Parrot 3.5.0 "Menelaus" released | parrot.org | Log: irclog.perlgeek.de/parrot/restoday
Set by moderator on 25 June 2011.
00:03 whiteknight joined
cotto mikehh, one guy who hacks on space telescopes for $dayjob 00:04
also bumped into Util
whiteknight, ohai. im in ur timezone
whiteknight isn't it so much better than the other timezones? 00:05
pmichaud waves to all of the yapc::na attendees
cotto not the first day
dalek kudo/nom: 3c22598 | pmichaud++ | src/core/ (2 files):
Update Array.new and List.new to be more robust.
 00:07
00:26 benabik joined
benabik 'ello, #parrot! 00:29
whiteknight hello benabik. How's the work going? 00:30
00:32 daniel-s left
benabik whiteknight: Not very quickly. Been travelling. :-D 00:32
whiteknight traveling schmaveling. You can type while you drive
benabik Uhm, no?
Okay, just wanted to poke my head in here to mention I exist again. Heading off to scout for food. 00:36
cotto benabik, you at yapc? 00:38
benabik cotto: Yes
cotto aloha, going to yapc::na?
aloha cotto: going to yapc::na is cotto or dukeleto or atrodo or Util or kid51 or benabik or soh_cah_toa
cotto sweet
benabik I have no idea what I'm doing, but I'm here. :-D
dalek rrot/whiteknight/packfilewrapper: 2fbd0d2 | Whiteknight++ | / (6 files):
Rename functions to Parrot_pf_serialize and Parrot_pf_deserialize. cotto++
cotto if you're on foot, there's a mexican place 10-15 minutes away that's not bad
benabik I drove here, so am highly moble. 00:39
cotto I flew here, so I'm not. ;{
benabik Driving was cheaper than flying two people. And let me stop at my folk's on the way. 00:40
If people get tired of hotel-centric stuff, I'm willing to do some ferrying. :-D
cotto benabik, I'm sure we'll take you up on that. 00:41
00:41 contingencyplan left
benabik is off to investigate "The Thirsty Monk". Pub grub! 00:43
whiteknight is getting thirsty just reading this 00:44
luckily, I have a fridge stocked with all the fixins for making mohitos 00:45
dalek rrot/whiteknight/packfilewrapper: e10d9e0 | Whiteknight++ | src/pmc/packfileview.pmc:
Combine trigger_init and trigger_load methods into a single trigger(STRING *flag) method. This change looks forward to a time when packfiles can flag Subs more flexibly than we do now. Right now the only accepted values are 'load' and 'init'
 00:47
whiteknight cotto++ for that suggestion too 00:48
cotto whiteknight++ for doing things 00:50
whiteknight the things are easy. The good ideas are hard
cotto whiteknight, does that branch build for you? 00:53
whiteknight yes
you need to reconfig to make sure the new pmcs get included in the makefile 00:54
cotto My compiler complains about some c89 violations
whiteknight oh, that's very plausible
cotto or did. let me see if they're still ther
e
they are
do you run Configure.pl with --no-stupid-c89-limits ? 00:55
whiteknight my adherance to C89 is pathetic at best
I compile with clang
let me try again with g++
the crotchety, whiney, complaining senior citizen of the compiler world
dukeleto lulz 00:56
whiteknight: does that make gcc the senior citizen that just doesn't give a shit? 00:57
whiteknight clang is the hip young ne'er do well with his rock music 00:58
cotto whiteknight, what about some tests? 00:59
whiteknight I'm getting there 01:00
cotto ok 01:01
whiteknight I just compiled with g++ no problems 01:12
...and passed coretest 01:16
cotto: I need to provide information in PackfileView to either get the total number of constants of each type, or return an array of all constants of each type. Any particular way you would like to see that done? 01:21
right now I provide get_*_keyed_int vtables, but without bounds information or iterable interfaces that's useless
cotto whiteknight, why not make it iterable? 01:23
whiteknight There are three groups which we would need to iterate over: PMC, STRING, and FLOATVAL 01:24
01:24 woosley joined
cotto I see. 01:24
whiteknight I feel like I would prefer to add a method to either return counts, or return arrays of constants
but adding three methods for the purpose seems clunky 01:25
and adding one method that takes "STRING", "PMC", or "FLOATVAL" string keys seems stupid too
cotto yeah
and having an iterator that returns all three types isn't amazing either 01:26
whiteknight how would that even work, shift_* on the iterator? 01:27
cotto poorly
whiteknight and then a whole new iterator type just for this rare purpose?
cotto having three methods seems like the least bad approach atm 01:28
whiteknight howabout inspect_str("counts") returns a 3-element FIA with counts. I can wraie up some library code in Winxed or PIR to do fancy stuff with it 01:29
cotto what about INTVAL?
whiteknight INTVAL constants are stored in the bytecode stream directly, not stored in the constants table
cotto oh, right
how special
whiteknight magical and lovely
actually, it's good for performance so I won't complain 01:30
cotto yeah
I like that idea. I'd prefer a hash, but the concept is good.
whiteknight inspect_str seems like the best. The functionality is not commonly-needed, so I don't want to add methods for it
hash is fine too
cotto harder to mess up 01:31
whiteknight {"pmc" : x, "string" : y, "num" : z }
okay, I'll get started on that stuff tomorrow. For now, bed 01:32
cotto 'night
01:38 rohit_nsit08 left 01:43 alester joined 01:48 whiteknight left
dalek kudo/nom: 8812718 | pmichaud++ | / (11 files):
Create an opcode for building Parcels; the opcode also handles replacing
 01:52
01:55 lichtkind left 02:00 Kulag joined 02:02 Drossel left 02:06 bluescreen left 02:16 Coke left 02:17 Coke joined 02:36 soh_cah_toa joined 02:39 rohit_nsit08 joined
cotto NotFound, ping 02:39
02:40 rurban_ joined 02:43 rurban left, rurban_ is now known as rurban
cotto hi soh_cah_toa 02:43
where are you at?
benabik pub good, grub fail. 02:44
But there's an excellent cafe downtown that serves amazing southern-type food.
soh_cah_toa cotto: in my room
the wireless here is teh sux0rz 02:47
buncha perl geeks clogging it up
atrodo ya, i can imagine how slow internet's going to be this week 02:48
cotto I'm using dukeleto's tethered wireless. no need to clog up the hotel's
benabik soh_cah_toa: Yes, just noticing that.
atrodo I'm hoping my boss leaves his new 4g mifi on 02:49
soh_cah_toa msg kid51 i think i remember you saying that yapc attendees should message you when they arrive. well, i'm here. yay :) 02:50
aloha OK. I'll deliver the message.
soh_cah_toa 18 hours in car == not fun 02:52
atrodo Aye, no doubt. I only had 6 hours
cotto argh 02:53
that's about as much time as I spent on planes and in airports
02:54 bubaflub left
benabik Maybe I'll turn on my iPad's cell modem again. :-/ 02:54
cotto soh_cah_toa, where did you come from?
soh_cah_toa new jersey
dukeleto soh_cah_toa: you are hardcore for driving that long 02:55
atrodo thats' dedication soh_cah_toa++ 02:56
soh_cah_toa dukeleto: not me. i got free ride w/ parents :)
i'm a mooch, yes
atrodo soh_cah_toa-- i remove my karma then
cotto also an option
soh_cah_toa ha!
benabik 16 hours for me, but I split it in two.
moderator Parrot 3.5.0 "Menelaus" released | parrot.org | Log: irclog.perlgeek.de/parrot/today 02:58
atrodo cotto or dukeleto, can either of you add a post-hook on github? I'm going to rollout the new isparrotfastyet, consequences ignored 03:02
cotto one of us can
It's not me
atrodo I wasn't sure, so I'll wait for dukeleto 03:03
soh_cah_toa i'm looking at the site now. what's oofib? 03:04
atrodo one of the benchmarks included in parrot 03:05
cotto examples/benchmarks/oofib.pir
object-oriented fibonacci
atrodo hmmm, looks like i forgot to add the descriptions i "painstakingly" created
dukeleto atrodo: what do you need? 03:07
03:14 alester left
atrodo goodnight #parrot, see you tomorrow 03:14
soh_cah_toa see ya 03:15
cotto 'night 03:16
dalek kudo/nom: ab1dda5 | pmichaud++ | src/core/ (15 files):
Convert some more pir:: opcodes to nqp:: .
 03:18
soh_cah_toa cotto: are you gonna be at the breakfast? 03:32
cotto soh_cah_toa, I didn't realize there was one. 03:34
sure
soh_cah_toa yeah, at 9:00. it's on the schedule on the website
cotto 9 isn't too bad
soh_cah_toa i'll try and look for dukeleto since he's the only one i'd recognize 03:39
cotto soh_cah_toa, trac.parrot.org/parrot/wiki/Yapc10B...ouppicture 03:41
the badges will help 03:42
soh_cah_toa oh yay, i get a badge 03:43
cotto sleeps
soh_cah_toa me too 03:44
03:44 soh_cah_toa left 03:46 bubaflub joined
dukeleto bubaflub: werd 03:46
bubaflub hola dukeleto
dukeleto bubaflub: how goes? 03:47
bubaflub dukeleto: not bad, not bad. you with all the peeps at yapc?
dukeleto bubaflub: yeps. Busily writing my talk for tomorrow 03:48
bubaflub dukeleto: very nice. wish i could be there with y'all
dukeleto bubaflub: how is parrot-gmp ? 03:59
bubaflub dukeleto: not bad. had some progress since the recent blog - all the random number functions require a specific struct and initialization / seeding so i got that in place with tests. monday i should have all the integer functions totally covered. 04:02
dukeleto: after that it's hooking up distutils to do the heavy lifting, checking we've got the right version of parrot and GMP, and adding some more basic docs. 04:03
dukeleto bubaflub: awesome. You are making good progress 04:08
bubaflub dukeleto: woot. those are all the mid-term deliverables. i'll probably have some extra time so i'll try to get some NQP examples together. also, i've been talking with plobsing about some NCI stuff. 04:09
dukeleto bubaflub: what kind of NCI stuff? 04:11
bubaflub dukeleto: getting docs up to date first and foremost. with recent deprecations many parrot bundled examples broke, so i'm working slowly on a script and a patch for ncidef2pir.pl to get those working again. 04:12
dukeleto: mostly just pestering him with questions so i can extract some of that NCI knowledge. 04:13
04:16 Kulag left, Kulag joined
dukeleto bubaflub: are the examples seperate files? If so, we should have tests that make sure they work 04:21
bubaflub dukeleto: right now there is only one example - but how should i make sure they work? check the output?
dukeleto: er, i mean, the examples are full blown programs on their own. maybe a Test::More that shells out to run something?
04:24 Kulag left, Kulag joined
dukeleto bubaflub: sure. We need to make sure our examples continue to work, or they are pretty useless 04:25
bubaflub dukeleto: right right. sorry, for a moment i thought you were talking about Parrot-GMP examples of which there is only one but it works. some bundled parrot NCI examples do not work because of NCI deprecations. 04:27
dukeleto: those examples still compile from definition -> pir -> pbc fine, but through a runtime error about an invalid signature and what not. i can fix the examples that we have in the repository but ideally we still want tests to actually execute them to see if they still work. i'll keep thinking about how we can do that.
goodnight, #parrot. 04:32
04:35 Kulag left, Kulag joined 04:41 Kulag left, Kulag joined 04:45 bubaflub left, jsut joined 04:50 jsut_ left, Kulag left, rohit_nsit08 left, Kulag joined
dalek kudo/nom: c58a3dd | pmichaud++ | NOMMAP.markdown:
More NOMMAP notes.
 05:14
05:19 fperrad joined
dalek kudo/nom: 27cef7c | pmichaud++ | src/core/ (2 files):
Gather needs to execute its body in sink context.
 06:45
kudo/nom: 8c2ecf4 | pmichaud++ | src/core/ (2 files):
Add Any.grep, Any.min. Fix &infix:<cmp> to handle infinity arguments.
06:48 Eclesia joined
Eclesia hi 06:48
sorear hi 06:49
06:51 ligne left, ligne joined 07:03 Kulag left, Kulag joined
dalek kudo/nom: b1195a8 | moritz++ | t/spectest.data:
three more passing spectest files
 07:05
07:05 zby_home left 07:28 jsut_ joined 07:33 jsut left 08:04 mj41 joined 08:09 Kulag left 08:11 Kulag joined 08:12 contingencyplan joined 08:17 Kulag left 08:36 ligne left
dalek kudo/nom: 01af805 | moritz++ | NOMMAP.markdown:
add missing regexes to NOMMAP
 08:53
09:12 JimmyZ joined
dalek kudo/nom: 30f5cc9 | pmichaud++ | src/core/MapIter.pm:
Add last/next/redo handling to map loops in MapIter. Unfortunately,

for much of the MapIter body, although it's very straightforward code. Fortunately, this gives us another huge speed boost, such that using map to iterate over a range is twice as fast as it was before
  (about 80% faster than master).
while loop with an increment variable.		 09:15
09:19 jsut joined 09:24 jsut_ left 09:25 contingencyplan left
tadzik great news 09:54
dalek kudo/nom: 3f99683 | moritz++ | t/00-parrot/09-pir.t:
adapt t/00-parrot/09-pir.t to current reality
 10:02
10:03 daniel-s joined 10:05 woosley left 10:40 rurban_ joined 10:43 rurban left, rurban_ is now known as rurban 11:31 lichtkind joined 11:53 mj41 left 11:59 mtk joined 12:04 Kulag joined 12:25 Drossel joined 12:27 Kulag left 12:34 whiteknight joined
whiteknight good morning, #parrot 12:34
12:40 bubaflub joined
Eclesia hi whiteknight 12:41
whiteknight Hello Eclesia. How are you doing today? 12:42
Eclesia fine, working in java, like all monday ^^ 12:43
12:44 benabik left
bubaflub ~ 12:45
whiteknight good morning bubaflub 12:51
bubaflub morning whiteknight
how's life?
whiteknight Good. I'm on vacation next week, so I'm looking forward to that
dukeleto ~~ 12:52
whiteknight hello duke! It's a pleasure to see you online this early in the morning 12:53
but since you're probably so bleery-eyed, I won't talk too much until you get coffee :) 12:54
mikehh hi whiteknight, how you doin' today 12:55
whiteknight hello mikehh! I'm doing surprisingly well, for a monday.
dukeleto whiteknight: i already went for a run and took a shower sans coffee
12:56 benabik joined
dukeleto goes to get his conf badge 12:56
mikehh just built a new system, core i7 2600 16GB Ram and for the moment 2TB Hard drive 12:57
haven't installed op system yet
benabik is at breakfast. Red shirt, brown hat. :-)
whiteknight mikehh: oh wow. Sounds like a monster.
mikehh still working with my old 4 core amd phenom 8GB RAM 12:58
12:59 moritz left
mikehh the new one should be faster, morew memory and 4 cores but 8 threads 13:00
Eclesia he calls that old ... 13:04
mikehh Well I built it about 2 years ago... 13:05
Eclesia it has more memory then my current one lol
mikehh My son just got a 6 core/12 threads Extreme edition with 24GB RAM and one of those Top end ATI graphics cards (spent about twice what I did) 13:07
whiteknight My son bit me on the shoulder and drew blood 13:08
benabik ( whiteknight's kid )--
mikehh even my grandson's have stopped doin' that sort of thing 13:09
mind you they are 9 now, will be 10 in december
13:10 cotto left
mikehh thinks I must be gettin' old or somethin' - so I am told anyway :-} 13:11
13:11 bluescreen joined 13:12 cotto joined 13:13 fperrad left 13:16 fperrad joined 13:19 cotto left, cotto joined 13:28 cotto left 13:32 cotto joined 13:33 lucian joined 13:35 cottoo joined 13:36 cotto left
dalek rrot: 643dba0 | mikehh++ | src/oo.c:
fix codetest failure - linelength, update copyright
 13:39
Eclesia copyright is bad :p 13:42
mikehh Eclesia: seems to be required by our coding standards, so may as well keep it up-to-date :-} 13:43
and copyleft seems to have problems too 13:45
atrodo =~ Good morning from asheville! 13:48
whiteknight good morning atrodo
mikehh hiho atrodo
have you connected with our other intrepid parrot developers there 13:49
atrodo not yet, just got setup
But i'm pretty sure I just saw Util
benabik I see Util. He's in the lecture room thing. 13:50
I like the dispy bot in #yapc. Displays the title for any link posted. 13:52
atrodo That is a nice feature 13:53
this is a very long room 13:55
13:59 cottoo left 14:03 PacoLinux joined
dukeleto atrodo: where are you in this ridiculously large room? 14:08
atrodo far right, second row from the front and in the middle
yellow brown shirt 14:09
14:11 contingencyplan joined
dalek nxed/include: 68eab5e | NotFound++ | winxedst1.winxed:
implement $include
 14:19
nxed/include: 462f52d | NotFound++ | winxedst1.winxed:
fix local variable hiding parameter with the same name
 14:24
mikehh All tests PASS (pre/post-config, make corevm/make coretest, smoke (#16468), fulltest) at 3_5_0-45-g643dba0 14:25
Ubuntu 11.04 i386 (gcc --optimize)
14:36 hercynium joined 14:37 ambs joined 14:44 alester joined 14:46 bluescreen left 14:51 bluescreen joined
dalek nxed: fb73827 | NotFound++ | winxedst1.winxed:
fix extern function at root namespace level
 14:51
14:54 jsut_ joined 14:55 jevin left
dalek nxed/include: fb73827 | NotFound++ | winxedst1.winxed:
fix extern function at root namespace level
 14:56
nxed/include: 51482b9 | NotFound++ | winxedst1.winxed:
Merge branch 'master' into include
14:56 jevin joined 14:59 jsut left
Eclesia NotFound: just asking, is namespace which are not identifier support, somewhere on your roadmap ? 15:17
NotFound Eclesia: yes, but is not a big priority right now. 15:28
Maybe after extern function and $include got finished, but I don't think I'll add more features before the next stable parrot release. 15:29
Eclesia hm.. so I'll have to change all my namespaces .. ok 15:34
15:35 benabik_ joined 15:36 benabik left, cottoo joined, benabik_ is now known as benabik 15:43 fperrad left, fperrad joined
atrodo dukeleto++ 16:02
16:05 Eclesia left
cottoo kid51++ 16:07
dalek p: fa94a60 | pmichaud++ | src/PAST/NQP.pir:
Add first pass at handling lexical returns; create a new :pasttype<lexotic>
 16:20
16:29 theory joined
dalek kudo/nom: f647b57 | pmichaud++ | / (3 files):
Switch to using lexical returns. A return invocation now attaches

purely dynamic. This implementation uses continuations directly instead of Parrot's exception handling, which may be both good and bad. We'll have to see how it all works out.		 16:30
kudo/nom: 8b62352 | pmichaud++ | t/00-parrot/09-pir.t:
Merge branch 'nom' of github.com:rakudo/rakudo into nom
16:39 hercynium left
whiteknight I've been wanting NQP to do that same exact thing for a long time 16:43
use return continuations instead of return control exceptions
16:45 lucian left 16:52 cottoo left 16:58 benabik left
dalek nxed/include: 4282a77 | NotFound++ | t/ (2 files):
test $include
 16:59
JimmyZ wonders why not use include instead of $include 17:20
17:23 ambs_ joined 17:26 ambs left, ambs_ is now known as ambs 17:38 dmalcolm joined 17:40 mj41 joined
dalek kudo/nom: 3cefeb4 | pmichaud++ | / (4 files):
Make fail() into a soft-fail.
 17:41
17:49 JimmyZ left 17:51 aloha left 17:52 aloha joined 17:57 moritz joined
dalek kudo/nom: c461a7d | pmichaud++ | / (2 files):
Update NOMMAP (remove fail() from things-to-do).
 18:09
kudo: 1fcb142 | (Solomon Foster)++ | src/core/Complex.pm:
Fix (-1) ** -i bug by switching to Complex log.
 18:10
18:18 benabik joined 18:31 hercynium joined 18:37 kid51 joined 18:38 cotto joined 18:40 rurban_ joined 18:43 rurban left, rurban_ is now known as rurban 18:46 Eclesia joined
Eclesia hi 18:47
dukeleto ~~ 18:48
Eclesia: hola
Eclesia what are the ~ for ?
whiteknight just a short way to say hello 18:49
dukeleto: how's the conference going? 18:50
cotto no injuries so far 18:51
Coke ~ === wave
18:57 hercynium left
Eclesia ... 18:59
NotFound: _ are valid in namespace in winxed ? 19:00
19:00 kid51 left, hercynium joined
dukeleto whiteknight: quite well 19:02
whiteknight: we had a parroty lunch with many parrot hackers 19:03
whiteknight: cotto and I have been thinking about what kind of language we should use to generate M0 19:04
whiteknight: have any ideas?
cotto at this point, criteria are more important than a specific language recommendation 19:08
whiteknight for the lowest-level system language that compiles down to M0, I suspect we want something syntax-light and familiar enough to get started working with it 19:09
something that's easy and efficient to compile seems like a good idea too, but easy-to-implement compilers might take precidence 19:10
cotto a very straightforward translation between the language and M0 is important 19:13
some magic, but not much 19:14
whiteknight I do keep coming back to the idea of something C-like, since those kinds of compilers are well understood and the syntax is very familiar to most parrot hackers
benabik +1 19:15
whiteknight although, something lighter on syntax like a ruby or python subset (minus much of the whitespace issues) would probably work just as well
cotto it's also important to keep in mind that we'll be writing a ton of code in this language
whiteknight if we could modify Winxed to do it, I think I would prefer that the most. I don't know if winxed is in that kind of condition. A lot of winxed features rely on PIR features, which might not map well at this stage 19:16
cotto c-like is great
19:16 zby_home joined
whiteknight so you would need a pared-down winxed subset, which isn't unthinkable 19:16
cotto if we wanted winxed, we'd need to fork it anyway 19:17
whiteknight right
and like I said, a lot of winxed syntax would need to be cut out
namespaces, class, etc
Eclesia if I understand correctly : cotto want pir, but with c-like syntaxe. nothing else ? 19:18
cotto yeah. the object system will need minimialism
Eclesia, no
whiteknight Eclesia: we want a very low-level language to replace PIR
cotto and C
Eclesia replace C ? 19:19
whiteknight cotto: I say we take a list of capabilities of M0, then take C and start ripping out everything from C that isn't covered by M0 or that doesn't map nicely, or that isn't required for our purposes
Including the entire C preprocessor
because implementing macros is too big a pain for too little benefit
cotto whiteknight, I like that 19:20
whiteknight We can write that compiler in any language. Winxed might be a very good choice for that purpose
sorear will C-M0 have structs?
cotto we don't need a strict subset of C, but something close to C would be great
sorear, it will have to
whiteknight right, but that's a good base to start with. Remove what we don't need, add what we do need, and try to keep it easy and familiar 19:21
19:22 kid51 joined
sorear if you're going to make a C-like language I have two small requests to make 19:22
cotto sorear, please do
sorear 1. int* x, y; declares two pointers, like C# (int* is just a type, no "declaration mirrors use" magic) 19:23
2. functional casts int(x); ((int)x) is obnoxious to write and parse
cotto that bit of C isn't well thought out. +1
also a nice idea
any suggestion for a name for this thing? 19:24
Eclesia (remove pointers, values by copy, object by ref, much easier to understand)
19:25 p6eval left
dalek rakudo/nom: 9a33668 | pmichaud++ | src/ (3 files): 19:25
sorear Eclesia: this is the language we're going to implement the object system *in* 19:26
cotto yes
sorear Eclesia: it won't have objects
except as a matter of convention (PMC*) 19:27
Coke cotto: why not just write it in C to start with. then you can see what you needed or not. Rather than adding an entire new language to the pile of things that needs to be created. 19:28
19:28 dalek left
whiteknight cotto: M0C 19:28
Coke (design it ex post facto)
cotto Coke, it needs to compile to M0 19:29
whiteknight Coke: because we need the language to build M0. We're going to need to write a compiler anyway to generate the M0 code we want
so we can try to modify an existing C compiler to do what we want, including ignoring all the C syntax we don't want or can't support
Coke Ah, so you can't use C for free to start with.
whiteknight right. Would if we could 19:30
cotto whiteknight, I'm not sure about that approach. I guess it depends on the quality of the code of the compiler and how easy it is to retarget.
benabik I really think int* x,y being two pointers would be very confusing. Esp since I'd write int *x, y; 19:31
Although I guess int* and int * could be different...
whiteknight cotto: Yeah, I don't recommend it. I'm simply pointing that out as the alternative, if we don't make our own compiler for it
well, before talking about declarations, what types does M0 handle? 19:32
cotto not having to write a full compiler is nice though
whiteknight do we have more than just intval, floatval, string, pmc, and maybe a void* ptr type?
If our type system is confined enough, we don't need things like * for pointers or other gabbledegak
benabik True, and possibly better. 19:33
cotto I like that idea.
whiteknight we only need "int*" if we want to differentiate between "pointer to an int" from "just another pointer"
cotto not sure if it'll be feasible though
19:34 p6eval joined
benabik Although isn't there no difference between string and pmc in M0? 19:34
whiteknight I haven't read the M0 design docs in a little while, I don't even remember what all types it can work with
cotto benabik, right.
NotFound A backend for a subset of winxed is doable. Is what winxedxx does.
cotto there'll probably also be primitive strings that aren't entirely unlike C-strings
whiteknight Notfound: Could you start writing up some details about how to do that? I would like to use Winxed or something winxed-alike if possible 19:35
NotFound The basic recipe is simple: just use the "target": "parse" in the options passed to the compiler, and then do things with the result ;) 19:37
19:37 kid51 left, dalek joined
NotFound The only problem is the high coupling with the implementation of the compiler. 19:38
cotto I've got a gist I'll use to start a wiki page. 19:40
NotFound BTW I think the clean and simple solution to the "int* x, y;" thing is not allowing multiple declarations. 19:41
19:41 benabik left
cotto or make int* the type 19:41
NotFound cotto: if the goal is to make the language close to C, that is a big source of confusion. 19:42
cotto NotFound, yeah
I wonder if this is one of those things like "say" that will be different but still feel natural and sensible 19:44
NotFound About the functional cast, there is also a solution: being compatible with a subset of C++ instead of a subset of C.
19:50 zby_home left
whiteknight nobody wants compatibility, I don't think. Just familiarity 19:50
it doesn't need to be C or C++, just needs to have some of the same kinds of syntax 19:51
We could just as easily be inspired by JavaScript, and not lose anything
NotFound Comptibility has the advantage of easier testability of proofs of concept. 19:52
cotto whiteknight, yes 19:55
C is nice because Parrot's currently written in it.
Eclesia interrupt one second : how to load a pbc in winxed at any time. 19:56
Eclesia ?
NotFound Eclesia: load_bytecode(filename)
whiteknight eclesia: load_bytecode("foo.pbc")
:)
Eclesia thanks
19:57 cjh left
cotto I suspect (but can't verify) that attempting to maintain compatibility with an existing language will be more work than it's worth 20:00
NotFound We have powerful tools for writing languages, the hypothesis may be easily verifiable. 20:02
20:02 hercynium left 20:04 daniel-s left
whiteknight cotto: I don't want compatibility. Just familiarity 20:05
take ideas from existing syntax, but don't be a slave to it
cotto whiteknight, that's great. I agree.
whiteknight Basically, when I say a "C-alike syntax" I'm talking about post-circumfix parenthesis being a function call, brackets to delimit blocks, etc 20:06
infix operators for arithmetic, common keywords like "if", "else", "while", etc.
that's all I want
cotto and when I say "lisp-like", the room suddenly gets quiet
;)
whiteknight (we (don't (want (lisp))) 20:07
cotto (which is one of the many reason I'm not suggesting it)
whiteknight or, let me rephrase that. (I (don't (want (lisp))))
cotto the first version was fine
whiteknight I won't speak for (we)
cotto or at least some subset of "we" 20:08
whiteknight I put the "I" in "we"
NotFound sing We are the world
cotto I are the world? 20:09
20:10 benabik joined
NotFound #define I We 20:10
cotto is watching kid51++ talk at yapc 20:16
whiteknight oh, what's he talking about?
cotto this might sting a bit
recording now
whiteknight what might sting? 20:17
cotto the talk
benabik "Open Source Software Projects: Two Stages and Ten Questions" 20:18
20:18 soh_cah_toa joined
whiteknight why would it sting? I'm not following 20:22
dukeleto ~~ 20:24
dukeleto is listening to kid51++ talk
atrodo is also listening to kid51++ talk 20:25
benabik Ditto, ditto, etc etc
Eclesia suffer a parrot overdose. going to bed +++ 20:26
20:26 Eclesia left
whiteknight if you're going to overdose on anything, Parrot is a good choice 20:27
benabik Yar, a little bit of sting so far. Brutal honesty and all. :-) 20:29
NotFound I'm thinking about how and where to install winxed include files. The 'inst_inc' section in distutils seem a bit limited.
20:30 whiteknight left
dukeleto bubaflub: ping 20:33
bubaflub dukeleto: pong 20:35
dukeleto bubaflub: i have a parrot-gmp repo. parrot setup.pir doesn't do anything yet. What do I do to take it for a test drive? 20:37
bubaflub dukeleto: ah, yes. sorry about that. i haven't gotten to setup.pir yet.
dukeleto: i've included the generated PIR from the Winxed sources. are you at HEAD? 20:38
dukeleto bubaflub: yep, just cloned a minute ago
bubaflub parrot -o src/GMP/Common.pbc src/GMP/Common.pir
parrot -o src/GMP/Integer.pbc src/GMP/Integer.pir
ooooh, sorry, before that 20:39
parrot -o src/GMP/raw.pbc src/GMP/raw.pir
then you should be able to do parrot-nqp t/harness
dukeleto bubaflub: this leads me to believe that you should spend 5 minutes to make setup.pir work :) 20:40
bubaflub dukeleto: indeed.
dukeleto $ parrot-nqp t/harness 20:42
"load_bytecode" couldn't find file 'rosella/harness.pbc'
bubaflub: i have rosella, but not installed 20:43
bubaflub: how do I tell the harness where to find it?
bubaflub dukeleto: ah, i'm not entirely sure... i just have rosella via plumage. perhaps there is some way to include a path with parrot-nqp?
dalek rrot: 8352b99 | fperrad++ | tools/dev/fetch_languages.pl:
[languages] Winxed moves to Github
 20:44
dukeleto bubaflub: i can try to install via plumage. you should document that that is the preferred way to get parrot-gmp and rosella to play nice 20:46
bubaflub dukeleto: while i figure that out, you should be able to run winxed examples/basic.winxed
dukeleto bubaflub: yay, it worked 20:48
bubaflub dukeleto: *phew*
dalek kudo/nom: b79c7b3 | (Martin Berends)++ | src/core/Cool.pm:
[src/core/Cool.pm] add sub equivalents of 11 Str methods
 20:50
20:51 fperrad left
benabik "shoving Parrot up Firefox" Interesting phrase 20:51
atrodo a parrot up a fox... hmmm...
dalek rrot-gmp: 6882f70 | bubaflub++ | examples/basic.winxed:
we've got a VTABLE override, let's use it
 20:52
rrot-gmp: 49d9982 | bubaflub++ | README.md:
clarify dependencies
20:55 ambs left
cotto and back 20:58
dalek kudo/nom: 0594102 | (Martin Berends)++ | t/spectest.data:
[t/spectest.data] uncomment tests for chomp chop and flip
 21:08
atrodo Oh, good, looks like ipfy will have some new data soon 21:12
21:12 bluescreen left 21:33 soh_cah_toa left
bubaflub seen plobsing 21:41
aloha plobsing was last seen in #parrot 3 days 3 hours ago saying "wasn't that ES4?".
21:41 soh_cah_toa joined 21:42 Psyche^ joined 21:47 Patterner left, Psyche^ is now known as Patterner 21:52 daniel-s joined
dalek rrot: c899706 | pmichaud++ | compilers/pct/src/PAST/ (2 files):
[pct]: Add 'signature' attribute to PAST::Node, to explicitly request a particular signature for evaluating children.
 21:56
benabik Ahhh, more PAST changes! ;-) 21:57
dalek p: 342cd53 | pmichaud++ | tools/build/PARROT_REVISION:
Bump PARROT_REVISION to get PAST::Node signatures.
pmichaud this change to PAST is pretty small. :) 21:58
but it increase flexibility a lot.
*increases
21:59 jsut joined
benabik 'sall good. 21:59
21:59 soh_cah_toa left
jnthn__ pmichaud: ooh 21:59
pmichaud: How's that one used?
NotFound dukeleto: about the load_bytecode problem, you can try to set the PARROT_LIBRARY env var. 22:00
benabik PCT needs more usage doc across the board. :-/
22:02 benabik left
jnthn__ pmichaud: Wait, does this mean we can say which node in a multi-node element is the result one? :) 22:03
22:04 jsut_ left 22:07 lichtkind left
NotFound pmichaud: There is command or something to update nqp-rx stage 0 files? 22:08
pmichaud jnthn__: yes, that's exactly what it does. :) 22:09
similar to what we can do for pirops. But I'd only use it on PAST::Stmt, PAST::Stmts, and PAST::Op :pasttype<stmts> 22:10
in other cases it's likely to not dtrt
jnthn__ Najs! :D
dalek kudo/nom: dc16a4f | pmichaud++ | / (3 files):
Refactor return handling to be in a single place (DRY). Handle the case where someone tries to return from an already-exited subroutine.
kudo/nom: 3efe8f8 | pmichaud++ | NOMMAP.markdown:
NOMMAP update.
pmichaud mainly it means you can specify a sequence of things to be done but you're no longer limited to the last one (as long as you don't have more than ten things to be done... but if that's the case you can factor it anyway :) 22:11
...limited to the result being the last one.
In this case, I had some cleanups that needed performing after the calculations but I didn't want them affecting the overall result. Fixing it with if nodes and temporary registers was going to be a pain. 22:12
this is a lot more straightforward :)
22:12 whiteknight joined 22:13 whiteknight left 22:14 whiteknight joined, cotto left
dalek kudo/nom: 872a536 | jonathan++ | NOMMAP.markdown:
Comment on a couple of nommap items.
 22:15
jnthn__ pmichaud: I can see that reducing pain in a few cases I've had in the PAST, yes.
22:15 lichtkind joined
whiteknight sorear++ 22:23
dalek kudo/nom: 4fc90f1 | jonathan++ | src/Perl6/Metamodel/ParametricRoleHOW.pm:
Fix role-based dispatch up a bit more. Should be able to distinguish by Positional and Associative now.
 22:31
22:35 PacoLinux left 22:39 eternaleye left 22:40 mj41 left, eternaleye joined
dalek kudo/nom: eda6101 | jonathan++ | src/Perl6/Metamodel/ParametricRoleHOW.pm:
is_composed in ParametricRoleHOW so we can stub a role.
 22:48
kudo/nom: 144660f | jonathan++ | src/core/traits.pm:
Stub Positional/Associative/Callable in preparation for making sigils map to roles. Need to do this really early.
22:51 cotto joined 22:56 dmalcolm left 23:05 daniel-s left
dalek kudo/nom: 0035ddb | jonathan++ | / (5 files):
Roles should be narrower than Cool/Any/Mu.
 23:07
23:22 bluescreen joined
NotFound I'm glad to see winxed mentioned in a presentation at YAPC, but a little disapointed seen it descibed as a library. 23:26
whiteknight who described it as a library? 23:27
NotFound James E Keenan to parrot-dev - Text in Open Office format: thenceforward.net/perl/yapc/YAPC-NA...1/ossp.odt 23:29
23:34 cottoo joined 23:35 cotto left
dalek kudo/nom: f670236 | jonathan++ | src/Perl6/Actions.pm:
Dispatch by sigil.
 23:36
kudo/nom: d6c905b | jonathan++ | / (2 files):
Fix &infix:<cmp> proto.
 23:41
23:53 Drossel left, Kulag joined

Powered by Cro and the Raku® Programming Language.

Please report any issues / comments / feature requests as an issue on App::Raku::Log.

Thank you!