Camelia, the Perl 6 bug

IRC log for #parrot, 2008-08-04

Parrot | source cross referenced

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

All times shown according to UTC.

Time Nick Message
00:09 AndyA joined #parrot
00:13 dalek r29978 | jkeenan++ | scriptconfigure:
00:13 dalek : [configure] To prepare for scripted configuration, move some code that is
00:13 dalek : specific to handling of command-line options into
00:13 dalek : Parrot::Configure::Options::Conf::CLI.  Adjust as needed in
00:13 dalek : lib/Parrot/Configure/Options.pm, lib/Parrot/Configure/Options/Conf.pm and
00:13 dalek : t/configure/001-options.t.
00:13 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29978
00:15 dalek r29979 | jkeenan++ | scriptconfigure:
00:15 dalek : Delete commented-out code.
00:15 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29979
00:19 dalek r29980 | jkeenan++ | scriptconfigure:
00:19 dalek : Delete commented-out code.
00:19 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29980
00:30 dalek r29981 | Whiteknight++ | gsoc_pdd09:
00:30 dalek : [gsoc_pdd09] a few minor changes. We aren't hanging anymore, we're segfaulting now
00:30 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29981
00:35 AndyA joined #parrot
00:36 teknomunk__ joined #parrot
00:36 Whiteknight plz to hav no segfaults?
00:39 jonathan Whiteknight: Keep plugging away, you'll get there!
00:39 jonathan You're doing a job I certainly don't envy.
00:44 bacek joined #parrot
00:48 Whiteknight thanks, I can use all the encouragement I can get
00:48 Whiteknight I wish I was doing cool flashy high-level work like you
00:48 AndyA joined #parrot
00:50 ruoso joined #parrot
00:54 Whiteknight actually, I take that back. I would be terrible at it
01:01 jonathan Whiteknight: You're doing stuff that enables some of the higher level stuff to happen. :-)
01:01 MeGaMiC joined #parrot
01:01 jonathan It's all important work - no use having a compiler without a VM that can run the code it generates reliably and, preferably, fast.
01:02 Whiteknight well, at the moment I don't have "reliably" or "fast" implemented
01:03 Whiteknight those are pending more free tuits :)
01:05 jonathan :-)
01:05 jonathan Well, sleep time for me...it's getting late here in Europe. :-)
01:06 jonathan night
01:07 Whiteknight goodnight
01:11 Infinoid Whiteknight: even a step in the right direction is very much appreciated.  :)
01:19 dalek r29982 | Whiteknight++ | gsoc_pdd09:
01:19 dalek : [gsoc_pdd09] todo a test which isn't currently passing, and may not ever pass. ever.
01:19 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29982
01:56 dalek r29983 | Whiteknight++ | gsoc_pdd09:
01:56 dalek : [gsoc_pdd09] a codingstd trailing whitespace fix, since I can't find anything substantial to fix.
01:56 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29983
01:57 Andy joined #parrot
02:08 teknomunk_ joined #parrot
02:11 rafl joined #parrot
02:18 kid51 joined #parrot
02:18 dalek r29984 | jkeenan++ | scriptconfigure:
02:18 dalek : [configure] First pass at file-based configuration.  Code to handle that
02:18 dalek : approach placed in Parrot::Configure::Options::Conf::File. 'perl Configure.pl
02:18 dalek : --script=foobar' succeeds, but plays havoc with configuration and step tests
02:18 dalek : for reasons still to be explored.
02:18 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29984
02:38 dalek r29985 | jkeenan++ | scriptconfigure:
02:38 dalek : [configure] Fixed way in which lib/Parrot/Configure/Options.pm receives data
02:38 dalek : from Options/Conf/File.pm and then returns that to Configure.pl.
02:39 dalek : t/configure/033 and 044 still failing; all other configure and step tests
02:39 dalek : passing.
02:39 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29985
02:54 dalek r29986 | jkeenan++ | scriptconfigure:
02:54 dalek : Was testing non-existence of file named 'foobar' -- but Parrot developers may very well have such a file lying around.  Generate file name from process ID instead.
02:54 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29986
02:54 dalek r29987 | jkeenan++ | trunk:
02:54 dalek : Was testing non-existence of file named 'foobar' -- but Parrot developers may very well have such a file lying around.  Generate file name from process ID instead.
02:54 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29987
02:56 dalek r29988 | jkeenan++ | trunk:
02:56 dalek : Was testing non-existence of file named 'foobar' -- but Parrot developers may very well have such a file lying around.  Generate file name from process ID instead.
02:56 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29988
02:56 dalek r29989 | jkeenan++ | scriptconfigure:
02:56 dalek : Was testing non-existence of file named 'foobar' -- but Parrot developers may very well have such a file lying around.  Generate file name from process ID instead.
02:56 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29989
02:59 dalek r29990 | jkeenan++ | scriptconfigure:
02:59 dalek : Delete commented-out code and superseded Dumpers.
02:59 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29990
03:35 teknomunk__ joined #parrot
04:20 bacek joined #parrot
04:38 uniejo joined #parrot
04:48 davidfetter joined #parrot
05:36 dalek r29991 | chromatic++ | trunk:
05:36 dalek : [include] Made PMC_IS_NULL into a macro when used in Parrot's core.  The
05:36 dalek : function calling overhead is statistically significant (around 3.5% on the
05:36 dalek : Rakudo benchmark).
05:36 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29991
05:42 Psyche^ joined #parrot
05:52 dalek r29992 | chromatic++ | trunk:
05:52 dalek : [src] Promoted more constant strings to CONST_STRINGs.
05:52 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29992
05:56 dalek r29993 | pmichaud++ | trunk:
05:56 dalek : [rakudo]: spectest-progress.csv update: 116 files, 2003 passing, 12 failing
05:56 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29993
06:17 uniejo joined #parrot
06:23 dalek r29994 | pmichaud++ | trunk:
06:23 dalek : [rakudo]: Add S29-str/substr.t (+20 passing tests) to spectest_regression
06:23 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29994
06:56 Ademan joined #parrot
06:57 masak joined #parrot
06:58 MeGaMiC joined #parrot
07:05 cotto_home anyone with some basic C- and Perl-fu mind doing a quick patch review?
07:05 cotto_home http://rt.perl.org/rt3/Tic​ket/Display.html?id=57468
07:21 Ademan joined #parrot
07:29 dalek r29995 | fperrad++ | trunk:
07:29 dalek : [Lua] fix assignlist
07:29 dalek : - patch from Robert G. Jakabosky
07:29 dalek : - see http://rt.perl.org/rt3/Publi​c/Bug/Display.html?id=57568
07:29 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29995
07:42 dalek r29996 | moritz++ | trunk:
07:42 dalek : [rakudo] autounfudge now rewrites skip as todo tests if started with --unskip
07:42 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29996
07:45 Theory joined #parrot
07:53 AndyA joined #parrot
07:56 dalek r29997 | fperrad++ | trunk:
07:56 dalek : [Lua] fix environment caching
07:56 dalek : - patch from Robert G. Jakabosky
07:56 dalek : - see http://rt.perl.org/rt3/Publi​c/Bug/Display.html?id=57568
07:56 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29997
08:18 dalek r29998 | fperrad++ | trunk:
08:18 dalek : [Lua] fix Lua bytecode loader
08:18 dalek : - patch from Robert G. Jakabosky
08:19 dalek : - see http://rt.perl.org/rt3/Publi​c/Bug/Display.html?id=57568
08:19 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29998
08:32 neopallium joined #parrot
08:43 justin joined #parrot
09:13 moritz who gets commit 30k?
09:13 dalek r29999 | moritz++ | trunk:
09:13 dalek : [rakudo] assignment tests in spectest_regression
09:13 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29999
09:19 masak moritz: you could always fix a typo I found in languages/perl6/ROADMAP
09:19 masak on line 44, 'co-ercion' should be 'coercion'
09:26 dalek r30000 | moritz++ | trunk:
09:26 dalek : [rakudo] typo in ROADMAP spotted by masak++
09:26 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=30000
09:26 masak wai~! 30k!
09:47 ambs joined #parrot
10:18 dalek r30001 | fperrad++ | trunk:
10:18 dalek : [Lua] Perl/yacc compiler
10:18 dalek : fix assignlist (see r29995)
10:18 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=30001
10:22 * jonathan feels proud of contributing to r30000
10:22 jonathan ...by making questionable use of the hyphen and thus prompting the patch.
10:41 Whiteknight joined #parrot
10:58 kid51 joined #parrot
11:06 dalek r30002 | jkeenan++ | scriptconfigure:
11:06 dalek : Fix lack of accounting for 'reconfigure' mode.
11:06 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=30002
11:10 ruoso joined #parrot
11:28 Casan joined #parrot
11:33 dalek r30003 | fperrad++ | trunk:
11:33 dalek : [Lua] bytecode translation
11:33 dalek : - add a test
11:33 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=30003
11:44 dalek r30004 | fperrad++ | trunk:
11:44 dalek : [Lua]
11:44 dalek : - show a fornum issue
11:44 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=30004
11:46 dalek r30005 | fperrad++ | trunk:
11:46 dalek : [Lua] fix fornum
11:46 dalek : - patch from Robert G. Jakabosky
11:46 dalek : - see http://rt.perl.org/rt3/Publi​c/Bug/Display.html?id=57568
11:46 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=30005
11:46 UltraDM joined #parrot
11:54 leo joined #parrot
11:58 MeGaMiC joined #parrot
12:09 MeGaMiC joined #parrot
13:39 gryphon__ joined #parrot
14:26 nopaste joined #parrot
15:09 Infinoid http://news.perlfoundation.org/2008/0​8/2008q3_grant_proposal_tcltk_ac.html is interesting - a proposal to add a Tk module for rakudo
15:09 moritz yes
15:09 Infinoid if it were available for other HLLs, we might support actual tcl/tk *grin*
15:10 moritz I thought "what about partcl?", but I think that TPF might not be that interested ;)
15:10 Infinoid yeah :)
15:10 Infinoid the thing that caught my eye was: his "deliverable" section mentions a CPAN module
15:11 Infinoid how far away are we from being able to release external rakudo modules on CPAN?
15:11 * moritz feels quite far away from that :/
15:12 Infinoid if it is a pure perl implementation, it might not be so bad
15:12 Infinoid is the NCI stuff available to rakudo perl scripts?
15:19 particle no, no nci yet
15:19 particle and rakudo's 'use' is a kludge
15:22 Infinoid hmm.  I think the guy who proposed that grant is in for a rough ride...
15:22 Andy joined #parrot
15:32 particle it's not a detailed grant.
15:32 particle proposal
15:32 purl i heard proposal was like the first few minutes of a job interview.  If you come in scabbily dressed, scratching my balls and slurring, we're not going to hire you.
15:32 particle purl! she has balls!
15:32 ambs :-S
15:33 ambs purl is a shemale?
15:33 cognominal fuck you purl, you don't judge a programmer by his dress and I scratch my ball and fart when want
15:33 cognominal s/want/I want/
15:36 * tetragon pictures a farting, ball scratching, kilted programmer and decides that she would rather that that programmer fart outdoors
15:38 cognominal that could be an argument for telecommuting..
15:38 cognominal my programs are clean, I don't need to be :)
15:39 tetragon Just so long as there isn't anyone in your vicinity.
15:44 cognominal I was joking.
15:58 Whiteknight joined #parrot
16:08 pmichaud Aloha
16:08 moritz hi
16:09 moritz (one last time online before my vacations ;)
16:09 moritz we made the 2000 passing test mark. YaY ;-)
16:17 iblechbot joined #parrot
16:18 pmichaud yes.  We're actually at 2022 or something like that.
16:19 moritz yes
16:20 moritz I enhanced autounfudge to s/skip/todo/ some tests and then I unfudged some todo fudges by hand
16:20 moritz which explains the leap from 2003 to 2022
16:25 pmichaud oh.  I added 20+ tests in substr.t, so we may be above that then.
16:31 cjfields joined #parrot
16:32 moritz and we have 9 failures due to 'if $!' not working
16:32 moritz (I don't know how much my unfudged tests actually contribute, though)
16:34 cognominal is there a C specialist around here?  I am writing an article about dtrace, and I want to say that main() is not really the first fonction entered when executing a process. But I would like to be more specific.
16:34 cognominal a pointer or an URL would be welcome
16:34 pmichaud 2102 passing tests as of r30005
16:35 cognominal and yes, I write this article as a way to gain knowledge to add probes in parrot.
16:35 moritz cognominal: istr that _entry was called first, but I might be 100% off
16:36 cognominal I know that one can do weird stuff with env vars like PRELOAD or whatever...
16:36 tetragon The traces I'm looking at place start before main
16:36 cognominal tetragon, I don't get what you said. I am French and dense :)
16:37 tetragon I have a load of stack traces open from the many crashes I'm getting.  The ones with a good trace have start as the first mentioned function
16:38 cognominal you are running Solaris or Mac OS X?
16:38 tetragon OS X
16:38 tetragon I had started looking at DTrace probes before my iBook went away
16:39 cognominal I hope my macbook will live one more year or so. There will probably be quadricore by then.
16:40 cognominal There is also a linux distro with linux (meaning the kernel replaced by opensolaris).
16:40 tetragon My MacBook Pro is doing well in its first week
16:40 cognominal s/with/without/
16:41 tetragon So many crashes when using a 64-bit build of parrot, though
16:41 cognominal the plastic of my macbook is getting porous. my macbook is becoming a trashcan
16:42 cotto_home cognominal, Nexenta uses the OpenSolaris kernel and the Debian userspace
16:42 cognominal I guess that with the pid provider ones already get interesting stuff with Parrot.
16:42 cognominal cotto_home, thx, I was searching the name...
16:42 tetragon My (almost three years old) iBook wasn't porous, the techs just forgot how to reassemble the things without breaking them
16:42 cotto_home no problem
16:43 cotto_home work &
16:45 cjfields_ joined #parrot
16:49 toddr joined #parrot
16:49 toddr left #parrot
17:05 dalek r30006 | julianalbo++ | trunk:
17:05 dalek : fix parrot_debugger after pdd25cx merge
17:05 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=30006
17:06 dalek r30007 | fperrad++ | trunk:
17:06 dalek : [Pipp] fix pdd25cx merge
17:06 dalek : - move to Parrot_ex_throw_from_c_args
17:06 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=30007
17:28 dalek r30008 | Whiteknight++ | gsoc_pdd09:
17:28 dalek : [gsoc_pdd09] update to trunk r30005
17:28 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=30008
18:04 gryphon__ joined #parrot
18:30 Ivatar joined #parrot
18:33 tetragon joined #parrot
18:37 jhorwitz joined #parrot
18:50 dalek r30009 | fperrad++ | trunk:
18:50 dalek : [Lua] fix bisect output
18:50 dalek : - patch from Robert G. Jakabosky with improvements
18:50 dalek : - see http://rt.perl.org/rt3/Publi​c/Bug/Display.html?id=57568
18:50 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=30009
19:14 dalek r30010 | fperrad++ | trunk:
19:14 dalek : svn:ignore parrot_debugger
19:14 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=30010
19:19 tetragon left #parrot
19:19 tetragon joined #parrot
20:17 rdice joined #parrot
20:19 contingencyplan joined #parrot
20:26 Ademan joined #parrot
20:30 Whiteknight joined #parrot
20:31 peepsalot joined #parrot
21:32 Andy left #parrot
21:50 Limbic_Region joined #parrot
22:11 dalek r30011 | Whiteknight++ | trunk:
22:11 dalek : [docs/pdd] update pdd09 to include more descriptions, more information and some much-needed clarity. These are all lessons i've learned the hard way.
22:11 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=30011
22:15 Whiteknight anybody here now?
22:17 Whiteknight oh nevermind. "make realclean" solves all my problems, once again
22:21 Infinoid my house needs a make realclean
22:23 particle my browser has decided that the parrotvm links no longer need css
22:24 Infinoid I've had occasional issues with browsers caching a temporary lack of css, when it failed to fetch it or something... clearing the browser cache fixed that
22:25 Whiteknight browsers, you give them an inch and they take a mile
22:26 ruoso joined #parrot
22:38 gryphon__ joined #parrot
22:56 dalek r30012 | coke++ | trunk:
22:56 dalek : Followup to r30010; no need to ignore pdb anymore now that it's spelled parrot_debugger
22:56 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=30012
22:58 dalek r30013 | coke++ | trunk:
22:58 dalek : [docs] Patch from RT #57602, Courtesy Bob Wilkinson <bob at fourtheye.org>
22:58 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=30013
23:00 kid51 joined #parrot
23:01 cotto_work kid51, ping
23:02 kid51 cotto_work pong
23:02 cotto_work do you mind doing a quick review of the patch in http://rt.perl.org/rt3/Tic​ket/Display.html?id=57468
23:03 cotto_work it seems to be in your area of expertise
23:07 kid51 First reaction:  Could you add a bit to the POD that distinguishes vsnprintf from snprintf?  I have no clue as to what the difference is.
23:08 kid51 Or perhaps even before that:  Coke asks whether we should be keeping this particular conditional.  Should we?  What do we gain by incurring the expense of an additional config step?  (These questions should be answered in RT.)
23:10 kid51 We would also need a t/steps/auto_vsnprintf-01.t.
23:11 cotto_work Thanks.  I'm glad I asked.
23:12 kid51 Generally speaking, we have no pollicy to evaluate whether a new config step should be added or not.  People just do it -- often without even posting an RT!  So I'm glad that this is already on RT.
23:14 kid51 At YAPC, had conversation with particle and chromatic about configuration.   particle promises development of a more formal plan for evolution of Parrot configuration -- but not until many other things have been planned and implemented.
23:14 kid51 So right now we have no formal criteria by which to evaluate whether a particular step should be added.  All I can do is recommend that you make your case in that RT.
23:15 cotto_work I don't care too much either way.  I just want to get that ugly BLAH... macro replaced with something reasonable.
23:16 teknomunk joined #parrot
23:18 kid51 The C functions/macros are definitely *not* my area of expertise.
23:19 kid51 Parrot_secret_snprintf ????
23:19 Whiteknight shhh!!! it
23:19 Whiteknight s a secret
23:21 cotto_work looks like a workaround for when a system doesn't have C99 snpritnf.
23:21 cotto_work I wonder how common that is.
23:23 DietCoke joined #parrot
23:24 DietCoke I also have no idea why this macro was added, which is why I posted the RT in the first place.
23:24 DietCoke Honestly, a simpler patch is to simply remove it.
23:25 DietCoke since no one has spoken up in the several days since I posted it, I'd go ahead and rip it out, cotto.
23:25 cotto_work the question is how often will we need Parrot_secret_snprintf?
23:25 cotto_work too late.
23:25 particle cotto_work: it's very common on non c99 systems :)
23:26 cotto_work Andy would be good to ask, since he seems to have added it.
23:28 particle i believe vsnprintf is safer
23:28 cotto_work yes, as are all *n*printf functions
23:29 particle and v means take a va_list instead of varargs
23:31 particle so, seems to me the idea was to convert all usage of vsprintf to vsnprintf
23:31 cognominal joined #parrot
23:31 particle i agree with that idea
23:31 particle but, the implementation has possibly fallen short
23:31 particle likely got ENOTUITS or some other blocking signal
23:33 cotto_work particle, afaict the only use of vsprintf is in Parrot_secret_snprintf
23:35 dalek r30014 | coke++ | trunk:
23:35 dalek : [docs] internal_exception is gone.
23:35 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=30014
23:37 particle ask andy about the intent, this was his baby and i don't know how well he's following list traffic lately
23:37 * particle runs over to feed the neighbor's dog
23:39 cotto_work seen petdance
23:39 purl petdance was last seen on #parrot 29 days, 1 hours, 3 minutes and 30 seconds ago, saying: WHOO LOOK I EDITED THE WIKI PAGE WHOOO  [Jul  6 22:35:41 2008]
23:39 cotto_work seen Andy
23:39 purl Andy was last seen on #parrot 9 days, 2 hours, 21 minutes and 49 seconds ago, saying: (which I don't)  [Jul 26 21:17:31 2008]
23:40 kid51 cotto_work:  Recommend posting to RT, bcc-ing Andy.  If he doesn't respond "soon", use your best judgment.
23:41 cotto_work sounds like a plan
23:56 particle yep, feel free to delete after reasonable time

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

Parrot | source cross referenced