Parrot 2.10.1 Released | parrot.org | Log: irclog.perlgeek.de/parrot/today | Long live Git! github.com/parrot/parrot | git clone git://github.com/parrot/parrot.git
Set by moderator on 18 November 2010.
whiteknight rfw: yes, those tests look nice 00:00
rfw is that all you need for GEMM?
whiteknight rfw: same tests with non-numeric alpha and beta, and yes we should be good
rfw awesome
whiteknight rfw++
rfw is incremented o/ 00:01
whiteknight it's channel karma
karma rfw?
aloha rfw? has karma of 0.
whiteknight rfw++
karma rfw?
aloha rfw? has karma of 0.
rfw karma rfw
whiteknight karma rfw
aloha rfw has karma of 2.
whiteknight that's it
I'm being stupid tonight
rfw now i have too much :D
jnthn karma jonathan
aloha jonathan has karma of 103.
jnthn *that's* too much :)
rfw hehe
whiteknight rfw: don't worry, you'll grow into it
jnthn Or grow out of it... 00:02
rfw whiteknight: what happens when you *do* gemm with a string alpha/beta? 00:03
i can't test it :(
whiteknight this software problem is turning into a real draf
drag
I'm going to have to set up archlinux in a VM and figure it out 00:04
let me test it for you
rfw hm, can't you just run it on your instance?
i just need to know what happens
i don't need you to figure out why my vm is broken, lol
whiteknight I'm working on building it all
rfw that's going to take a while :( 00:05
whiteknight I'm already done. my box here is pretty speedy
rfw ah
Kovensky what happens if one does ++rfw++
rfw thanks for going to all that trouble for lil' ol' me :( 00:06
karma Kovensky
aloha Kovensky has karma of 0.
rfw :<
whiteknight Kovensky: I don't think we can pre-increment it
jnthn On #perl6 we use pre-increment as a way of encouraging people with potential karma if they do something in the future. :) 00:07
But you have to Do Stuff to get postfix karma that counts. :P
Kovensky #perl6 looks... empty
whiteknight it's on freenode
Kovensky ic
you have a dalek? cool 00:08
reminds me to resume watching dr who, I'm at like ep7
nopaste "Whiteknight" at 192.168.1.3 pasted "results of ComplexMatrix2D.gemm() with a string for alpha for rfw++" (8 lines) at nopaste.snit.ch/26197 00:12
rfw yeah i know
lol
i had 0+01
isntead of 0+0i
whiteknight: pastebin.com/4qaxLBjY 00:13
Kovensky <+nopaste> "Whiteknight" at 192.168.1.3 <-- cool IP
rfw ddoses 192.168.1.3 00:14
ohshi-
whiteknight rfw++
rfw i feel like i'm getting too much karma :(
whiteknight it's cheap 00:15
for every 10,000 you get, we send you a fresh donut
rfw oh boy
jelly donut?
whiteknight sure, why not. We've never done it so the first person gets to set precident
Kovensky lol
rfw whoops, my $alpha := pir::box__Pi(2);
not my $alpha := pir::box_Pi(2);
Kovensky does rfw-- for balance 00:16
karma rfw
aloha rfw has karma of 5.
jnthn Dva underscores. :)
rfw jnthn: not quite used to it :<
jnthn You'll get there.
whiteknight he is picking it all up very quickly
rfw every programming language is like every other, just some have weirder symbols 00:17
Kovensky like ruby?
or APL?
rfw ohgoddonteventalktomeaboutruby.jpg
haha apl
jnthn bf? :) 00:18
Kovensky TECO?
rfw well now you're just sily
jnthn You could always program in the Whitespace language. It's totally free of weird symbols, I here.
*hear 00:19
rfw whit--
goddammit
jnthn: saves paper to print too
er
s/paper/ink/
NotFound Some pople consider hard tabs weird.
jnthn The paper is easily recycled, though.
Plus you can use it in a script tag in HTML without having to add <!-- --> around the code :)
whiteknight rfw: your assert_throws calls don't have a ); at the end
rfw oh 00:20
whoops
Kovensky whiteknight: missing a sad face? :>
rfw i was treating it like a statement
Kovensky NotFound: they are weird
whiteknight it's a method
rfw assert_throws(Exception::Brazilian, "is brazilian", { my $kovensky := new Kovensky(); });
that took a while to type 00:21
Kovensky why is there a pascal-like assignment there
in b4 that made it into perl6 D:
rfw yup
:D
Kovensky cries
rfw pats Kovensky
it's okay
there are still other pure languages out there!
like python!
Kovensky or perl5!
00:21 Infinoid left 00:22 hercynium joined
rfw Kovensky: now you're just being silly 00:22
Kovensky perl5's major deficiency for me is the fail unicode support on windows
it's like... useless ._.
00:23 Infinoid joined
Kovensky the only implementation that works properly with unicode is cygwin's, but nobody uses cygwin 00:23
rfw whiteknight: does it work fine with the );? 00:24
whiteknight: if it works then i think i'm done lol 00:30
though my tests are a little half-assed
whiteknight does that last nopaste have all your tests, or are their more? 00:31
there*
rfw for gemm only yeah
i have a bunch of other tests in other test files
like complex conjugate with non-square
and item_at with complex numbers
i can test the rest myself fine :p 00:32
00:34 dngor_ joined
rfw i haven't written any of the method test_VTABLE_get_string_keyed() etc. 00:35
it'd take me a while to learn parrot bytecode ._.
00:37 TypeNameHere_ is now known as TypeNameHere_____, dngor left
rfw whiteknight: still there? 00:40
whiteknight of course
rfw ah
did you try it with the );
whiteknight yes 00:41
rfw did it work?
whiteknight yes 00:42
the assert_throws tests are still weird. I'm looking into it
rfw hm
the other two work fine, right?
sorry for making you do part of my task ^^ 00:43
whiteknight it's okay, you're dealing with unforseen difficulties
cotto Does Parrot have clearly defined goals or is it mostly "try to get faster and stop annoying Rakudo and Partcl"? 00:49
whiteknight none that I've heard explicitly defined 00:51
mikehh All tests PASS (pre/post-config, make corevm/make coretest, smoke (#1365) fulltest) at 9a7cfab - Ubuntu 10.10 i386 (g++-4.5 with --optimize) 00:52
whiteknight cotto: why do you ask?
rfw whiteknight: hm from the C source i can't see anything that would go wrong...
whiteknight rfw: is that sarcasm?
rfw no 00:53
well
now that you mention it i probably needed to rephrase it
i can't see anything in the gemm function that would make it behave too weirdly
cotto whiteknight, I'm working out the core inclusion policy, and that depends somewhat on what Parrot's goals are.
rfw whiteknight: what exactly does happen? 00:54
whiteknight "try to get faster and stop annoying Rakudo and Partcl" really seems short-sighted
rfw: what do you mean?
sorear cotto: parrot *had* clearly defined goals at one point, but I think we've thrown out most of them as conflicting with "try to get faster and stop annoying Rakudo and Partcl"
rfw whiteknight: what makes it behave weird
whiteknight I'm not sure yet. I'm still digging into it 00:55
rfw ah
does it pass or fail?
whiteknight it fails, so it either doesn't throw or doesn't throw what you're expecting
cotto sorear, what do you mean?
rfw ah 00:56
whiteknight rfw: the exception is coming from Complex PMC's parse routine, so it's probably not OutOfBounds
sorear cotto: successor to p5vm and support all dynamic languages
whiteknight but again, I'll play with this one to make it pass. When the rest of your tests are ready you can call it complete
rfw hm hold on
whiteknight cotto: A nice mission statement would be a very good thing for us to hammer out 00:57
cotto +1
whiteknight being faster is nice, but that's a means to an end, not the end itself
we could be extremely fast if we cut out all our features and had 0 code
cotto It'd be good to come up with something for the pds.
whiteknight yes
at least a first draft to present and get feedback on
rfw whiteknight: get_complex_value_from_pmc(PARROT_INTERP, PMC * value, FLOATVAL * real, FLOATVAL * imag) is where it parses it, right 00:58
cotto right
whiteknight rfw: sort of. It hands the string off to the Complex PMC type and uses that for parsing and extracting 00:59
rfw ah 01:00
whiteknight The answers lie somewhere in the source of the Complex PMC
rfw ah okay
cd ..
erm
i amvery bad at alt-tabbing today
whiteknight cotto: Something like "to become the premier dynamic language runtime environment" is short but a nice start
cotto It needs a bit of expansion, but I like it. 01:01
whiteknight "...and to not piss off partcl any more than we already have"
cotto we'd probably fail at that goal before we could finalize the mission statement 01:02
whiteknight :)
cotto not a good thing, but likely nonetheless
NotFound "and get some ponies"
whiteknight My hope is that one day using Parrot as a base for new dynamic languages will be the best decision because the things Parrot provides are so far ahead of what any other project could expect to do itself in a reasonable time 01:03
NotFound whiteknight: I thins it already is. 01:04
rfw whiteknight: do you prefer a patch or a fork+pull 01:06
dalek rrot/generational_gc: f6c8b56 | nwellnhof++ | src/gc/gc_ms2.c:
Fix pointer aliasing problem in optimized build

Inlining gc_ms2_set_gen_flags broke GCC's strict aliasing assumptions. Fixes the optimized build on Linux/i386.		 01:07
rrot/generational_gc: e9d88fa | nwellnhof++ | src/gc/gc_ms2.c:
Don't overwrite next pointer of free list.
rrot/mappedbytearray: 663cc80 | NotFound++ | src/pmc/mappedbytearray.pmc:
add init and get_bool vtables and close method, change the handling of the custom destroy flag
 01:11
rfw whiteknight: dl.dropbox.com/u/2880866/0001-Added...-GCI.patch 01:12
whiteknight: can you take a look at it and see if it's acceptable? 01:20
dalek rrot/mappedbytearray: 7f0b8de | NotFound++ | / (2 files):
first tests for MappedByteArray PMC
 01:25
rrot/mappedbytearray: 7344aaf | NotFound++ | t/pmc/mappedbytearray. (2 files):
ooops, wrong filename
 01:27
rrot/mappedbytearray: cc4a836 | NotFound++ | MANIFEST:
wrong MANIFEST too
 01:29
NotFound (Note to self: don't commit when tired)
whiteknight rfw: either is fine. 01:35
rfw whiteknight: link up there
whiteknight got it 01:36
rfw: it all looks good 01:37
rfw \\o/
www.google-melange.com/gci/task/sho...9033196801 01:38
wait i have to submit a link
whiteknight actually, is all this in a public repo? I'll just pull it
rfw ah no
it's in a local repo
you can apply it with git am
whiteknight yeah, I know
rfw alright, work submitted
dalek rrot/mappedbytearray: d4c2cd2 | NotFound++ | src/pmc/mappedbytearray.pmc:
codingstd fixes
 01:40
whiteknight blah. it won't apply. Is there some secret? 01:41
rfw hm?
it should do
i put it on a clean copy of the repo
i even recloned it lol 01:42
whiteknight what's the command, git am foo.patch?
rfw yeah
should be
cotto whiteknight, there's some extra stuff at the top that might be messing up patch
rfw cotto: it should be fine with git am 01:43
whiteknight: i could push it i guess
if that's any easier
whiteknight what's your github id?
rfw rofflwaffls
hold on, let me fork 01:44
"Hardcore Forking Action"
oh you, github
whiteknight I added you as a contributor to mine
just push it there
rfw oh 01:45
whiteknight I'm trying to find a way to waste the least amount of your time at this point
rfw hehe it's fine
done
whiteknight perfect 01:46
marked complete. Congrats
thanks for the help
rfw any time
you open source developers are all really awesome
whiteknight and we're all really handsome
:) 01:47
NotFound rfw: s/you/we
whiteknight I'm signing off for the night. I'll see you later
rfw see you
NotFound: one of us?
01:47 whiteknight left
NotFound rfw: you have contribued, you are a open source developer too. 01:47
rfw hehe
dalek rrot/mappedbytearray: cdb2fac | NotFound++ | / (61 files):
Merge branch 'master' into mappedbytearray
 01:48
rrot-linear-algebra: f4ba8df | rofflwaffls++ | t/methods/complexmatrix2d/ (10 files):
Added ComplexMatrix2D test cases (GCI).
 01:50
cotto msg whiteknight Do you have the tuits to come up with a first draft of a mission statement? 01:52
aloha OK. I'll deliver the message.
cotto rfw, thanks for contributing! 01:53
rfw so problem :)
no*
cotto pla is mostly over my head, but I'm happy to see any parrot-based project getting some love. 01:54
NotFound A program that simulates some physical process with PLA and draws it with opengl will be a nice demo of parrot abilities. 01:57
A pogram written in some parrot language, of course. 01:58
02:32 TypeNameHere_____ left
sorear NotFound: have you seen moritz' Math::Model work? 02:45
dalek TT #1873 created by jkeenan++: Windows2000 version requirement impeded build 03:10
TT #1873: trac.parrot.org/parrot/ticket/1873
03:41 kid51 joined
kid51 I'm getting a build failure in master 03:42
I'm at commit 9a7cfabd75c82e447606fcca53c1fb98059c0a5e
The failure begins here: src/extra_nci_thunks.c:3151: error: expected ')' before string constant 03:43
Observed on linux/i386.
nopaste "kid51" at 192.168.1.3 pasted "master: build failure at 9a7cfab" (750 lines) at nopaste.snit.ch/26200 03:46
kid51 Surprisingly, I'm *not* getting this failure on darwin/ppc. 03:50
In the build failure output, am observing several warnings in src/gc/fixed_allocator.c 03:53
03:54 TypeNameHere_____ joined
kid51 I built successfully at this earlier commit: ba88b03c6b3702cd1fa66bea3fd7b8bdbe95e502 04:01
nopaste "kid51" at 192.168.1.3 pasted "master: A differernt build failure at the 2nd oldest commit" (23 lines) at nopaste.snit.ch/26201 04:11
GeJ Just built master successfully on FreeBSD 8.1 amd64. 04:17
Coke . 04:19
kid51 Hmm. I wonder if this is a git side effect. I have been switching back and forth among branches within a single directory. 04:20
The position of certain generated files many have changed over time, leading to 'make realclean' not doing the job.
Hmm. Hypothesis somewhat confirmed. extra_nci_thunks.c now in src/nci/. 'make' once again succeeding at head 04:23
04:27 kid51 left 04:49 theory left 05:20 rurban_ joined 05:22 rurban left, rurban_ is now known as rurban 06:05 diakopter joined
diakopter any nqp users around? 06:06
cotto diakopter, what would you like to find out? 06:07
diakopter oh; hi
I'm getting an error message 06:08
Method 'list' not found for invocant of class 'ResizablePMCArray'
cotto can you nopaste your code? 06:09
diakopter well it's any method on that
parrot-nqp -e "my @a := (1,2); say(@a.WHAT)" 06:10
cotto The list method is part of the nqp setting. It might not be getting loaded.
diakopter hm, oh
seems hard to imagine though
this is part of jnthn's port of nqp to the CLR backend
it's a few thousand lines of nqp-rx that work 06:11
I see plenty of places where for @($somelist) { } works fine 06:12
cotto: thanks for your helpfulness :) 06:19
sorear diakopter: nqp-rx doesn't use the nqp setting, so @( LIST ) won't work 06:22
I'm pretty sure for @( ) is a special case
I don't know about JnthnNQP and how it differs, though 06:23
diakopter JnthnNQP is written in nqp-rx. 06:27
@( ) works
when it's on an object that's a member of a class built in nqp-rx 06:28
that implements the proper methods
as I've just learned
(see my last push - I resolved it by using the object itself as storage) 06:29
moritz bacek: still a few segfaulting tests in rakudo 06:35
plobsing moritz: have you determined a reliable way of reproducing them? 06:39
moritz plobsing: yes, run "make spectest" in rakudo, you always get at least one failure 06:44
plobsing moritz: but do you get the *same* failure. also, it would make the problem a lot more tractable if I could run ./perl6 buggy-test.p6 in stead of waiting an hour for maybe something to happen. 06:46
moritz I know
maybe if I create a sufficiently long-running p6 script...
plobsing curses, you've seen my intentional loophole 06:47
diakopter I mean, if the demand for something reproducible weren't serious, it might be comical.. 06:48
moritz: try man-or-boy with higher inputs
plobsing moritz: there are 3 potential sources of randomness that I know of (more) 06:49
1) address space randomization 2) hash-seed 3) user-injected (using rand op or such) 06:50
moritz 1) can be disabled (but the segfault is still random
plobsing I suspect you aren't using the third, and I suspect you've turned off the first (sop for parrot debug)
moritz 3) doesn't appear in that many test files
plobsing that leaves the second as the likely culprit (I was hunting one of these that magically dissapeared half a year ago) 06:51
cotto 2) is easy to play with if you're running the parrot executable directly
plobsing and you *can* tweak the rakudo spectest harness to do exactly that
hash-seed is naively based on time. something along the lines of 'parrot --hash-seed $(time | tee timestamp.name_of_rakudo_test) perl6.pbc rakudo_test' 06:54
should emulate the behaviour exactly (and hopefully bring out the bug with a reproducable hash seed) 06:55
cotto you'd need to make time spit out an int, but that'd work
nice idea 06:56
plobsing time/date/some-command
cotto date +%N would be pretty random 06:57
plobsing the exact behaviour is epoch time in hex 06:58
diakopter how does one disable (or raise) the recursion limit 07:03
oh, I suppose that could be a rakudo thing. 07:04
cotto There's apparently a recursion_limit method on ParrotInterpreter 07:05
diakopter wonders if that's accessible from rakudo's Q::PIR or whatever 07:06
cotto should be
just use getinterp
whether it's a good idea is up to you 07:07
diakopter for trying to reproduce this segv
plobsing what's a test that often segfaults? 07:56
I'm going to probe for hash seeds that cause it to segfault
moritz plobsing: t/spec/S02-literals/types.t 08:02
S32-trig/tan.t 08:03
I just did
grep S32-trig t/spectest.data > t/localtest.data
make localtest
=> t/spec/S43-trig/{sin,cos,tan}.t failed
plobsing ooh fun, the trig tests. sinh.t was failing last time I was hunting this bug 08:04
(if its the bug I think it is) 08:05
moritz plobsing: t/spec/S02-literals/types.t is much shorter, and also aborted twice here 08:06
for i in `seq 100`; do echo $i; ./parrot_install/bin/parrot --hash-seed=$i perl6.pbc t/spec/S02-literals/types.t; done 08:08
segfaulted at 2
(but still with address space randomization)
plobsing probing with: 'parrot --hash-seed $(perl -e "printf qq{%x\\n}, time" | tee seed) perl6.pbc t/spec/S02-literals/types.t' 08:09
moritz: I cannot get it to fail. just ran it 20 times without event. 08:14
can you run the probe command I provided to get the hash seed of a segfaulting run (and then verify that that hash seed causes consistent failure)?
08:29 fperrad joined
plobsing moritz: what kind of platform are you on? 08:31
09:02 rfw left 09:29 kurahaupo joined 10:07 mj41 joined
dalek rrot: be82239 | bacek++ | src/gc/gc_ms2.c:
One more fix of GC MS2 for tracing stack pointers. Strings are stored in different Pointer_Array
 10:27
10:43 pjcj left 11:08 rurban left 11:14 rurban joined 11:22 mikehh left 11:32 pjcj joined 11:44 pjcj left 11:45 pjcj joined 11:51 PacoLinux left 11:52 wesjdj left 11:56 bacek left 12:00 bacek joined
dalek rrot: 0753977 | mikehh++ | t/steps/auto/revision-01.t:
fix test to aviod passing TODO's

I think this test is redundant at present as it was svn related and should probably be removed from the repository		 12:05
12:12 wesjdj joined 12:22 mikehh joined 12:39 PacoLinux joined
mikehh All tests PASS (pre/post-config, make corevm/make coretest, smoke (#1379) fulltest) at 0753977 - Kubuntu 10.10 amd64 (g++-4.5) 12:46
dalek rrot/mappedbytearray: 3e42be9 | NotFound++ | src/pmc/mappedbytearray.pmc:
method MappedByteArray.get_string
 13:06
13:19 rurban_ joined 13:22 rurban left, rurban_ is now known as rurban 13:26 Patterner left 13:29 mikehh left 13:36 Psyche^ joined, Psyche^ is now known as Patterner 14:00 mikehh joined 14:24 nwellnhof joined 14:35 janus joined
mikehh All tests PASS (pre/post-config, make corevm/make coretest, smoke (#1382) fulltest) at 0753977 - Ubuntu 10.10 i386 (g++-4.5 with --optimize) 14:48
nwellnhof I'm still getting lots of random segfaults in Rakudo's spectest. 14:52
15:20 whiteknight joined
whiteknight good afternoon, #parrot 15:27
msg cotto: I won't be on the computer most of the day today (son's birthday party). I will mull it over in the back of my head and send you something tonight
aloha OK. I'll deliver the message.
15:29 simon_ left
mikehh rakudo (82f3de0) - builds on parrot (513-g0753977) - make test PASS, make spectest_smolder[(#1383) roast (52c1e70)] FAIL - Ubuntu 10.10 i386 (g++-4.5 with --optimize) 15:31
26,308 ok, 51 failed, 609 todo, 1,854 skipped and 1 unexpectedly succeeded
15:38 kid51 joined
kid51 ~~ 15:39
whiteknight hello kid51 15:41
dalek rrot/mappedbytearray: 58cb23b | NotFound++ | src/pmc/mappedbytearray.pmc:
implement MappedByteArray functionality for mingw32
 15:44
whiteknight msg dukeleto I have a GCI task marked "ActionNeeded" but I can't take any action on it. Can you do something about it? Thanks 15:46
aloha OK. I'll deliver the message.
15:46 wesjdj left
kid51 whiteknight Good morning 15:46
15:48 wesjdj joined
Coke yawns. 15:51
tadzik don't do thi... YAWN
15:53 bluescreen joined 16:01 whiteknight left
wesjdj moritz: How would you translate 'Code Browser and Cross Reference tool? 16:02
moritz no idea 16:03
wesjdj moritz: Ok, I'll have a look on leo.. 16:04
16:12 cognominal left 16:13 nwellnhof left 16:15 fperrad left
dalek rrot/tt532_headerizer_refactor: 2377b3b | jkeenan++ | t/tools/dev/headerizer/02_methods.t:
Add tests for 'macro_match' option.
 16:32
16:34 cognominal joined
dalek rrot/tt532_headerizer_refactor: 7f94d34 | jkeenan++ | tools/dev/headerizer.pl:
Provide better documentation for '--macro' option.
 16:38
mikehh ran rakudo stresstest (after spectest-smolder) got a couple failures (different ones) mostly numeric though seg fault after tests (probably trying to free something already freed) 16:54
onr backtrace that I got was trying to free an invalid pointer 16:58
one
17:03 kid51 left 17:07 fperrad joined 17:14 nwellnhof joined 17:15 TiMBuS left, TiMBuS_ joined 17:29 fbrito joined, mikehh left 17:32 diakopter left, mikehh joined
fbrito Andy: hello. you are Andrew Whitworth, right? 17:32
tadzik fbrito: look for whiteknight 17:33
seen whiteknight?
aloha whiteknight was last seen in #parrot 1 hours 46 mins ago saying "msg dukeleto I have a GCI task marked "ActionNeeded" but I can't take any action on it. Can you do something about it? Thanks".
fbrito oh, my bad :D
tadzik: thank you for the information :) 17:34
tadzik fbrito: you're welcome
Coke seen jnthn ? 17:56
aloha Sorry, I haven't seen jnthn .
cotto ~~
Coke seen jnthn?
aloha jnthn was last seen in #perl6 55 mins 45 seconds ago saying "s".
Coke msg bacek "seen jnthn ?" and "seen jnthn?" should probably do the same thing. 17:57
aloha OK. I'll deliver the message.
jnthn Coke: Looking for me?
18:17 theory joined
mj41 Hi. Coverage reports cv.perl6.cz/ ( tapir2.ro.vutbr.cz/cover/ ) are back. 18:40
fbrito quick question: does PIR has the += operator? 18:42
ok, found it :D 18:45
19:53 nwellnhof left 19:54 allison joined
allison I need confirmation of a rakudo bug fix for debian. What's the best way to ask? 19:56
IRC, ticket queue?
tadzik what kind of bug? 19:57
allison tadzik: bugs.debian.org/cgi-bin/bugreport.cgi?bug=592277 19:58
earlier versions of Rakudo depended on having the SVN revision for Parrot, even if it was an installed/packaged version
It's almost certain it has been fixed (since Parrot doesn't even use SVN anymore) 19:59
but I need confirmation before I close the bug
tadzik allison: this must've been git versions of Rakudo 20:10
and specific ones
some just did have "2.8.0" in build/PARROT_REVISION
allison tadzik: he does specifically say in the bug report that he was building a git checkout of Rakudo 20:16
tadzik: but, it should be possible to build a git checkout of rakudo from an installed version of Parrot, shouldn't it?
tadzik allison: then it's obvious 2.6.0 isn't sufficient, if a git version required some specific svn revision
allison will this work for 2.9.1? 20:17
tadzik allison: only if git version needs this installed version of Parrot. There is build/PARROT_REVISION in rakudo tree which specifies which version of Parrot is needed. Releases usually have sth like 2.10.0 or something similar, git checkout in the middle of the month usually test some bleeding-edge Parrot features
allison: depends 20:18
allison tadzik: I'm building the debian 2.9.1 packages for parrot today
moritz allison: rakudo 2010-11 builds from an installed parrot
specifically from 2.9.1
tadzik what moritz said
moritz but random git revisions of rakudo usually require some parrot development versions
allison okay, I've talked through this with pmichaud in figuring out the Debian release cycle for Rakudo 20:19
I think this just means we need to get the Debian packages up for Rakudo releases
I'm happy to help with that 20:20
moritz sounds good
allison I think we've lost our Debian Rakudo packager
tadzik CRUX updates Rakudo & Parrot simultanously, monthly
allison the current version of Rakudo on debian is [rakudo_0.1~2010.01.orig.tar.gz], which I'm assuming is considered *really* old 20:21
tadzik *cough*
you are *really* right
allison tadzik: which version of Rakudo builds on Parrot 2.9.1? 20:24
moritz 2010.11
tadzik allison: ā†‘ 20:25
allison tadzik: okay, will package that one
tadzik hmm, wasn't there a version using 2.10.1?
isn't 2010.10 using 2.9.1? 20:26
moritz wait
tadzik 2010.11 is the newest, and using 2.10.1 IIRC
moritz what's the latest parrot release?
tadzik 2.10.1
moritz sorry, then tadzik is right
tadzik allison: they you want either Parrot 2.10.1, or Rakudo 2010.10
sorry for misguiding you
allison tadzik: no worries. Debian only packages supported version of parrot, so it has to be 2.9.1 20:28
tadzik: so, will have to package Rakudo 2010.10
tadzik alright
allison tadzik: are there any known showstopping bugs from 2010.10 that were fixed in 2010.11?
tadzik/moritz: I could backport limited changes to the source into the package, if needed 20:29
tadzik Well, there are bugs in every release. But nothing more than usual in 2010.10 I think
allison tadzik: okay, that's simple enough
20:41 perlite_ joined 20:44 perlite left 20:45 perlite_ is now known as perlite 20:57 wesjdj left
fbrito how to use 2 dimension arrays with NQP? i can only use 1 dimension (with $var[x]). I have tried $var[x][y] and $var[x;y] but no success 20:58
tadzik I think you'll need to play with Opcodes directly 20:59
sorear I think you'll need to create the second dimension manually
tadzik create an ResizablePMCArray full of ResizableIntegerArray, or something
moritz fbrito: in what did you not have success? 21:00
fbrito actually I am trying to use Parrot-Linear-Algebra
after creating a matrix with NumMatrix2D.new() i was able to get the elements using $matrix.item_at(x,y) 21:02
but how can I modify the value of an element? i tried $matrix.item_at(x, y) := 1, but didn't work. I could modify elements on the first row using $matrix[0] or $matrix[1] = 1, for example 21:04
I could only modify elements on the first row*
I am probably missing something silly :P. this is my first day with parrot, NQP, PIR, Perl and everything :D 21:05
sorear I guess $matrix.set_item_at or something of that nature 21:06
Perl is new to you, eh? What's your background? 21:07
fbrito Ruby
I am just a teenage trying to do some tasks on Google Code In :D 21:08
Ah, found how to solve my problem 21:15
21:20 rurban_ joined 21:22 rurban left, rurban_ is now known as rurban 21:26 dngor_ is now known as dngor 21:30 fperrad left 21:48 Kulag left, Kulag joined 21:57 gbacon joined 22:16 kurahaupo left 22:28 bluescreen left 22:34 TypeNameHere_____ left 22:42 bluescreen joined 22:45 bacek left 22:57 bacek joined
dalek rrot: 4e441aa | allison++ | ports/debian/ (2 files):
Updating Debian packaging files for 2.9.1 supported release.
 23:03
23:37 cotto left 23:42 bacek left 23:53 TypeNameHere_____ joined

Powered by Cro and the RakuĀ® Programming Language.

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

Thank you!