Parrot 0.9.1 Released | parrot.org/ | 451 RTs left!
Set by moderator on 24 February 2009.
00:08 AndyA joined
cotto notes that testing rakudo's build would probably be a good idea in general 00:16
00:23 Coke joined
cotto rurban, you may be able to fix the rakudo build by changing the VTABLE functions in that commit to GETATTR_Key_key_str macros. can't test atm 00:27
rurban not me. I'll go to sleep now. masak maybe, he has commits 00:28
but there's a nervous ticket by patrick in trac
00:41 HG` joined
Coke . 00:53
irclogs? 00:59
purl i think irclogs is irclog.perlgeek.de/parrot/today or see also: infrared clogs
01:02 kid51 joined 01:09 Whiteknight joined 01:35 jimmy joined
kid51 One of our MSWin32 smoke testers is reporting a failure in t/perl/Parrot_Docs.t. See: smolder.plusthree.com/app/public_pr.../18445/393 01:35
shorten kid51's url is at xrl.us/behycb
kid51 I don't have MSWin32 available. I can't reproduce this failure on Linux. And what appear to be *other* MSWin32 smoke testers are *not* reporting this failure. Is there anybody around who can test on Win32? 01:36
jimmy me
wait 01:37
kid51 The test is pure Perl, you don't even have to configure.
jimmy updating 01:38
All tests successful. Files=1, Tests=25, 2 wallclock secs ( 0.02 usr + 0.09 sys = 0.11 CPU) Result: PASS 01:39
kid51 thanks. I suspect there's some impurity on the smoke tester's box.
jimmy maybe 01:40
maybe the his perl env
kid51 But unfortunately we don't have a way of identifying the source of Smolder reports.
dalek website: coke++ | TWIP - This week in parrot, February 28th, 2009 01:44
website: www.parrot.org/news/twip-week-parro...-28th-2009
shorten dalek's url is at xrl.us/behyc4
Coke ttp://www.parrot.org/news/twip-week-parrot-february-28th-2009
www.parrot.org/news/twip-week-parro...-28th-2009
feedback welcome. 01:45
01:47 tetragon joined
kid51 Coke: Looks good; it's first time I've looked at it. 01:52
Coke: Any response to my last comment in trac.parrot.org/parrot/ticket/310 ?
01:57 HG` joined
Coke kid51: commenting now. 01:58
01:59 Andy joined
Coke kid51: done. 02:02
dalek rrot: r37067 | coke++ | trunk/DEPRECATED.pod:
Mention TT #310
 02:05
kid51 Thx
So, I can't work on it till after the April release. 02:07
02:11 bacek joined
Coke Immediately after the 1.0 release. 02:34
DEPREC* just refers to releases, not svn. 02:35
and you can always branch it up, get a patch ready, etc.
02:42 bacek joined
dalek kudo: 0d369db | pmichaud++ | perl6.pir:
Initial version of importing setting exports into global namespace.
 02:44
shorten dalek's url is at xrl.us/behyh5
nopaste "kid51" at 70.85.31.226 pasted "make testj errors at r37067" (17 lines) at nopaste.snit.ch/15752 02:56
kid51_at_dinner I believe rurban reported the same errors earlier today, but I can't locate where right now. 02:57
kid51 Ah, he reported it on list. 02:58
dalek rrot: r37068 | coke++ | trunk/DEPRECATED.pod:
Whoops; Removed a deprecation notice from a work in progress; add it back in until that commit hits.
 03:01
kid51 must sleep
purl $kid51->sleep(8 * 3600);
dalek rrot: r37069 | cotto++ | trunk/src/packout.c:
[PMC] unbreak Rakudo, fixing whatever broke it in TT #388
 03:16
cotto pmichaud, that commit should take care of the recent rakudo build breakage.
I'm glad that was a quick fix, because I'm out of brainpower. 03:26
Hmmm. I can very easily see how that could be a cow issue. 03:27
==--
kerma ==
karma ==
purl == has karma of -1
03:31 Eevee_ joined 03:34 TiMBuS joined 03:35 janus joined
jimmy cotto: it is the same (type == PFC_STRING && ct->constants[i]->u.string == s) || (type == PFC_NUMBER && ct->constants[i]->u.number ==VTABLE_get_integer(interp, key)) 03:35
cotto jimmy, can you add information to the ticket? 03:36
03:36 Theory joined
jimmy yes,怀I'm glad to upload a new patch 03:37
cotto thanks. I done for the night 03:39
03:39 contingencyplan joined
pmichaud cotto, jimmy, others: it appears that r37069 does indeed fix the problem with building rakudo. Im' running a spectest now. 03:48
04:10 buildbot joined 04:17 Eevee joined 04:35 Tene joined 05:41 Theory joined 06:04 rurban joined
rurban /msg buildbot status 06:04
07:24 uniejo joined
dalek rrot: r37070 | fperrad++ | trunk (4 files):
[chm] HtmlHelp compiler accepts only HTML file. So, all file must be htmlized.
 07:41
07:50 masak joined 08:35 slavorg joined
moritz pmichaud: if r37069 fixes a build failure, should we also bump build/PARROT_REVISION? 08:47
dalek kudo: 50279cf | (Moritz Lenz)++ | (11 files):
Merge branch 'master' of git@github.com:rakudo/rakudo
 08:51
shorten dalek's url is at xrl.us/behy5h
jimmy What's been changed? I can't see any changes. 08:52
dalek rrot: r37071 | fperrad++ | trunk/config/gen/makefiles (2 files):
[chm] add target htmlhelp
 09:29
moritz what's the minimum perl version of building parrot? 09:35
I guess some 5.8.*, but I don't know exactly which
@tell pmichaud it would be really neat to have a fixed URL that contains always the latest rakudo release - any chances of getting that soonish? 09:37
jimmy when I saw ā€˜Merge branch 'master' of git@github.com:rakudo/rakudoā€™ I don't know what's been changed? I can't see any changes. 09:38
moritz jimmy: open gitk, you'll see what's changed there 09:40
in this case it was [setting] added a stub .perl in Object.pm 09:41
jimmy thanks, I usually saw it at github online. 09:47
masak moritz: maybe you noticed, bu no lambdabot in here, so no @tell. 10:03
moritz masak: uhm yes, it's msg here... 10:04
masak kicks purl 10:11
purl NEWSFLASH: LAMER ATTACKS INFOBOT. FILM AT 11
10:25 alvar joined 10:29 jimmy left
dalek rrot: r37072 | fperrad++ | trunk (2 files):
[install] add installable_pbc_to_exe that is just a copy
 10:51
rrot: r37073 | fperrad++ | trunk/tools/dev/mk_language_shell.pl:
[install] remove some build_dir, refactor some targets 'clean'
 10:59
10:59 Ademan joined 11:22 gaz joined
cotto Can someone tell me why PackFile_find_in_const is comparing a FLOATVAL and an INTVAL? 11:33
Apparently because I changed it. 11:37
self fail 11:38
I guess that's what happens when commits end in 666. 11:42
The rest of that commit looks sane. What a strange typo. 11:43
smolder? 12:05
purl well, smolder is sourceforge.net/projects/smolder or web-based smoke test aggregator used by developers and testers to upload (automated or manually) and view smoke/regression tests using the Test Anything Protocol (TAP). or smolder.plusthree.com/app/public_pr..._reports/8
dalek rrot: r37074 | cotto++ | trunk/src/packout.c:
[PMC] fix an old uncaught mistake in PackFile_find_in_const, minor refactor
 12:06
12:24 UltraDM joined
dalek rrot: r37075 | fperrad++ | trunk/MANIFEST:
[install] put mk_inno_language.pl in devel package
 12:57
13:28 DietCoke joined 13:50 Whiteknight joined 13:54 Andy joined
dalek rrot: r37076 | fperrad++ | trunk/tools/dev/mk_inno_language.pl:
[inno] add various license file names
 13:59
14:02 gryphon joined
dalek rrot: r37077 | rurban++ | trunk (6 files):
[t] TT #357: native_pbc status updates according to smolder.

  - fill in slots for _7 and _8.
  - linebreak in pod, update copyrights		 14:15
14:25 Eevee joined 14:46 mberends joined 14:48 contingencyplan joined
Coke wonders how aweesome it would be to get the 2008 GSOC work merged into trunk before 2009. 14:49
er, GSOC2009.
14:50 cotto joined 14:53 rg joined 15:02 Andy joined 15:06 PacoLinux joined
Coke gets "make fulltest" to avoid using the deprecated options. 15:12
rg you have got to be kidding me. trac won't let me reopen my own ticket?! 15:23
Coke what is your trac account? 15:24
rg rg
this would apply to every regular user then 15:25
Coke any authenticated user, ayup. 15:26
I'm trying to figure out which of the dozen TICKET_ perms is the one that allows this.
please try again? (no clue if you have to logout/login) 15:27
rg ah yes, that worked. 15:28
now i also have options to generally modify ticket status
Coke Which is probably not what we want. :| 15:29
rg and change all properties
purl rg: that doesn't look right
Coke I'll leave it as is for now. If it gets abused, we'll reshuffle.
rg i don't think that's what you want for everyone (although i'd sure like to keep it ;))
Coke let me check something.
meh. just changed it so you have that perm, not all authenticated users. 15:30
thank you for your patches. 15:31
"Please bear with us as we adjust to our strange new infrastructure"
rg coke: thanks a lot. now i can finally add myself as a cc to the tickets i'd like to keep an eye on without having to subscribe to the whole list. 15:32
still you might want to figure out what's wrong with reopening tickets. after all, if trac is offering the option it should be possible. 15:34
maybe the problem is that it's also trying to delete "resolution: fixed" 15:35
seen rurban 15:37
purl rurban was last seen on #parrot 9 hours, 32 minutes and 30 seconds ago, saying: /msg buildbot status
15:39 Theory joined
Coke rg; worst case, you can always comment on the ticket and say "I don't think this is resolved, please reopen." 15:46
15:51 Tene joined
dalek rrot: r37078 | NotFound++ | trunk/src/pmc/coroutine.pmc:
[cage] add a cast and fix some spellings in coroutine.pmc
 15:58
rrot: r37079 | coke++ | trunk/t/pmc/packfileconstanttable.t:
TODO a test (TT #385) that is causing a LOT of smoke and blocking progress on other tickets.
 16:02
rg coke: so who would have an idea how to fix that test? 16:04
Coke If I wanted to try to track that down, I'd see who modified the test file before it started breaking, or who modified the PMC it's testing. 16:05
I am just trying to get "make fulltest" running again, which it happily fails to do with a failing test. 16:06
dalek rrot: r37080 | coke++ | trunk/lib/Parrot/Harness/Options.pm:
Properly map t/harness options to parrot options. (e.g., -b => --runcore=bounds)
 16:15
16:21 cas joined
dalek rrot: r37081 | coke++ | trunk (38 files):
RT #46503: [DEPRECATED] Remove individual runcore command line flags...

  ./parrot -b no longer works: -R bounds or --runcore=bounds is now required.		 16:36
kudo: 831c21f | pmichaud++ | docs/spectest-progress.csv:
spectest-progress.csv update: 315 files, 7087 passing, 0 failing
 16:40
shorten dalek's url is at xrl.us/behzt3
dalek kudo: 67770c3 | (Moritz Lenz)++ | docs/spectest-progress.csv:
Merge branch 'master' of git@github.com:rakudo/rakudo
 17:02
shorten dalek's url is at xrl.us/behzww
17:58 szabgab joined 18:12 Psyche^ joined
rg why does pbc_to_exe load config.fpmc when it could get the values from its own interpreter conf_hash? (or couldn't it?) 18:16
18:22 davidfetter joined 18:47 szabgab joined
dalek imitivearc: 91d25ef | (Francois Perrad)++ | (4 files):
re-run mk_language_shell.pl
 19:07
shorten dalek's url is at xrl.us/beh2jv
19:13 barney joined
Coke having to manage multiple branches at work is really making me dislike svn. 19:45
s/making/helping/ 19:46
pmichaud prepares for the inevitable 'git' chorus
Coke yah. I'd certainly consider it if the cost of switching was not so high. 19:47
Cheaper to push for shorter lived branches. 19:48
19:58 estrabd joined 19:59 mikehh joined
Coke pmichaud: when you were doing branch2, branch3, did you have a tool to do that? did you apply individual commits or just a single commit that was "everything that happened on the last copy of the branch" ? 20:05
barney Is the env var PARROT_TEST still used ? 20:08
Coke $ ack -a PARROT_TEST ? 20:10
barney only in PARROT_TEST and some tests 20:16
only in Parrot::Test and some tests
pmichaud Coke: no, I didn't have a tool -- I did it by hand. But it was easy to do by hand (more) 20:17
in the old branch: svn log --stop-on-copy would tell me the revision number where the old branch began 20:18
I'd then do: svn copy trunk branch_new
cd branch_new
svn merge -r 'old_rev_creation:head' branch_old
thazzit.
dalek kudo: 66595b0 | pmichaud++ | build/gen_metaop_pir.pl:
Add R metaops for standard infix operators.
 20:20
shorten dalek's url is at xrl.us/beh2ux
mikehh I just did a build at r37081 - smolder test passes, however it fails the build tools test after configure
Failed 1/38 test programs. 3/996 subtests failed
t/pharness/02-get_test_prog_args.t (Wstat: 768 Tests: 16 Failed: 3) 20:22
Failed tests: 2, 5, 13
smolder.plusthree.com/app/public_pr...ails/18484 20:23
shorten mikehh's url is at xrl.us/beh2vf
20:33 Khisanth joined
dalek rrot: r37082 | NotFound++ | trunk/t/op/01-parse_ops.t:
[t] fix skip on -j in 01-parse_ops.t
 20:47
20:53 justin joined 20:57 particle1 joined 20:58 mikehh joined
rg eek. the svn web interface link on parrot.org/dev still links to perl.org. could someone fix that? 21:24
GeJ Good morning everyone
21:25 bsdz joined 21:26 mikehh joined
Infinoid rg: it's on my list, I just haven't had a spare moment in the last few days 21:28
dalek 9plus: d50b283 | (Bernhard Schmalhofer)++ | (3 files):
Rerun mk_language_shell.pl
shorten dalek's url is at xrl.us/beh254
rg infinoid: ok, at least you know ;) 21:30
Infinoid That whole page needs a review 21:31
rg the page looks fine to me otherwise (aside from the fact that it could probably be moved to the wiki) 21:32
Infinoid parrotblog and parrotcode are going away (they will be CNAMEd to parrot.org if I understand correctly) 21:33
and the irc logs link needs to go to irc.perlgeek.de. otherwise it doesn't look too bad, as you said 21:34
rg oh right. i missed that. 21:35
dalek pp: 9c31985 | (Bernhard Schmalhofer)++ | lib/P (6 files):
Use Pipp::Test for testing
 21:41
shorten dalek's url is at xrl.us/beh29f
dalek pp: 9f8f981 | (Bernhard Schmalhofer)++ | (35 files):
Use Pipp::Test for testing.
shorten dalek's url is at xrl.us/beh29h
dalek zy-k: 9021e34 | (Bernhard Schmalhofer)++ | (3 files):
Rerun mk_language_shell.pl. Courtesy of fperrad.
 21:48
shorten dalek's url is at xrl.us/beh3bj
barney Argh, github hangs 21:55
PerlJam it's a sign!
purl and it says STOP
PerlJam ;)
barney is reading the signs 21:56
Coke PerlJam: I spent most of today fighting with svn. thought you would appreciate that.
PerlJam Coke: did you win?
Coke not yet. 21:57
I have nearly vanquished the daemons.
I had one branch that got delivered to trunk. I then tried to update the 8 other branches that were in various states of disrepair.
Having 'rebase' would have made it the work of about 5m. =-) 21:58
bsdz anyone here might be able to help wity
an nci question
dalek rrot: r37083 | fperrad++ | trunk/runtime/parrot/library/config.pir:
[config] refactor with FileHandler PMC & exception handling (step for TT #345)
PerlJam Coke: yep. svn is really nice ... for simple needs. As soon as you start doing anything moderately complex, it's time to consider something else. 21:59
bsdz: you could just ask. I'm not an NCI expert, but I may have the bit of knowledge that you need :) 22:00
Coke PerlJam: at the moment, my choices are svn and cvs. 22:01
rg interesting commit. just how is that going to help?
Coke so, I picked the right one, I think.
dalek : ba41e71 | (Bernhard Schmalhofer)++ | (3 files):
Rerun mk_language_shell.pl, Courtesy of fperrad
shorten dalek's url is at xrl.us/beh3ex
dalek : 3b6e8e5 | (Bernhard Schmalhofer)++ | VERSION:
A more regular VERSION
shorten dalek's url is at xrl.us/beh3e3
PerlJam Coke: oh, indeed. You suffer the tyranny-of-being-poor :-)
Coke: I take it, this is some work thing and that's why you're choice-limited? 22:02
Coke es.
yes.
local server team doesn't want to manage any version control system. we have a sourceforge-like service inside the company that provides cvs/svn hosting "for free". 22:03
bsdz thanks PerlJam. I'm trying to return data via pointers in my dll function. specifically using sig func = dlfunc lib, "foo", "ii3". My func is int foo(int x, int *y) { *y = 2*10; return x; }. Now if I foo(x,y) where x = 10 and y is a .local int I don't get the return I expect iny only -888. I did a little debugging and looks like the GC is destroying the return value before it gets to me.
Coke (which given how each department bills every other department for everything, is nice.)
dalek pp: 386010a | bschmalhofer++ | lib/Pipp/.Test.pm.swp:
Could not figure out how to delete accidently checked in file.
 22:05
confound Coke: where do you work?
shorten dalek's url is at xrl.us/beh3ga
Coke confound: I'm currently deployed at GE. 22:06
PerlJam bsdz: what types are your x and y that you're passing to foo() ? 22:07
bsdz PJ: both defined as ".local int x,y"
PerlJam bsdz: pretty sure you need a PMC for y 22:08
anything pointery needs a PMC
Coke isn't that what autoboxing is for?
bsdz PJ: should it be a managed or unmanaged struct?
Coke ah. you don't just mean any ole pmc. 22:09
PerlJam bsdz: neither. It should be an 'Integer' PMC (IIRC)
PerlJam memory is fuzzy here, so caveat lector and play with it :) 22:10
bsdz PJ: ah okay. i'll try that
dalek pp: bd13b0b | (Bernhard Schmalhofer)++ | :
Remove .Test.pm.swp
shorten dalek's url is at xrl.us/beh3hf
bsdz PJ+: that worked 22:11
PJ: thanks :) 22:13
PerlJam sure. 22:16
what are you using NCI for btw?
bsdz i'm building a library interface into lapack 22:17
PerlJam cool.
bsdz early stages at the moment but got some basic matrix ops working
PerlJam I just hope that perl6 (through parrot :) has all of the PDL goodness readily available. 22:18
22:19 donaldh joined
bsdz should be useful for most HLLs. just need to create a driver to flatten out internal reps of matrices to flatten lapack ones :) 22:20
GeJ PDL ?
purl PDL is at www.aao.gov.au/local/www/kgb/perldl/ or perl data language or at pdl.perl.org
PerlJam purl: no, PDL is pdl.perl.org 22:23
purl okay, PerlJam.
PerlJam pdl?
purl i heard pdl was pdl.perl.org
bsdz is there any easy way to do a _dumper on a ManageStruct in pir? 22:28
dalek rrot: r37084 | NotFound++ | trunk (3 files):
[examples] pir example of HLL interoperability
 22:39
NotFound I hope that this example will be more readable than the version pirric for people not exposed to old Basic ;-) 22:42
Coke PerlJam: if you pass an int to something expecting a PMC, you'll get an Integer (or the HLL equiv) 23:02
23:08 basic joined
cotto Does anyone know the PackFile code? 23:09
rg i guess rurban does 23:10
cotto seen rurban 23:11
purl rurban was last seen on #parrot 17 hours, 7 minutes and 20 seconds ago, saying: /msg buildbot status
23:18 sjn joined
cotto seen chromatic 23:26
purl chromatic was last seen on #parrot 2 days, 22 hours, 45 minutes and 8 seconds ago, saying: Removing deprecated ops will change PBC_COMPAT. [Feb 28 00:40:38 2009]
23:36 bacek_ joined
Infinoid cotto: I've used it once or twice and wrote half of the Packfile PMCs, if that helps. 23:51
cotto Could you take a look at TT #388? 23:52
23:52 kid51 joined
Infinoid Got a backtrace handy, or do I have to build rakudo to get one? 23:53
cotto no rakudo needed
see the last comment, although I can nopaste a backtrace too 23:54
Infinoid I'll try to build parrot then
nopaste "cotto" at 96.26.202.243 pasted "bt of tt #388 for Infiniod" (31 lines) at nopaste.snit.ch/15757 23:55
cotto s/io/oi/
(note that the "feather" in that nopaste is unrelated to the machine run by juerd) 23:56
Infinoid ok, thanks. Is the trunk/src/packout.c hunk of that patch valid? 23:57
The previous code was checking pointer equality, not string equality
cotto it passes make test for parrot
I thought it'd be better to check string equality than pointer equality. 23:58
Infinoid It may have unconstanted a string, though I'm not sure why that would kill imcc.
pmichaud I suspect checking pointer equality is important there.
Infinoid can dig further when he gets home
pmichaud I doubt that the original intent was string equality. 23:59
(if it was, then that code was seriously brain damaged)
Infinoid Yeah, it looks a bit suspect to me too, though I'm not really familiar with the constant handling stuff.

Powered by Cro and the RakuĀ® Programming Language.

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

Thank you!