»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'perl6: say 3;' or rakudo:, niecza:, std:, or /msg camelia perl6: ... | irclog: irc.perl6.org | UTF-8 is our friend!
Set by sorear on 25 June 2013.
00:06 denis_boyun left 00:10 ajr_ left 00:15 ggoebel6 left, benabik joined, ggoebel6 joined 00:20 hlin_laptop joined
[Coke] stares blankly at the remaining opcodes. 00:24
diakopter what are they 00:25
[Coke] gist.github.com/coke/6799633 00:26
00:26 FROGGS joined
diakopter I don't understand these 00:27
[Coke] the opcodes, or the way in which they are presented?
diakopter the last section means there are ops that are documented that don't exist in moar?
[Coke] diakopter: yes
or my code isn't finding where they are defined.
and I've only documented ops that exist somewhere. (usually nqp-jvm) 00:28
diakopter jnthn: so all I need to do is implement the bottom section of [Coke]'s gist? :)
00:29 ivanshmakov left
[Coke] no. that's doesn't count things that aren't documented. :) 00:29
they'll probably help, though, sure.
jnthn diakopter: we'll need msot of those, yeah 00:30
00:32 benabik left, FROGGS left
[Coke] if someone groks the lex opcodes, this would be a great list to write up some ops for. If someone gets me the basics, I can format it all properly. 00:33
gist.github.com/coke/6944137
jnthn what, you mean the names aren't, like, totally logical? :D 00:38
00:38 ivanshmakov joined
diakopter they're completely descriptive. 00:38
jnthn [get|bind]lex[_\w]? are all just normal lexical lookup
bindlexdyn/getlexdyn just walk the caller chain
getlexcaller walks the caller chain but at each step walks the lexicals from that frame's point of view 00:39
getlexouter is normal lexical lookup but skips the current frame
getlexrel* do lexical/dynamic/caller lookup relative to the specified ctx
lexotic is something else altogether 00:40
It's used to implement return
[Coke] danke 00:48
I'll see if I can convert that into docs.
jnthn pzh 00:49
'night o/ 00:52
00:57 grondilu left 00:58 benabik joined 01:03 btyler left 01:08 ivanshmakov left 01:09 ivanshmakov joined
dalek ecs: fd70666 | grondilu++ | S99-glossary.pod:
MRO in S99

adding "Method Resolution Order" for MRO
01:15
01:23 ivanshmakov left 01:24 ivanshmakov joined 01:56 ivanshmakov left 01:58 ivanshmakov joined 02:10 ammoun joined
ammoun $row =~ s/$clitic//g; I want to substitute anything in $clitic with nothing, means just remove it 02:11
$clitic = "(’|:|-|’S|’D|’M|’LL|’RE|’VE|N’T|’s|’d|’m|’ll|’re|’ve|n’t)"; 02:12
Do you guys see any reason why this doesn't work?
diakopter ammoun: this channel is for Perl 6 questions, but I think you need //eg instead of //g 02:14
geekosaur I don't think /e helps there at all 02:16
ammoun e stands for expression right?
diakopter <- kinda dull; ignore me
geekosaur it stands for eval
you should ask in #perl. expect them to want to see more than 2 lines of your script (please use the paste site from the /topic) 02:17
ammoun Thank you guys
diakopter tries to understand why someone would want to remove contraction suffixes 02:18
hm, I guess for dictionary generation
er, wordlist
geekosaur I thought aspell had a dictionary generation mode, seems like a good idea to farm that out to something with a couple years' worth of intelligence in it 02:20
02:22 cognominal left 02:37 fgomez joined, fgomez left 02:39 jnap left 03:23 Dorado_Kirscht joined 03:36 Mouq left 03:38 araujo left 03:42 jnap joined 03:43 jnap left 03:50 preflex left, preflex joined, ChanServ sets mode: +v preflex 04:27 BenGoldberg left 04:34 kaare_ joined 04:45 Dorado_Kirscht left 05:02 dayangkun joined 05:32 dayangkun left 05:40 dayangkun joined 05:42 SamuraiJack__ joined 05:47 SamuraiJack joined 05:50 SamuraiJack__ left 06:09 dansamo joined 06:20 ammoun left
moritz \o 06:25
06:54 darutoko joined 07:02 FROGGS[mobile] joined
FROGGS[mobile] o/ 07:07
07:08 dayangkun left 07:24 kurahaupo joined 07:30 xenoterracide__ left 07:49 kaare_ left 07:54 kurahaupo left 07:55 dansamo left 08:05 dansamo joined 08:07 itz_ left 08:11 xenoterracide__ joined, Rotwang joined 08:21 berekuk joined 08:39 araujo joined 08:47 kurahaupo joined 08:51 kurahaupo left 08:53 darutoko left 08:58 xenoterracide__ left 09:02 darutoko joined 09:03 FROGGS[mobile] left 09:08 odoacre left 09:15 fhelmberger joined 09:16 kurahaupo joined 09:21 kurahaupo left 09:23 Rotwang left 09:24 rindolf joined, rindolf left 09:29 hlin_laptop left 09:33 FROGGS[mobile] joined 09:45 grondilu joined 09:46 SamuraiJack left 09:54 kurahaupo joined 09:55 fhelmberger left 10:05 woosley left 10:18 odoacre joined 10:20 rindolf joined 10:31 labster left, denis_boyun joined 10:45 cognominal joined 10:51 darutoko left, daniel-s_ joined 10:58 darutoko joined 10:59 daniel-s__ joined 11:01 daniel-s_ left 11:09 labster joined 11:22 pernatiy joined 11:27 denis_boyun left 11:33 darutoko left 11:43 rindolf left 11:44 daniel-s__ left, darutoko joined 11:50 zakharyas joined
masak good day, #perl6 12:08
quiet today, huh_
?
jnthn shhh!
nwc10 good day to write a FAQbot? :-) 12:09
12:10 Bzek left
masak heh. 12:16
no, I'm still on t4 reviews, thank you very much. 12:17
colomon \o 12:24
masak jnthn++ # 6guts.wordpress.com/2013/10/11/nqp-...ther-news/ 12:25
jnthn: oh, and in case you were wondering "Allocating Less Cursors" -- that would be a good time to use "fewer", since "cursor" is not used as a mass noun there. 12:26
colomon masak++
jnthn masak: That distinction is probably gone by the next generation, much like whom is mostly gone to mine. :) 12:29
masak please, no karma. I feel bad enough already being a grammar nazi... :)
jnthn So by getting it "wrong" I'm simply ahead of my time :P 12:30
masak jnthn: I'm a firm believer of language being a flexible, evolving thing. but that argument is silly, and I hope you realize that.
jnthn masak: I wasn't being entirely serious about the second bit, but there's at least a grain of truth in the first. 12:31
12:32 spider-mario joined, SamuraiJack joined
masak there are certainly "less/fewer" fights not worth having. such as ss/6 items <(or less)>/or fewer/ -- see itre.cis.upenn.edu/~myl/languagelog...03775.html 12:32
but in the case of "less cursors", that just sounds wrong to me because the distinction makes sense to uphold. 12:33
JimmyZ are you talking about designing english? 12:34
nwc10 masak: I also thought this
encrypted.google.com/search?q=10+i...CAcQ_AUoAQ 12:35
masak JimmyZ: no, just following it in its current implementation.
jnthn JimmyZ: No, just about a non-native speaker being upset about a construction a native speaker knows is technically wrong, but would never actually react to or intuitively feel is wrong.
masak :P
jnthn: oh, get off your low horse.
jnthn When I saw the "5 items or less" article I was like..."so what's wrong with it" :) 12:36
masak me too.
that, to me, is over-reacting.
hm, interesting. I would want it as "5 or fewer items", not "5 or less items". but "5 items or less" is fine to me. 12:37
which means that the "or less" probably refers to a *hidden mass noun* of some sort. not the items, but (maybe) the amount representing the items. 12:38
12:44 Rotwang joined
JimmyZ to avoids this, you could 'use chinese;' 12:46
12:48 SamuraiJack left
dalek kudo/moar-support: e1056ec | jnthn++ | src/vm/moar/ModuleLoaderVMConfig.nqp:
First pass at a ModuleLoaderVMConfig for Moar.
12:49
kudo/moar-support: af53dfe | jnthn++ | src/Perl6/ModuleLoader.nqp:
Make ModuleLoader.nqp Moar friendly.
kudo/moar-support: aeb4e8b | jnthn++ | / (2 files):
Stub configure/makefile for MoarVM building.

Doesn't even get passed the first file yet, and will want various things fixing along the way, but enough to start the porting work.
jnthn uh, get *past*
masak JimmyZ: :P
JimmyZ: what, because it's impossible to make grammatical errors in Chinese? um, yeah, that totally aligns with my studies of the language. :) 12:50
jnthn masak: Maybe there's fewer opportunities for them :P 12:51
masak .oO( it's not funny when you use a grammatical construction *correctly*... ) :P
jnthn: s/NFa/NFA/ at some point in the post. 12:52
jnthn I'm quite sure if it wasn't for the preceding discussion I'd have written "less" there :P
Fixed the NFa thing, thanks 12:53
masak "less opportunity" is fajn, fwiw. 12:54
tadzik fajn, haha
JimmyZ masak: it's possible. but not 'less or fewer' part
masak oh, granted. 12:56
JimmyZ masak: if you grammatical errors in Chinese, you could { use english; .... }
masak tadzik: pre-emptive en.wikipedia.org/wiki/Muphry%27s_law :P
JimmyZ: the important thing is keeping track of which language you're in. :P
jnthn "if you grammatical errors" :D
.oO( autopun? )
12:57
JimmyZ by use '{ }'
masak jnthn: I'm not sure it's an autopun if it's completely unintenional... 13:00
13:02 LylePerl joined
masak that is, the utterer doesn't have to be aware, but at least the messenger has to be. 13:03
LylePerl! \o/
LylePerl hi
Just a quick visit unfortunately
www.linkedin.com/groupItem?view=&am...VClIbbmBY1
sorry for the long url
there is a Perl 6 thread on linked in, which could probably do with a knowledgeable response 13:04
I thought we had a bot that automatically shortened urls?
masak long ago, I think.
LylePerl how are you masak? Enjoying married life? 13:08
masak :)
clkao 11^W 13:09
dalek p: 60a7992 | jnthn++ | src/vm/moar/QAST/QASTOperationsMAST.nqp:
Map nqp::bindhllsym on MoarVM.
13:10
13:11 spider-mario left
Ulti am I the only person who gets angry when seeing the phrase "in production" in relation to Perl5 or Perl6? no one asks how many people use BASH in production or sed, awk or any other language like tool that's just useful regardless... a better question is anyone using Perl6 for something useful thats hard to accomplish in Perl5, if that question isnt yes who cares if its used in production if there isn't a benefit to doing so! 13:13
masak Ulti: many people are not early adopters. 13:14
Ulti: they basically want an answer to the question "did someone else do something serious with it yet?"
I can see where they're coming from.
Ulti plenty of things are really specific too, lisp in the form of CLIPS is used "in production" for hardware control systems specific to space flight... is that interesting?
masak: sure but the problem is everyone else isn't vocal in the perl world 13:15
masak what's that got to do with anything?
Ulti I've taught about ten people Perl in just the last few years, none of them use the "community" or would think to talk about the tools they use
its not something they care about
they just assume software tools are like hammers
they will always be around in one form or another
perl6 could have a thousand early adopters and you wouldnt know unless they were financially invested in everyone else using perl6 13:16
masak to some people, Perl 5 is exactly that: a hammer. they don't go to hammer conferences or workshops -- the entire notion of doing that is meaningless to them.
Ulti but thats my point perl5 is never going to die, its just become a hammer instead of some luxury gadget worth chatting about 13:17
13:18 spider-mario joined
dalek p: 60e2222 | jnthn++ | src/vm/moar/ModuleLoader.nqp:
Bring MoarVM module loader in line with others.
13:20
Ulti if Perl6 is the next version of Perl I dont get why anyone would expect it to suddenly generate startups wanting flashy hip devs
masak Ulti: hm, interesting point.
Ulti: I think we're far from that time yet, though. of Perl 5 becoming just a hammer.
Ulti I'd say Perl is essentially expected to exist on any UNIX system like awk and sed 13:21
Ulti is trying to remember if his raspberry pi had Perl... my NAS did
13:22 spider-mario left, spider-mario joined
Ulti the tiny embedded swarm robot I used for my research had perl which I used with swig to write my controllers :S 13:23
and if in production means Perl6 takes the place of Perl5 to that level its a long long way away just because of lag in all the different people who will have to package it 13:24
if anyone has the latest Rakudo* installed, could you try doing 'panda install LWP::Simple' and see if you get: No such method 'postcircumfix:<( )>' for invocant of type 'Signature' 13:43
LylePerl checkout my new hammer 13:46
Ulti if the swiss army made hammers it would be perl6 13:49
13:54 SamuraiJack joined
diakopter Ulti: japhb__ mentioned here the other day he used rakudo-jvm to parallelize something nifty at his Googler job 13:55
sounded like a personal tool (not critical path of delivery of service to customer) for syseng stuff 13:56
Ulti yeah I've played at doing the same though I haven't used all the nice regexes Util made for me yet 13:57
13:58 dmol joined
Ulti my use case for perl day to day usually involves using GNU parallel alongside, it would be nice to have simple built in syntax for doing a lot of that 13:58
13:59 SamuraiJack_ joined 14:01 SamuraiJack left 14:13 spider-mario left 14:14 PZt left 14:18 spider-mario joined, PZt joined 14:30 SamuraiJack_ left 14:37 fgomez joined, fgomez left 14:43 xinming left 14:44 xinming joined
timotimo r: grammar Nazi { } 14:45
camelia ( no output )
tadzik :D 14:48
14:50 ssutch left
tadzik oh wow. I found a website that's giving hugs: thenicestplaceontheinter.net/ 14:51
14:53 pmurias joined
pmurias [Coke]++ # awesome op docs 14:53
timotimo those docs are op! 14:55
jnthn No longer must we put op with undocumented ops! 14:56
[Coke] there are about 500 failing op tests still!
jnthn aww 14:57
:)
Still a big improvement ;)
pmurias how do you generate the list of ops?
moritz jnthn++ # blog post 14:58
[Coke] pmurias: t/docs/opscodes.t - I dig through the source of each vm and look for where ops are added. 14:59
15:01 denis_boyun joined
[Coke] bindlex takes a long, a callframe, and an int - lexes are stored by position, not name/ 15:02
?
(what's the int index into the callframe? level up?)
jnthn [Coke]: nqp::bindlex takes a name and an obj 15:03
[Coke]: But there are overloads used by QAST::Var compilation
[Coke]: Those ones calculate indexes and number of frames out
Grrrr jnthn++ for submitting an NPW2013 talk 15:06
15:06 zakharyas left
jnthn Grrrr: It's pretty convenient for me to attend this year :) 15:07
Grrrr jnthn: well, somewhat less so than 2 years ago, I am sure ;)
jnthn Grrrr: True, though I was doing a little more than attending 2 years ago ;) 15:08
15:13 mtk left 15:16 denis_boyun left
[Coke] ... ok. documenting these not as easy as it seems, then. :) 15:18
15:18 kaare_ joined
dalek Heuristic branch merge: pushed 220 commits to nqp/nativecall-jvm by jnthn 15:25
jnthn arnsholt: I merged latest master into nativecall-jvm 15:26
15:27 kurahaupo left 15:28 kurahaupo joined 15:29 kurahaupo left, kurahaupo joined 15:31 kurahaupo left, kurahaupo_ joined 15:33 kurahaupo_ left, kurahaupo joined 15:34 kurahaupo left, kurahaupo joined
arnsholt jnthn: Excellent! 15:35
jnthn arnsholt: I had to change the r_atpos to atpos to get NativeCall to compiler on JVM; I think we can actually switch it to that without busting it on Parrot these days too 15:37
arnsholt: I can't seem to get t/nativecall/01-basic.t to work... 15:39
15:39 kurahaupo left, kurahaupo joined
arnsholt Hmm. That's not good... 15:40
What kind of error do you get?
jnthn java.lang.UnsatisfiedLinkError: Unable to load library '': Native library (win32-x86-64/.dll) not found in resource path ([file:/C:/consulting/rakudo/nqp/, file:/C:/Program%20Files%20(x86)/QuickTime/QTSystem/QTJava.zip]) 15:41
oh...
arnsholt Argh. That might be a bug in the test script 15:42
jnthn nah, it's looking for printf
And I guess trying to use the "current process" trick..
arnsholt Yeah, that's the idea 15:43
Writing portable tests for NativeCall is really tricky
15:44 PZt left
jnthn Yeah, the current process thing doesn't work on Windows... 15:47
Did the tests actually pass for you?
On non-Widnows? 15:48
masak those links, so unsatisfied...
arnsholt When I last worked on this it worked on Linux, yeah
15:48 Rotwang left
arnsholt And they still do on Linux 15:49
jnthn Looks like JNA may give us a portable way...trying it.
load me...and then just link me...till I can get my satisfaction... 15:50
diakopter m
jnthn arnsholt: ooh, I managed to get it to do ok - printf :) 15:51
strdup ain't foudn though...grr
arnsholt Anyways, my plan forward was something like this: 15:52
jnthn oh...but if I make it _strdup it works :) 15:53
arnsholt 1) Get get_boxed_ref working, so that we can run the NativeCall test suite proper and see what happens
2) Fix the bugs that shakes out in the stuff that's already supposed to be working
3) CArray 15:54
4) CStruct
5) Callbacks
moritz 6) everything else :-) 15:56
arnsholt Yeah =)
Perhaps an even more fruitful approach would be making 3) Whatever's needed to get DBIish working
moritz fwiw I'm not sure if DBIish on the JVM should also use NativeCall 15:58
or rather the JDBC drivers
arnsholt That's a very good point 16:00
jnthn We'll still need NativeCall either way, though. 16:01
arnsholt Oh, definitely! 16:02
16:03 Rotwang joined
dalek p/nativecall-jvm: ddb9cd4 | jnthn++ | src/vm/jvm/runtime/org/perl6/nqp/runtime/NativeCallOps.java:
Some NativeCall error handling tweaks.
16:03
p/nativecall-jvm: a37b601 | jnthn++ | src/vm/jvm/runtime/org/perl6/nqp/runtime/NativeCallOps.java:
Do "current process" loading in a portable way.
p/nativecall-jvm: 20c10a9 | jnthn++ | t/nativecall/01-basic.t:
NativeCall test tweak; now passes on Windows.
nwc10 there's a DBD::JDBC and a DBD::ODBC for Perl 5, so it would make sense to have JDBC style drivers also available for Rakudo on the JVM, even if the "usual" is "direct"
pmurias arnsholt: nqpy has been suspended?
jnthn arnsholt: Well, now I have those passing on Windows I can no I didn't bust it too badly :) 16:04
*know 16:05
dalek kudo-js: f345298 | (Pawel Murias)++ | / (2 files):
Update the nqpjs to the new code organisation.
kudo-js: 6d08ea0 | (Pawel Murias)++ | bin/nqp-js.nqp:
Use HLLBackend::JavaScript from nqp-js.nqp.
kudo-js: fff8aba | (Pawel Murias)++ | / (5 files):
Emit modules instead of .js files so that ender doesn't cause problems.
kudo-js: e088488 | (Pawel Murias)++ | / (13 files):
fixjsstyle --strict --nojsdoc nqp-runtime-core/bootstrap.js nqp-runtime-core/code_ref.js nqp-runtime-core/ctx.js nqp-runtime-core/hash.js nqp-runtime-core/helpers.js nqp-runtime-core/null.js nqp-runtime-core/reprs.js nqp-runtime-core/runtime.js nqp-runtime-core/serialization_context.js nqp-runtime-core/serialization.js nqp-runtime-core/serialization_writer.js nqp-runtime-core/sixmodel.js
rakudo-js: f33afda | (Pawel Murias)++ | nqp-runtime-core/ (6 files):
rakudo-js: Fix some errors reported by make lint (gjslint).
arnsholt pmurias: Not really, but I've been busy with article writing and such at work lately 16:08
jnthn: Still work on Lunix. Hooray! \o/ 16:09
jnthn \o/
arnsholt: OK, I'll have a crack at (1) 16:10
16:12 kurahaupo left
arnsholt Cheers. That's the hardest one from my perspective 16:12
16:12 kurahaupo joined, [Sno] left 16:13 pecastro_ left
arnsholt pmurias: So if you want to hack on it, feel free. There are some bugs in the grammar that need sorting, but I mostly know how to if you're interested in hacking on it 16:18
timotimo repast.sourceforge.net/repast_3/rep.../nqpy.html - this nqpy? 16:19
:P
16:22 denis_boyun joined
arnsholt Heh, name's already taken ^_^ 16:23
Should've seen that one coming =)
16:30 rindolf joined 16:31 xenoterracide__ joined
dalek p/nativecall-jvm: ca138b9 | jnthn++ | src/vm/jvm/runtime/org/perl6/nqp/ (4 files):
Hold native callsite info at an extra indirection.

This will help support inlining it into a P6opaque.
16:35
p/nativecall-jvm: 43567c3 | jnthn++ | src/vm/jvm/runtime/org/perl6/nqp/sixmodel/reprs/NativeCall.java:
Teach NativeCall REPR how to inline itself.

Note that there's not currently a way to get hold of the inlined thing so NativeCall can use it; that will come next.
16:36 kurahaupo_ joined 16:37 kurahaupo left 16:39 hummeleB1 joined, ajr joined, ajr is now known as Guest55207
jnthn arnsholt: Would Rakudo need another thing in the classpath for JNA to work? 16:40
Seeing java.lang.NoClassDefFoundError: com/sun/jna/NativeLibrary
ah, jna.jar...
16:41 mtk joined
arnsholt Yeah, unless the nqp JAR includes the third part stuff inside it'll need that, yeah 16:42
16:42 REPLeffect left, kurahaupo_ left, kurahaupo joined
jnthn k 16:42
16:43 hummeleB1 left
arnsholt We should probably set up nqp.jar like that at some point, actually 16:43
16:44 Guest55207 is now known as ajr_, hummeleB1 joined
dalek volaj: b946f3f | jnthn++ | lib/NativeCall.pm6:
.exists => :exists
16:44
p/nativecall-jvm: 0040bf6 | jnthn++ | tools/build/Makefile-JVM.in:
Install jna JAR.
16:47
16:59 [Sno] joined
jnthn Time to sort out some dinner. bbl :) 17:00
timotimo "NQP is Copyright (C) 2013 by Paweł Murias." wat :) 17:02
moritz it's not? :-)
17:12 kurahaupo left, kurahaupo_ joined 17:23 denis_boyun left 17:25 spider-mario left 17:26 kurahaupo_ left, kurahaupo joined 17:29 kurahaupo_ joined, kurahaupo left 17:31 kurahaupo joined, kurahaupo_ left 17:33 kurahaupo_ joined, kurahaupo left 17:41 ssutch joined 17:43 kurahaupo_ left, kurahaupo joined 17:45 dmol left 17:48 kurahaupo_ joined, kurahaupo left, pecastro joined, kurahaupo joined, kurahaupo_ left 17:53 FROGGS joined 17:55 kurahaupo_ joined, kurahaupo left, kurahaupo_ left, kurahaupo joined 17:56 REPLeffect joined 17:59 kurahaupo_ joined, kurahaupo left
pmurias timotimo: ? 18:02
timotimo says so on the bottom of your readme; i'm confused about that 18:03
your readme == the one in rakudo-js 18:04
dalek kudo-js: e0ad3d3 | (Pawel Murias)++ | README.markdown:
Fix README.
18:06
18:07 pecastro left 18:08 kurahaupo joined, kurahaupo_ left 18:19 kurahaupo_ joined, kurahaupo left 18:26 kurahaupo joined, kurahaupo_ left
masak whoa, t4 is so deep-and-subtle. 18:26
moritz that's why I didn't find a good solution to it :-) 18:30
18:31 kurahaupo left, kurahaupo_ joined
dalek rlito: a3e2348 | (Flavio S. Glock)++ | / (2 files):
Perlito5 - perl6 - special var
18:35
18:36 darutoko left 18:42 kurahaupo_ left, kurahaupo joined
pmurias arnsholt: do we rename nqpy? 18:47
arnsholt I have no objections to a rename 18:49
Open to suggestions for a better name =)
lue
.oO(It's too bad Blender doesn't allow scripting for Perl 6. t4 would've been obvious then.)
pmurias arnsholt: inventing a new name is the hard part ;) 18:50
arnsholt Yup
Which is why I went with NQPy in the first place. Very little invetiveness required =) 18:51
lue PythoNQPythoNQP... ? 18:52
pmurias arnsholt: snake?
arnsholt Ooh, not bad 18:53
Snake in turn tempts me to suggest something like badger or mushroom, but snake is probably better =) 18:56
18:57 rindolf left
pmurias badger and mushroom will be the qast ruby and php implementations ;) 18:59
19:00 mtk left
arnsholt =D 19:00
19:01 cognominal left
rurban who is working on the libuv cygwin issues? My branch is at github.com/rurban/libuv/tree/cygwin but __io_poll is missing 19:03
19:03 mtk joined 19:06 cognominal joined
jnthn rurban: Apparently, you're working on it ;) 19:10
19:13 sivoais left
pmurias arnsholt: do we have dalek hooks for nqpy/snake? 19:15
arnsholt Not yet 19:16
What's needed to set that up? 19:17
pmurias github.com/perl6/mu/blob/master/mi...k-push.txt
(I can't do that as I have no access to the Admin panel of nqpy repo) 19:18
arnsholt: it's mostly clicking around on github and copy and pasting a url 19:22
arnsholt Indeed it is 19:25
arnsholt goes to rename
pmurias arnsholt: do we implement python 3 or python 2? 19:26
arnsholt 3 19:28
19:29 xinming left, xinming joined
arnsholt The section numbering in Grammar.nqp follows the numbering in docs.python.org/3/reference/ 19:30
19:31 kurahaupo left 19:32 kurahaupo joined, hummeleB1 left
arnsholt pmurias: While we're on the topic, do you know if there's something like a spectest suite for Python? 19:32
I've looked around a bit but haven't found anything yet 19:33
grondilu we can ask #python
pmurias arnsholt: found it
Lib/test in the python repo 19:34
19:35 kurahaupo left 19:36 kurahaupo_ joined
arnsholt Ah, brilliant! 19:37
19:37 kurahaupo_ left 19:38 kurahaupo joined 19:40 PZt joined
pmurias arnsholt: one problem with that is that those tests seem rather complex (we would have to implement classes and a lot of other stuff to get any of them to pass/run) 19:40
arnsholt: so we will have to write/find a set of "sanity" tests 19:42
arnsholt Yeah, can't be avoided I guess 19:45
grondilu what's t4?
felher grondilu: github.com/masak/p6cc2012 -- part 4 of the last p6cc 19:47
grondilu oh, ok 19:48
pmurias arnsholt: you changed the repo name? 19:52
found it 19:53
arnsholt Yeah, sorry I forgot to tell you
moritz if you want to have some fun, we can move it into the perl6 organization :-) 19:55
19:55 kurahaupo left, kurahaupo joined
pmurias arnsholt: did you add the dalek hooks? 19:58
arnsholt 'sec. Got sidetracked
pmurias np 19:59
dalek ake: 0f9e071 | (Arne Skjærholt)++ | src/NQPy/Grammar.nqp:
Handle comments properly in ws.
ake: 58db12f | (Pawel Murias)++ | / (12 files):
Rename nqpy to snake.
ake: 5e0b0a8 | (Pawel Murias)++ | / (2 files):
Add simplest test possible, add test target target to Makfile.
arnsholt It's a snake! 20:00
20:03 dwarring joined, cognominal left
jnthn won't badger anyone about Python commit messages on #perl6, 'cus they won't take up mush room... 20:05
arnsholt =D 20:06
20:07 kurahaupo left, kurahaupo_ joined
rurban jnthn: yes, but my io_poll idea looks awful. just duplicate the windows io_poll. maybe not_gerd or jimmyz got better ideas 20:10
20:14 shtt465 joined
felher looks at a review of t1: "Not sure type1, type2, type3, and type4 conveys the maximal amount of information to the interested reader."... hihi, priceless 20:17
dalek ake: 85167cd | (Arne Skjærholt)++ | src/Snake/ (2 files):
Make indentation handling less broken.

We're probably not all the way there just yet, but at least it's less hilariously broken this way than it was.
20:18
20:19 dmol joined
colomon snake? 20:19
timotimo ah, snake is not "translate nqp to python", it's a python implementation in nqp 20:20
colomon arnsholt++
arnsholt Yeah. Fallout from me reading the RaNIW notes =) 20:21
lue arnsholt: is that what I keep referring to as "the slides" ?
20:21 woolfy left
arnsholt I'd assume so, yeah 20:22
lue (specifically the "learn how to NQP" slides)
arnsholt Yup. That's the stuff
colomon will be very happy when he can find the time to read / play with those slides 20:23
dwarring I've been been extending the rubyish example from RaNIW
been a lot of fun
README summarises progress github.com/dwarring/nqp-rubyish 20:24
diakopter dwarring: feel free to submit a pull request; I'm sure someone will want it somewhere 20:25
moritz and here I was hoping that the RaNIW attendees actually hack on nqp and parrot
dwarring it's still all just in a single source file
moritz :-)
diakopter parrot?
dwarring nqp 20:26
dalek ecs: 919626d | Mouq++ | S99-glossary.pod:
Add LTM and NFG to S99
diakopter I thought LTM was already on there
arnsholt moritz: Well, I didn't actually attend O=) 20:28
20:29 gshank joined
arnsholt But I'm mulling over some NQP work as well, partially as fallout from the Python stuff 20:29
grondilu reads S99 and wonder if EPIC FAIL needs to be there. 20:30
20:32 lichtkind joined
dwarring not sure where to put rubyish - do we need a perl6/nqp_examples? 20:33
or a perl6/perl6_examples/nqp? 20:34
20:34 kaare_ left
moritz dwarring: or you can put it into perl6/perl6-examples in an .nqp file 20:34
dwarring: and maybe add a comment on top that it's actually nqp, not Perl 6
dwarring: no need to create a separate repo for a single file unless that's what you want 20:35
20:35 shtt465 left
dwarring moritz: makes sense 20:36
i'll add it there 20:37
20:37 sivoais joined
moritz dwarring: do you a commit bit already? 20:37
dalek ecs: 465527e | grondilu++ | S99-glossary.pod:
defining sigil
20:38
dwarring moritz: should have 20:39
jnthn grrr...I almost forgotted JVM bytecode 20:40
dwarring i've committed there before
arnsholt jnthn: Win32-API on JVM sounds awesome! \o/ 20:44
dwarring grondilu: I'll add rubyish.nqp to perl6_examplkes/interpreters
arnsholt jnthn++
dwarring *examples
dalek p/nativecall-jvm: 764b127 | jnthn++ | src/vm/jvm/runtime/org/perl6/nqp/ (3 files):
Make native call sites nested in P6opaque work.

With this, we can have NativeCall basically working on JVM (will need work on CArray and CStruct before we have full support, though).
jnthn arnsholt: OK, that's (1) done and since the win32 example worked right off, probably (2) also :) 20:45
dwarring: Maybe it belongs as an NQP example in the NQP examples/ dir
arnsholt: I'm just gonna check that my patches to NativeCall.pm6 doesn't bust it on Parrot
20:46 _espo joined
dwarring ok, i'll maybe add add to nqp/examples 20:49
arnsholt Ooh, that sounds very promising! 20:50
I'll try to get some stuff done on it in the morrow
jnthn arnsholt: Great!
20:51 aindilis joined
grondilu dwarring: maybe it's best if you put it in perl6-examples/parsers 20:54
20:55 ajr_ left
timotimo i misread "morrow" as "moarvm" and was like "wait, when did rakudo get onto moarvm already" 20:55
20:56 _espo left
jnthn arnsholt: grrr...nope, I busted arrays 20:59
grondilu dwarring: oops, haven't read enough. nqp/examples it is, then. 21:00
21:03 dmol1 joined, dmol left 21:04 fgomez joined, dmol1 left, dmol joined 21:05 fgomez left 21:10 dmol left 21:11 dmol joined 21:14 dmol left 21:16 dmol joined
dalek p/nativecall-jvm: 10abc40 | jnthn++ | src/vm/parrot/pmc/sixmodelobject.pmc:
Add missing 6model v-table forwarders.

With this we can use the normal nqp::atpos_* with CArray. This means NativeCall can be tweaked to build on JVM also without breaking it on Parrot.
21:20
jnthn arnsholt: Now that things work, is there any reason to keep this work in a branch?
arnsholt: If I merge it now I can just bump NQP_REVISION and update NativeCall, rather than have a branch there too...
21:22 dmol1 joined 21:23 dmol left 21:24 dansamo left 21:27 kurahaupo_ is now known as kurahaupo 21:30 dmol joined 21:31 dmol1 left 21:39 btyler joined, cognominal joined 21:43 dmol left, dmol1 joined 21:45 benabik left 21:46 benabik joined 21:55 colomon left 21:56 colomon joined 22:01 kurahaupo left 22:06 colomon left 22:12 kurahaupo joined 22:21 colomon joined
dalek p: 825038b | jnthn++ | src/vm/moar/QAST/QASTCompilerMAST.nqp:
Remove leftover code from the Moar nqp-cc.

This avoids loads of calls to nqp::getenvhash(...) in the eval-heavy t/qregex.
22:22
22:26 lichtkind left
lue Does anyone here with access to feather know why the percabal.org/syn and feather.perl6.nl/syn pages don't list the S99 entry? It bugs me that it's been missing for a while now. 22:26
22:29 kurahaupo left
grondilu Off Topic: guys, have you watched the latest spacex grasshopper test? www.youtube.com/watch?v=9ZDkItO-0a4 22:35
grondilu was skeptical about this thing but he's beginning to be impressed. 22:36
22:37 kurahaupo joined
timotimo look at that plume(age) 22:37
that thing is crazy accurate 22:38
22:39 xinming left
diakopter grondilu: that's truly amazing 22:39
22:40 dwarring left, xinming joined
benabik Shame the retrieval attempt on the v1.1 launch failed. 22:41
22:46 kurahaupo left
jnthn grondilu: wow! 22:47
Science/engineering win.
grondilu what are you guys waiting to build a rocket with a Perl6 control system? 22:48
:)
I'm kidding of course, what you guys do is almost as awesome as what SpaceX does :) 22:49
22:50 btyler left 22:56 kurahaupo joined 23:02 pmurias left, woolfy joined 23:05 kurahaupo left 23:10 Rotwang left 23:21 xinming left 23:23 xinming joined 23:30 grondilu left 23:34 fuad joined 23:48 kurahaupo joined