if you already have an existing backend like parrot, and wanna spit out a stage0 for, say, moarvm, you build a cross compiler
12:48 jnthn Woodi: Once long ago the first NQP compiler was written in PIR
12:48 FROGGS this cross compiler goes away when nqp@moarvm can build itself
12:48 jnthn Just like the first C compiler was written in something non-C.
12:48 jnthn The reason gcc doesn't include a C compiler is 'cus your system already has one :)
12:49 FROGGS right, the very first version must be written by hand or another compiler
12:49 Woodi hmm, so you want to break with C, for good ? :)
12:50 Woodi I have C compiler I trust it. I want to go from this point tp Perl6 via source code...
12:52 FROGGS Woodi: what if there is nothing written in C, like for the JVM backend?
12:52 jnthn Woodi: Well, your stage1 and stage2 (which is installed) NQP aer built from source you can inspect. If you are really paranoid, you can trace the whole history back to the PIR version in version control, I guess. :)
12:52 FROGGS btw, if you dont trust nqp's stage0, then you don't trust nqp
12:53 moritz somebody did that with Haskell some time ago
12:53 Woodi you alway need to trust someone or something :) asm for first C, hardvare for asm...
12:53 moritz erm, GHC I mean
12:53 moritz which is also a bootstrapped compiler
12:54 SamuraiJack__ joined #perl6
12:56 Woodi jnthn: I know it is possible to trace all the way to pir... but this way will be longer and longer and in case someone on the way hide something (even in few steps) then no one will want to do this (except excentric scientists or government) and trust for language will be lost...
12:58 Woodi the proprer way is shortest way from hardware...
13:00 Woodi how about other nqp implementations ? js for example ? it can bootstrap language on it's own ?
13:00 moritz if you really thing "the proprer way is shortest way from hardware", you should program direct in native executable formats, not in high-level languages
13:01 Woodi moritz: no, I should not :) thes days cpu are cheap, programmers are expensive :)
13:02 Woodi but to have decent preformance you still need to be clode to hw...
13:02 Woodi close*
13:02 jnthn Woodi: Yes, so stop expecting the very over-busy NQP programmers to maintain a second version of their system in something other than NQP.
13:02 jnthn :P
13:03 jnthn And yes, afaik JS version was getting towards being able to bootstrap too, if it didn't already get there.
13:03 FROGGS Woodi: I would not trust hardware btw, see hidden code in RNG or how easy it is to turn your mobile into something I can control (using the firmware of your boardband processor)
13:04 Woodi jnthn: I do installing all yours code, I trust you all. just wanting perfecter world :)
13:04 jnthn In a world where most people install their OS, browser, etc from bianries ratehr than building soruce, I really don't think we have to worry too much over people not wanting to use Perl 6 'cus it has a compiled stage0, like other bootstrapped langauges. :)
13:04 jnthn I the end, Windows folks will tend to grab an MSI, and others will use their favorite package manager...
13:05 jnthn *In
13:05 moritz and the hipsters will do a  wget -O - http://upgrade.rakudo.me | bash
13:06 jercos nah man, hipsters use curl
13:06 moritz jercos: you're right, but I was too lazy to look up curl's command line syntax :-)
13:06 jercos oddly enough, curl defaults to STDOUT
13:06 jercos :D
13:07 * jnthn adds "know curl syntax" to list of things requierd to be a hipster
13:07 jercos *Linux hipster
13:07 Woodi FROGGS: I do not trust hardware :) I use it with a smile :)
13:07 FROGGS hehe
13:07 jercos presumably there are plenty of hipsters out there who don't know a command line from a nazi
13:08 jnthn Most likely they did nazi a comamnd line before...
13:08 jnthn ...um, that pun is a stretch in text. :)
13:08 * jnthn gets back to his teaching...
13:09 jercos anne frankly without the pun it's just offensive.
13:10 Woodi my belief is that at the END knowledge will be finished and everything will be known :)
13:13 moritz the other day I was in an ultra-hipster restaurant. It had compostable one-time cutlery made from starch flour plastic or so
13:16 nwc10 jnthn: I'm still suspicious that you need one of these, to properly view the callbacks: http://www.gizmag.com/kohjinsha-d​ual-screen-laptop-on-sale/13479/
13:37 mathw oh, that's why I'd use curl instead of wget. Simpler syntax for stdout!
13:37 mathw also more unixy
13:38 moritz also maintained
13:38 FROGGS you don't need to maintain a software that just works :o)
13:39 moritz ... until it stops working.
13:39 FROGGS blasphemy!!
13:41 mathw hahaha
13:43 * raydiak wonders how much more creative he has to get with his nick until he stops running into sites where it's already taken :P
13:45 FROGGS :/
13:45 moritz raydiak: man pwgen # :-)
13:45 FROGGS mine is always taken too
13:45 FROGGS hehe
13:45 nwc10 there is only one schmorp :-)
13:46 mathw too many people online :(
13:47 raydiak hmmm...good point...several strategic EMPs might free up my name on wordpress.com...
13:49 FROGGS but please save the europeans :)
13:52 Woodi could we just stop making wars and do something usefull instead ? eg. [re]writing libraries of usefull code to cut down IT costs or something :)
13:52 Woodi usefull functions* of course :)
13:52 raydiak not interested in international conflict.  just my username.  :)
13:53 Woodi ok, I go google it now :)
14:00 dalek rakudo/moar-support: d01d6dc | (Tobias Leich)++ | tools/build/gen-moar-runner.pl:
14:00 dalek rakudo/moar-support: correctly create maor runner on windows
14:00 dalek rakudo/moar-support: review: https://github.com/rakudo/rakudo/commit/d01d6dc0d1
14:02 kaare_ joined #perl6
14:03 tgt joined #perl6
14:06 raydiak I suppose it would have been more accurate to say "sickened beyond recourse"..."not interested" sounds terribly irreverent of human life when applied to international relations
14:07 prevost joined #perl6
14:13 PacoAir joined #perl6
14:16 fhelmberger_ joined #perl6
14:20 rurban1 joined #perl6
14:21 c1sung joined #perl6
14:26 perl6 joined #perl6
14:34 langdon_ joined #perl6
14:37 langdon_ perl6: Array[Array[Bool]].new($(Array[Bool].new))
14:37 camelia rakudo-parrot e5fd34: OUTPUT«Type check failed in .new; expected 'Array[Bool]' but got 'Array[Bool]'␤  in block  at gen/parrot/CORE.setting:8300␤  in method reify at gen/parrot/CORE.setting:7244␤  in method reify at gen/parrot/CORE.setting:7139␤  in method gimme at gen/parro…»
14:37 camelia ..rakudo-jvm e5fd34: OUTPUT«Type check failed in .new; expected 'Array[Bool]' but got 'Array[Bool]'␤  in block  at gen/jvm/CORE.setting:8271␤  in block  at gen/jvm/CORE.setting:7259␤  in block  at gen/jvm/CORE.setting:7123␤  in method gimme at gen/jvm/CORE.setting:7541␤  in …»
14:37 camelia ..niecza v24-98-g473bd20: OUTPUT«Unhandled exception: System.NullReferenceException: Object reference not set to an instance of an object␤  at Builtins.InvokeSub (Niecza.P6any obj, Niecza.Variable[] pos) [0x00000] in <filename unknown>:0 ␤  at Niecza.Kernel.ToComposable (Niecza.ST…»
14:38 xinming joined #perl6
14:39 c1sung joined #perl6
14:39 mathw has anybody ever done print-format document generation and not gone insane?
14:41 langdon_ expected 'Array[Bool]' but got 'Array[Bool]'
14:41 timotimo oh wow :/
14:41 tgt joined #perl6
14:42 timotimo that's ... unhelpful ...
14:42 langdon_ yep
14:42 timotimo the problem is that if you build $(...), it won't give a type annotation
14:42 timotimo perl6: my $test = $(Array[Bool].new); $test.push: 1; say $test;
14:42 camelia niecza v24-98-g473bd20: OUTPUT«Unhandled exception: System.NullReferenceException: Object reference not set to an instance of an object␤  at Builtins.InvokeSub (Niecza.P6any obj, Niecza.Variable[] pos) [0x00000] in <filename unknown>:0 ␤  at Niecza.Kernel.ToComposable (Niecza.ST…»
14:42 camelia ..rakudo-jvm e5fd34: OUTPUT«Type check failed in .push; expected 'Bool' but got 'Int'␤  in block  at gen/jvm/CORE.setting:7634␤  in method push at gen/jvm/CORE.setting:7620␤  in block  at /tmp/7WuvVcCQhL:1␤  in any eval at gen/jvm/stage2/NQPHLL.nqp:1086␤  in any evalfiles at…»
14:42 camelia ..rakudo-parrot e5fd34: OUTPUT«Type check failed in .push; expected 'Bool' but got 'Int'␤  in block  at gen/parrot/CORE.setting:7663␤  in method push at gen/parrot/CORE.setting:7646␤  in method push at gen/parrot/CORE.setting:1562␤  in block  at /tmp/7VnDjUKwOW:1␤  in any  a…»
14:42 timotimo er ...
14:43 timotimo that seems to have flattened.
14:43 timotimo perl6: my @test = $(Array[Bool].new); @test.push: 1; say @test;
14:43 camelia rakudo-parrot e5fd34, rakudo-jvm e5fd34: OUTPUT« 1␤»
14:43 camelia ..niecza v24-98-g473bd20: OUTPUT«Unhandled exception: System.NullReferenceException: Object reference not set to an instance of an object␤  at Builtins.InvokeSub (Niecza.P6any obj, Niecza.Variable[] pos) [0x00000] in <filename unknown>:0 ␤  at Niecza.Kernel.ToComposable (Niecza.ST…»
14:43 timotimo there we go, that's better
14:43 timotimo so you see, just by making a list that has Array[Bool].new as its only item, it won't generate a container that says "only Array[Bool] may go in here!"
14:51 jnthn nwc10: I think that kind of laptop is standard issue for Node.js developers :P
14:54 ajr joined #perl6
15:03 benabik joined #perl6
15:24 timotimo i wonder how long until it makes sense to build an optimisation to turn big if trees or simple given/whens into things like computed jumps
15:26 jercos the lack of case structures in perl 5 had me using a lot of calls to subs in hashes... which of course, brings a similar increase in speed, though perhaps not quite as potent.
15:27 nwc10 do profiling tools yet exist to identify hot spots?
15:27 timotimo they kind of do, but ... :\
15:28 timotimo as you can tell by my constant guessing instead of measuring for attempting optimisations, i can't make heads nor tails out of kcachegrind :(
15:28 nwc10 strikes me that part of the answer to your question is "once we can measure it"
15:28 timotimo that does indeed sound sensible
15:29 jercos 'twould perhaps be simpler if the language offered an explicit multiple choice statement, more caselike than if chains, but less smartmatchy than given/when.
15:30 timotimo the thing is, if you just put constants into the whens, it's essentially the same :)
15:30 kaleem joined #perl6
15:30 * timotimo has an idea
15:31 timotimo when * < 20 { } is currently implemented as a call to a block
15:31 timotimo the optimizer doesn't inline it yet
15:34 TimToady jercos: that is precisely why the semantics of the switch are governed by the 'when', not the 'given'
15:35 jercos Indeed.
15:36 TimToady and it was the desire for that optimization that caused us to break the symmetry of the old smartmatch
15:36 TimToady Perl 4 actually did the jump optimization on switch-like structures, but I don't thing Perl 5 ever attempted it.
15:37 TimToady though the opcodes for it were in there for at least a while
15:37 timotimo jnthn: where do i find the code that decides wether or not to keep the ast of a block around for later in-lining? or something like that?
15:40 jnthn look for inline_info in setting
15:40 jnthn um
15:40 jnthn look for inline_info in actions
15:40 timotimo thanks :)
15:42 jnthn teaching done; time to go get the train :)
15:42 jnthn bbl
15:42 TimToady oh, I misread jercos++, who was speaking about Perl 5 in the first place, durr
15:42 TimToady o/
15:43 TimToady but Perl 5 is a language, and since we do talk about linguistics here, it's not off topic  :)
15:43 timotimo that was a close one, we would almost have had to kick jercos out
15:45 nnunley joined #perl6
15:47 hoelzro is NativeCall able to handle foreign functions that require native callback yet?
15:47 tadzik at least in parrot, yes
15:47 tadzik hoelzro, hoelzro when are we playing l4d2?
15:48 hoelzro hmm
15:48 arnsholt hoelzro: Callbacks are supported
15:48 hoelzro there has been talk of l4d2 in my home lately
15:48 hoelzro but no playing of it yet
15:48 hoelzro tadzik: one of the coming weekends, I think ;)
15:48 tadzik hoelzro: steam id?
15:48 arnsholt We don't support invoking function pointers returned from native code yet, but the other way should work fine
15:48 * hoelzro checks
15:48 tadzik there's apparently more of us on the channel :)
15:49 hoelzro ah yes
15:49 hoelzro my steam id
15:49 hoelzro is
15:49 hoelzro hoelzro.
15:49 hoelzro arnsholt: so do_this_after_a_second(void (*cb)(void *)) works
15:50 arnsholt If you need to be able to invoke function pointers returned from C, I think that should be relatively easy to fix too
15:50 hoelzro but not void (*)(void *) heres_something_to_pass_the_time(void)?
15:50 hoelzro I just need the former
15:50 arnsholt Exactly
15:50 hoelzro if I ever get off my butt and work on this stuff =P
15:50 xinming joined #perl6
15:50 arnsholt Yeah, the latter is unimplemented because I didn't see a pressing need for it
15:50 timotimo hoelzro is apparently not publically findable
15:50 hoelzro hmm
15:50 arnsholt The other way around OTOH was kind of important =)
15:50 hoelzro I'll change that when I get home tonight!
15:51 timotimo :)
15:51 tadzik timotimo: it is for me
15:51 timotimo huh, weird
15:51 timotimo well, if you have him in your friends list, i can find him via you
15:55 xinming joined #perl6
15:58 Psyche^_ joined #perl6
16:01 xinming joined #perl6
16:02 araujo joined #perl6
16:09 atroxaper joined #perl6
16:13 xinming joined #perl6
16:37 FROGGS joined #perl6
16:37 dalek rakudo-star-daily: 6cb5be3 | coke++ | log/ (5 files):
16:37 dalek rakudo-star-daily: today (automated commit)
16:37 dalek rakudo-star-daily: review: https://github.com/coke/rakudo​-star-daily/commit/6cb5be318d
16:37 dalek rakudo-star-daily: bf52277 | coke++ | log/ (5 files):
16:37 dalek rakudo-star-daily: today (automated commit)
16:37 dalek rakudo-star-daily: review: https://github.com/coke/rakudo​-star-daily/commit/bf52277069
16:37 dalek rakudo-star-daily: c147794 | coke++ | log/ (5 files):
16:37 dalek rakudo-star-daily: today (automated commit)
16:37 dalek rakudo-star-daily: review: https://github.com/coke/rakudo​-star-daily/commit/c147794c4d
16:37 dalek perl6-roast-data: c425356 | coke++ | / (3 files):
16:37 dalek perl6-roast-data: today (automated commit)
16:37 dalek perl6-roast-data: review: https://github.com/coke/perl6​-roast-data/commit/c42535681b
16:37 dalek perl6-roast-data: bff7276 | coke++ | p (2 files):
16:37 dalek perl6-roast-data: today (automated commit)
16:37 dalek perl6-roast-data: review: https://github.com/coke/perl6​-roast-data/commit/bff72764b4
16:39 timotimo wow, so many days today
16:43 spider-mario joined #perl6
16:53 spider-mario joined #perl6
16:53 c1sung joined #perl6
16:54 moritz p: say Date.today.comb(/\d+/).join.Int.is-prime
16:54 camelia rakudo-parrot e5fd34: OUTPUT«No such method 'comb' for invocant of type 'Date'␤  in block  at /tmp/12A4cFCpzW:1␤  in any  at /tmp/12A4cFCpzW:1␤  in any  at gen/parrot/stage2/NQPHLL.nqp:1146␤  in any eval at gen/parrot/stage2/NQPHLL.nqp:1133␤  in any evalfiles at gen/parrot…»
16:54 moritz p: say Date.today.Str.comb(/\d+/).join.Int.is-prime
16:54 camelia rakudo-parrot e5fd34: OUTPUT«True␤»
16:54 moritz p: say 29122013.is-prime
16:54 camelia rakudo-parrot e5fd34: OUTPUT«False␤»
16:55 ajr joined #perl6
16:55 moritz p: say 28122013.is-prime
16:55 camelia rakudo-parrot e5fd34: OUTPUT«True␤»
16:55 moritz p: say 12282013.is-prime
16:55 camelia rakudo-parrot e5fd34: OUTPUT«False␤»
16:56 moritz oh, it's not december yet :-)
16:57 moritz p: say (11292013, 29112013, 20131129)>>.is-prime
16:57 camelia rakudo-parrot e5fd34: OUTPUT«False False True␤»
16:57 moritz p: say Date.today
16:57 camelia rakudo-parrot e5fd34: OUTPUT«2013-11-29␤»
16:58 moritz p: say Date.today.Str.comb(/\d+/).join.Int
16:58 camelia rakudo-parrot e5fd34: OUTPUT«20131129␤»
17:00 moritz p: say (112913, 291113, 20131129)>>.is-prime
17:00 camelia rakudo-parrot e5fd34: OUTPUT«True True True␤»
17:02 TimToady p: say 1129.is-prime, 2911.is-prime
17:02 camelia rakudo-parrot e5fd34: OUTPUT«TrueFalse␤»
17:32 retupmoca is github:supernovus in here?
17:34 timotimo what have i done wrong?!
17:34 timotimo oh, wait, i'm the wrong timo :)
17:34 FROGGS retupmoca: he was, ages ago
17:34 stevan_ joined #perl6
17:37 retupmoca alright
17:37 retupmoca just did a pull request, wanted to poke him about it
17:45 [Sno] joined #perl6
17:52 SamuraiJack__ joined #perl6
17:53 FROGGS[mobile] joined #perl6
18:38 timotimo if anyone here cares about good games, spelunky is currently 75% off on steam. unfortunately there's no linux or mac port
18:39 timotimo i'm hoping it'll get picked up by the humble bundle guys and ported to mac and linux in the process, but i have no evidence to support that hope at all
18:53 lizmat joined #perl6
18:54 FROGGS my class Dummy {
18:54 FROGGS our proto method AUTOGEN(::T $: |) { * }
18:54 FROGGS }
18:54 FROGGS wth is the $: there?
18:54 timotimo it's an anonymous invocant
18:54 FROGGS ahh
18:54 FROGGS of course
18:56 timotimo are we even looking at the right line?
18:57 FROGGS seems so
18:57 FROGGS I get the same msg when I just run that blok
18:57 FROGGS block*
18:57 FROGGS umm, stupid me
18:59 FROGGS (it still stumbled in the setting, not my sript)
18:59 FROGGS script*
19:06 jnthn :P
19:06 jnthn I'm pretty sure it's the lack of reposession stuff like I mentioned last night
19:07 nwc10
19:08 woolfy joined #perl6
19:08 FROGGS jnthn: if it works for you I'd like to merge hope into moarvm master
19:08 jnthn FROGGS: If...what works? :)
19:09 FROGGS hope :o)
19:09 * TimToady wonders if he should have a commit bit to rakudo...
19:10 * TimToady turned in a CLA about ten years ago...
19:10 jnthn .oO( "Did you send in a CLA?" :D :D )
19:10 jnthn :P
19:10 jnthn TimToady: Then you can have one. :)
19:10 jnthn FROGGS: You...called a branch "hope"? :P
19:10 FROGGS yeah *g*
19:11 jnthn Will the next one be "glory"? :P
19:11 jnthn git checkout hope # ...
19:11 FROGGS since... I spent already four or five days with that thing a few months back, and now another one... so... that was its last chance to get fixed by me
19:12 FROGGS I hope there are no such beasts anymore
19:12 * jnthn is make installing
19:12 FROGGS jnthn: I fixed the runner in rakudo btw
19:12 FROGGS nqp is left untouhed
19:12 jnthn yay
19:12 FROGGS untouched*
19:12 jnthn Should I try building NQP anyway?
19:12 FROGGS hmmm, would be nice, yeah
19:13 jnthn ok, set it off
19:13 jnthn Trying builds is easy thing to do on the train.
19:13 FROGGS can you remember if you had problems using the debugger last time I closed handles?
19:13 * jnthn is surfing "med full fart!" again :)
19:13 FROGGS *g*
19:14 jnthn I don't think I had that problem...
19:14 * FROGGS imagines jnthn++ sitting in a train with his laptop on his knees blowing in the wind...
19:14 FROGGS k
19:15 FROGGS I dont think there will be problems since std filehandles won't be closed, only common filehandles will
19:16 jnthn NQP build, Rakudo build is underway
19:16 FROGGS cool
19:16 FROGGS I'll enjoy a Malzbier if that works out
19:17 jnthn I hope it won't Malzfunction, then...
19:17 FROGGS >.<
19:17 jnthn Reached CORE.setting
19:17 FROGGS damn, your machine is fast
19:17 jnthn And this is my slower one :P
19:17 FROGGS gah!
19:18 FROGGS you still need to tune it :o)
19:18 lue TimToady: Is it just me, or does Mutex not have a Perl 6 section?
19:18 FROGGS no excuses!
19:18 jnthn Not by much, though...
19:18 TimToady it's just you :)
19:18 jnthn lue: Didn't for me at first, then I refrash and it was there...
19:18 jnthn Caching I guess.
19:18 lue Huh. Hitting F5 helped. So yeah, I blame caching.
19:20 FROGGS that is how the interwebs work :o)
19:20 FROGGS if I comment out the curront blow-ups in the setting, I get:
19:20 FROGGS Type check failed in bind; expected Callable but got Block
19:20 FROGGS at src/vm/moar/Perl6/Ops.nqp:197  (blib/Perl6/Ops.moarvm:bind_error:42)
19:20 FROGGS from src/gen/m-CORE.setting:701  (./CORE.setting.moarvm:frame_name_1647:6931)
19:21 FROGGS dunno if that will be the next or if it is related
19:21 FROGGS current*
19:21 jnthn It's related, I suspect
19:21 jnthn I mean, I can directly attribute a way that could happen to the lack of reposession too
19:22 FROGGS jnthn: btw, we'd need to either gc_free objects when panicking in moarvm or need to walk the handles explicitly to hide libuv's nice message when we crash
19:22 FROGGS k
19:23 FROGGS I don't care about that atm though
19:25 jnthn I don't see any problem with hope.
19:27 jnthn As in, I get to the same point as I did without it
19:27 FROGGS \o/
19:27 jnthn perl6-m.bat gives me a different error than before though
19:28 FROGGS O.op
19:28 FROGGS -p
19:29 jnthn Ahl, it passes --lib-path as '.'
19:29 FROGGS you need to pull in rakudo then
19:29 woolfy It's dark in London. Tomorrow is LPW.  I'm hungry.  None of this is related to Perl 6.
19:30 TimToady how do you know?
19:30 FROGGS woolfy: there is no such thing as off topic :o)
19:30 woolfy Hehe...
19:31 TimToady OT is just one of those vagrancy laws we use to chase off people we don't like for other reasons...
19:31 FROGGS *g*
19:31 jnthn FROGGS: I did, ut should I reconfigure?
19:31 FROGGS I, I think so
19:32 FROGGS jnthn: yes, you need
19:32 FROGGS need to
19:32 FROGGS because I patched the Makefile.in to pass the libpath
19:32 jnthn k
19:33 jnthn woolfy: I heard that London hsa food!
19:33 * FROGGS wishes he was in scotland again :/
19:34 jnthn I wishky I was too...
19:34 xinming joined #perl6
19:34 xenoterracide joined #perl6
19:35 moritz Land of my high endeavor, land of the shining river, land of my heart forever, Scotland the brave!
19:36 TimToady maybe each backend should have its own national anthem on STDIN
19:37 FROGGS where was moarvm founded?
19:37 moritz p: say for lines
19:37 camelia rakudo-parrot e5fd34: OUTPUT«␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤␤»
19:37 moritz ugh.
19:37 FROGGS :/
19:37 moritz .say for lines
19:37 TimToady yipes
19:37 moritz p: .say for lines
19:37 camelia rakudo-parrot e5fd34: OUTPUT«Stille Nacht! Heilige Nacht!␤Alles schläft; einsam wacht␤Nur das traute heilige Paar.␤Holder Knab im lockigten Haar,␤Schlafe in himmlischer Ruh!␤Schlafe in himmlischer Ruh!␤␤Stille Nacht! Heilige Nacht!␤Gottes Sohn! O wie lacht␤Lieb´ …»
19:37 moritz it should complain about say without arguments, no? :-)
19:37 TimToady hah
19:37 FROGGS lockigten
19:38 dalek evalbot: 80b9f0d | moritz++ | stdin:
19:38 dalek evalbot: typo, FROGGS++
19:38 dalek evalbot: review: https://github.com/perl6/evalbot/commit/80b9f0d97d
19:38 * moritz just copied and pasted it from somewhere
19:39 FROGGS eww
19:39 TimToady hmm, I learned it "Nur das traute hochheilige Paar"
19:39 moritz TimToady: that's the more common version, yes
19:39 FROGGS moritz: that might be even the original text an not a typo
19:40 FROGGS I only can find this version
19:40 * moritz shrugs
19:40 xinming joined #perl6
19:40 colomon hoch is high?
19:41 FROGGS yes
19:41 TimToady ja
19:43 colomon moritz++ for changing the default text to something seasonal.
19:43 ssutch joined #perl6
19:43 woolfy It seems lizmat and diakopter are going to take care of my hunger.  It seems the hunger will be killed by Indian food.  Maybe some whisky too.
19:43 colomon whisky++
19:44 jnthn \o/
19:44 lue .oO(Aber es ist nicht noch Dezember, ja?)
19:44 moritz "ja ist denn heut' schon weihnachten?"
19:45 dalek evalbot: 42b3b4c | moritz++ | evalbot.pl:
19:45 dalek evalbot: another control command (because I am lazy)
19:45 dalek evalbot: review: https://github.com/perl6/evalbot/commit/42b3b4cd96
19:45 moritz evalbot control restart
19:46 camelia joined #perl6
19:46 colomon lue: as far as I'm concerned it's open season for Christmas music as of today.  Already listened to the new Bellowhead Christmas single about six times.  (To be fair, half of that was "Play it again!" from my son.)
19:47 moritz also with small children you have to start a bit earlier, otherwise they don't have a chance to learn the lyrics :-)
19:47 benabik joined #perl6
19:48 lue colomon: Yeah, I was just pointing it out for fun, considering December's only a couple days away :)
19:48 colomon actually Henry requested Jingle Bells a month ago already.  But he was satisfied listening to it three or four times in a row.
19:49 lizmat fwiw, all hotels / pubs / restaurants we've seen the past days, had christmas decorations or were in the process of putting them on
19:49 colomon we've got just enough snow here to make everything white, too.
19:50 jnthn My hotel in Stockholm decorated for it while I was there.
19:50 jnthn And today the fridge at the teaching place suddenly had jul must.
19:50 lizmat a must have, jul!
19:51 TimToady jnthn: sure, but that just means putting up some of those ⋀ lights in the window :)
19:51 woolfy oO(hunger will be vanquished) <- (diakopter is on way to our hotel) -> (ow, we must go downstairs now) -> bye
19:51 TimToady or has Sweden graduated to greater glories in the last decade?
19:52 lizmat dinner&
19:53 lue December's the time for building ☃ and watching that ☄ shower that happens every year abouts.
19:53 jnthn Well, it wsa mostly greenery and even more candles that I saw... :)
20:00 zakharyas joined #perl6
20:04 dalek rakudo/moar-support: 95eae40 | (Tobias Leich)++ | tools/lib/NQP/Configure.pm:
20:04 dalek rakudo/moar-support: fix --gen-moar=OPTION
20:04 dalek rakudo/moar-support: review: https://github.com/rakudo/rakudo/commit/95eae40ffd
20:04 FROGGS sorry TimToady, that was a stupid thinko of me :(
20:07 dalek rakudo/moar-support: eb24cea | (Tobias Leich)++ | Configure.pl:
20:07 dalek rakudo/moar-support: respect --gen-nqp if it was passed
20:07 dalek rakudo/moar-support: review: https://github.com/rakudo/rakudo/commit/eb24ceab00
20:10 * FROGGS .oO( already used the to be specced ¿¿ ¡¡ syntax )
20:12 berekuk joined #perl6
20:14 TimToady ˙°O(¡sdoo)
20:15 lue FROGGS: what is that meant to do? :)
20:15 TimToady (¡sdoo)O°˙
20:16 FROGGS lue: see the fix and judge yourself :o)  https://github.com/rakudo/rakudo/commit/95eae40ffd
20:16 jnthn There will be modules... :P
20:17 FROGGS with a single line in it, where other languages need hundreds :o)
20:18 lue FROGGS: I was referring to your ¿¿¡¡ comment
20:18 hummeleB1 joined #perl6
20:19 FROGGS lue: it is about the same, I mixed up the "if" and the "else" part...
20:20 FROGGS like ¿¿¡¡ would do the opposite of ??!!
20:20 * lue thinks again of    why { do_stuff; } because ( $x ~~ False )
20:21 FROGGS why { $x + $y } because ( $y < 10 )
20:22 FROGGS "why { $x + $y }" does not feel right
20:22 FROGGS why starts a question, but $x + $y is not part of it
20:23 FROGGS do { $x + $y } because ( $y < 10 )
20:23 FROGGS that feels righter to me
20:23 * FROGGS .oO( but then I took an ➴ to the knee )
20:26 TimToady Q: What's worse than bootstrapping a compiler?
20:26 TimToady A: Bootstrapping two compilers!
20:26 TimToady Q: What's worse than that?
20:26 nwc10 "Bootstrapping three compilers!" I guess
20:26 nwc10 and, how could we make it even better?
20:27 TimToady um, target .NET, .js, golang, etc....
20:28 TimToady half-hour later, the *Configure* is done...
20:29 FROGGS :/
20:29 nwc10 what did it build?
20:29 TimToady on the bright side, it seems to have configured
20:30 jnthn half an hour?!
20:30 TimToady well, it bootstrapped nqp on parrot, jvm, and moar
20:30 jnthn Does --gen-moar pass --optimize to Moar? :)
20:30 TimToady beats me
20:30 jnthn Was more asking FROGGS++ :)
20:31 FROGGS jnthn: it does not currently
20:31 stevan_ joined #perl6
20:32 jnthn I guess other option is make it the default in Moar's configure.
20:32 TimToady maybe that would be premature, if you want good debugging for now
20:32 jnthn You only really want a debug build if you're developing...
20:33 FROGGS ahh, it does
20:33 FROGGS nqp does it
20:33 jnthn FROGGS: ok :)
20:33 preflex jnthn: you have 3 new messages. '/msg preflex messages' to read them.
20:33 TimToady that would explain why my gdb trace was...terse...
20:33 FROGGS it passes --optimize in sub gen_moar
20:33 FROGGS >.<
20:34 FROGGS TimToady: the good thing is that you dont need to recompiler anything when reconfiguring/installing moarvm without --optimize
20:34 TimToady or is this an optimize flag aimed higher than C
20:35 lue FROGGS: it can change to why { $x + $y }? because ( $y < 10 ) if you like :)
20:35 FROGGS recompile*
20:35 FROGGS hehe
20:36 jnthn TimToady: No, --optimize in Moar is all about C
20:36 TimToady ooh, got past m-BOOTSTRAP!!!
20:36 TimToady now in stage parse
20:36 FROGGS \o/
20:36 colomon \o/
20:36 FROGGS that is the time where you want to make tee
20:36 TimToady 'course, it decided to do moar before parrot, so I'll probably end up sans parrot...
20:36 FROGGS (to drink it in stage mbc)
20:36 TimToady er, sans perl6-p
20:37 FROGGS make p-install is your friend
20:37 lue I get past stage mbc before it fails. Is that good?
20:38 FROGGS yes
20:38 FROGGS how does it fail ooc?
20:41 lue moar: 3rdparty/libuv/src/unix/loop.c:150: uv__loop_delete: Assertion `!((((*(&(loop)->active_reqs))[0]) == (&(loop)->active_reqs)) == 0)' failed.
20:41 lue (this was last night, so perhaps it's changed since.)
20:41 FROGGS lue: no, there must be an error message earlier
20:41 tgt joined #perl6
20:41 FROGGS it should have changed, yes
20:42 lue I would've given you the preceding error if there was one :) It goes  Stage mbc        :   8.069
20:42 lue moar: 3rdparty/libuv/src/unix/loop.c:150:
20:42 * lue will run the process again now, see what changes.
20:43 FROGGS I hope there is something else
20:43 jnthn 8.069?! Wow. It is like 0.554 here :)
20:43 lue jnthn: I think I was running other things on my computer at the time, it would've contributed to a slowdown.
20:43 TimToady yeah, well, my stage parse is 288 :/
20:44 FROGGS 184.777s here
20:44 lue TimToady: Mine was in the 1000's. Again, other stuff running.
20:44 lue (Incidentally I'm sure I'm running the same 'other stuff' now...)
20:48 telex joined #perl6
20:48 TimToady and stage mast takes 343
20:48 TimToady and, of course, it couldn't compile Test
20:48 FROGGS 220, and then "Cannot invoke null object"
20:49 TimToady my mbc was only .955  :)
20:49 FROGGS TimToady: it is unable to load the setting
20:49 FROGGS and only that strangely farting guy can fix that :P
20:49 TimToady yes, it's about where I'd expected it to fail from the reports
20:50 FROGGS I'd love to profile it, but when using goodle-pprof moar complains about invalid utf8
20:50 TimToady it was running in just under a gig, but makes me wonder whether it'd run faster with more memory
20:51 FROGGS don't think so, jnthn and timotimo both said it would take 950MB peak
20:51 * TimToady got up to 990 or so
20:51 jnthn detrain &
20:52 FROGGS my times are on a "Intel(R) Core(TM) i5-3320M CPU @ 2.60GHz" with turbo stuff enabled
20:53 TimToady anyway, for the record, succeeded with: perl Configure.pl --prefix="/home/larry/nom/install" --backends='jvm,parrot,moar' --gen-nqp=master --gen-moar=master --gen-parrot
20:53 TimToady eventually...
20:53 FROGGS that is nice that it works now
20:54 TimToady FROGGS++
20:54 FROGGS :o)
20:54 sjohnson heh
20:54 sjohnson that emoticon is cute.
20:55 FROGGS so am I *g*
20:55 FROGGS (according to my wife)
20:55 lue Once rakudo successfully compiles on moar, I'll start pestering MoarVM to default install to /usr/local, like every other project ever (even parrot!) . :)
21:02 xinming joined #perl6
21:15 cognominal joined #perl6
22:02 lue FROGGS: I got a more sensible "Cannot invoke null object" error now.
22:02 FROGGS lue: that is what I wanted to hear!
22:02 FROGGS :o)
22:02 lue (and no libuv assertion either!)
22:03 FROGGS hmmm, there should be one... but it doesn't hurt if it does not show up :o)
22:14 rurban1 joined #perl6
22:18 xinming joined #perl6
22:21 FROGGS rp: say "abc" ~~ / <ident> { make 42 } /; say $<ident>
22:21 camelia rakudo-parrot e5fd34: OUTPUT«「abc」␤ ident => 「abc」␤␤「abc」␤␤»
22:21 FROGGS rp: say "abc" ~~ / <ident> { make 42 } /; say $/.ast
22:21 camelia rakudo-parrot e5fd34: OUTPUT«「abc」␤ ident => 「abc」␤␤42␤»
22:21 ajr joined #perl6
22:35 preflex joined #perl6
22:37 FROGGS rp: sub a (:*@b) { }
22:37 camelia rakudo-parrot e5fd34: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/YAMqae4Oxrâ�¤Malformed parameterâ�¤at /tmp/YAMqae4Oxr:1â�¤------> [32msub a ([33mâ��[31m:*@b) { }[0mâ�¤    expecting any of:â�¤        statement listâ�¤        prefix or termâ�¤        prefix …»
22:37 FROGGS rp: class C { method a (:*@b) { } }
22:37 camelia rakudo-parrot e5fd34: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/Kf9dOiwIx0â�¤Malformed parameterâ�¤at /tmp/Kf9dOiwIx0:1â�¤------> [32mclass C { method a ([33mâ��[31m:*@b) { } }[0mâ�¤    expecting any of:â�¤        statement listâ�¤        prefix or termâ�¤â€¦Â»
22:37 FROGGS rp: class C { method a (:%h; :*@b) { } }
22:37 camelia rakudo-parrot e5fd34: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/ER9vSyZifzâ�¤Malformed parameterâ�¤at /tmp/ER9vSyZifz:1â�¤------> [32mclass C { method a (:%h; [33mâ��[31m:*@b) { } }[0mâ�¤    expecting any of:â�¤        statement listâ�¤        prefix or te…»
22:38 xinming joined #perl6
22:39 c1sung joined #perl6
22:43 yeltzooo6 joined #perl6
22:46 yeltzooo3 joined #perl6
22:47 FROGGS rp: my $x = "abc"; $x.=subst(m/a/, 'x')
22:47 camelia rakudo-parrot e5fd34: OUTPUT«Cannot call 'match'; none of these signatures match:␤:(Cool:D : $target, *%adverbs)␤  in any  at gen/parrot/BOOTSTRAP.nqp:1218␤  in any  at gen/parrot/BOOTSTRAP.nqp:1209␤  in method match at gen/parrot/CORE.setting:3435␤  in block  at /tmp/h32r…»
22:47 FROGGS rp: my $x = "abc"; $x.=subst(/a/, 'x')
22:47 camelia rakudo-parrot e5fd34: ( no output )
23:02 FROGGS if [&&] @provides>>.ACCEPTS($file) -> $/ { # I love it
23:02 kivutar joined #perl6
23:03 yeltzooo joined #perl6
23:06 hugme joined #perl6
23:10 ssutch joined #perl6
23:10 tadzik joined #perl6
23:13 Juerd joined #perl6
23:14 timotimo [&&] passes through the rightmost match?
23:14 Juerd feather1 and 2 are back
23:14 Juerd 3 is still booting.
23:14 FROGGS timotimo: yeah
23:16 FROGGS timotimo: there is only one match anyway, which comes back as a parcel
23:16 FROGGS somehow .flat doesnt work, but fortunately [&&] does
23:17 c1sung joined #perl6
23:19 FROGGS hmmm, I better use [||]
23:23 sricloud joined #perl6
23:25 BenGoldberg joined #perl6
23:27 fridim__ joined #perl6

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

Perl 6 | Reference Documentation | Rakudo