Camelia, the Perl 6 bug

IRC log for #parrot, 2008-03-07

Parrot | source cross referenced

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

All times shown according to UTC.

Time Nick Message
00:39 davidfetter joined #parrot
01:49 Psyche^ joined #parrot
02:06 dalek joined #parrot
02:25 svnbotl r26260 | coke++ | pdd17pmc:
02:25 svnbotl : [APL]
02:25 svnbotl : update pmc to pdd17 standards (I think).
02:25 svnbotl : Now only failing 6 sub-tests.
02:25 svnbotl diff: http://parrotvm.org/svn/parrot/revision/?rev=26260
02:34 IllvilJa joined #parrot
02:48 Andy HI EVERYONE LET'S WRITE SOME CODE
02:57 gryphon joined #parrot
02:58 Coke Andy: I have a double free. you want to track it down?
02:58 Andy nah
02:58 Andy I mena, I guess I could.
02:58 Coke I just don't think I have the ability to do so. =-)
02:59 Coke bah. I have a slightly modified checkout. I'll have to clean it up to see if it still happens.
02:59 Andy That's the first thing to check, yes. :-)
03:00 Coke or I could just check it in. =-)
03:00 Coke not tonigth.
03:00 Coke -> zzzz
03:08 svnbotl r26261 | chromatic++ | pdd17pmc:
03:08 svnbotl : [lib] Prefixed all C variables created in PMC C files by the PCCMETHOD
03:08 svnbotl : generator.  Now there's less chance of accidentally shadowing a variable in a METHOD in a PMC by choosing the wrong variable name.
03:08 svnbotl : I used one underscore for now.  If this needs to change in the future, we
03:08 svnbotl : should consider instead calling a couple of functions to perform the setup and
03:08 svnbotl : teardown appropriately rather than continuing the arms race required to slap
03:08 svnbotl : boilerplate code in PMCs.
03:08 svnbotl diff: http://parrotvm.org/svn/parrot/revision/?rev=26261
04:10 ewilhelm anybody know what's the story with the branches/gmc ?
04:11 ewilhelm it looks like that was a SoC2005 project -- was it successful, did it get merged or used in some other way?
04:17 cotto there isn't any gc_gmc.c in the current trunk
04:38 petdance joined #parrot
04:39 Ademan joined #parrot
04:43 diakopter joined #parrot
04:45 diakopter joined #parrot
05:11 cotto wow
05:12 cotto I just spent several days puzzling over a bug
05:13 cotto it turns out I forgot to return a value from add(), and gcc had been giving me a warning about it the whole time.
05:13 cotto compiler warnings FTW
05:29 particle joined #parrot
05:34 particle joined #parrot
05:35 petdance indeed.
06:06 Coke joined #parrot
06:19 cotto andy, do you know much about the code in src/pmc/hash.pmc and src/hash.c?
06:19 petdance no
06:19 petdance but what about 'em.
06:20 cotto I'm curious why hash_freeze and hash_thaw aren't referred to anywhere else in trunk
06:21 cotto especially when hash_visit is referred to in src/pmc/hash.pmc
06:26 cotto I'd expect Parrot_default_freeze to do something generic with the visit_info, which then would then call Parrot_Hash_visit => hash_visit => hash_freeze, but Parrot_default_freeze is empty.
06:31 cotto hmmm.  There don't seem to be any freeze/thaw tests for Hash.  Perhaps I shouldn't be assuming that it works.
06:32 cotto nm.  all the tests are in freeze.t
06:34 petdance back
06:35 petdance Yeah, I dunno.
06:47 cotto hmmm.  t/pmc/hash.t contains two copies of the same test
06:47 cotto at least they both pass
07:11 uniejo joined #parrot
08:26 iblechbot joined #parrot
08:29 AndyA joined #parrot
08:59 UltraDM joined #parrot
10:04 allison joined #parrot
10:28 wknight8111 joined #parrot
10:39 ruoso joined #parrot
11:20 wknight8111 joined #parrot
11:48 kid51 joined #parrot
12:03 rdice joined #parrot
12:18 arcady joined #parrot
12:28 cognominal joined #parrot
13:39 * Coke wonders if it would be worth having a 'make depend' target.
13:39 confound joined #parrot
13:42 gryphon joined #parrot
13:49 AndyA joined #parrot
13:55 cognominal joined #parrot
14:00 wknight8111 joined #parrot
14:28 IllvilJa joined #parrot
14:42 petdance joined #parrot
14:53 pmichaud (from scrollback)  branches/gmc is RT#47888
14:55 mj41_ joined #parrot
15:57 Coke pmichaud: have you used google docs presentation editor?
15:57 particle ooh, good idea, coke
15:58 particle i should put my outline for 'contributing to rakudo' up there and share it with pmichaud
15:58 Coke I may have to do a presentation for yapc::na just so I can use it. =-)
15:58 Coke even has "export to PDF".
16:01 purl joined #parrot
16:11 c9s joined #parrot
16:11 Coke (though save as HTML would have been spiffier.)
16:12 particle well, created the presentation, shared with pmichaud. coke++ for the idea
16:12 Coke "wah, share it with me too!"
16:14 avar Coke: I was reasonable happy with the stuff I used to generate http://search.cpan.org/src/AVAR/Talk-No​thingIsFaster510-0.02/slides/index.html
16:15 shorten avar's url is at http://xrl.us/bha7x
16:15 avar It's on CPAN and you can write your talk in POD
16:15 avar (Pod::S5)
16:15 Coke if I was goign that route, I'd use spork.
16:15 avar What's that?
16:15 purl well, that is what he is using :)
16:16 * Coke pats purl onna head.
16:16 Coke spork ishttp://search.cpan.org/~ingy/Spork-0.20/spork
16:16 Coke spork is http://search.cpan.org/~ingy/Spork-0.20/spork
16:16 Coke uses kwiki as a base instead of pod.
16:16 Coke but for collaboration google docs beats spork+svn.
16:17 peeps[work] joined #parrot
16:17 avar Ah, I didn't know about spork
16:18 avar ::S5 has syntax highlight though which is nice:)
16:19 particle i like s5
16:19 particle but it's not perfect
16:20 avar No, I had some troubles, submitted some patches to it even:)
16:21 cosimo_ joined #parrot
16:29 sjansen joined #parrot
16:51 davidfetter joined #parrot
16:55 cjfields joined #parrot
17:02 pmichaud I've tended to use spork for my presentations.  Quick, easy, and I can use 'vi' to create my slides :-)
17:02 pmichaud one of my wiki authors has embedded S5 into PmWiki, and he uses it for all of his presentations, though
17:03 avar what does spork have over s5?
17:03 avar do you have some presentations made with spork online?
17:03 pmichaud as far as presentation quality goes, s5 is likely better
17:03 pmichaud last time I looked at s5, though, one had to basically write HTML in order to use it
17:04 pmichaud my spork presentations
17:04 pmichaud http://www.pmichaud.com/2007/pres
17:04 pmichaud http://www.pmichaud.com/2008/pres
17:04 pmichaud the nice thing about spork is that it's plain text (wiki markup) and not html
17:05 pmichaud the bad thing about spork is that its wiki markup isn't very robust
17:05 pmichaud particle:  are you planning to go to yapc::na?
17:08 particle pmichaud: i'm in the process of deciding
17:08 particle i can probably do sat-wed
17:09 particle i can't stay afterwards, have to head to PA for a wedding
17:09 particle you?
17:09 purl rumour has it you is ;)
17:09 pmichaud what talks are you planning to give?
17:09 pmichaud I'm definitely planning to attend
17:09 particle haven't submitted any
17:09 pmichaud I need to submit my talks
17:09 pmichaud (deadline is 03/15, yes?)
17:09 particle urk. iunno
17:10 particle i was hoping to hear about oscon by the time i submit for yapc::na
17:10 pmichaud it usually doesn't happen that way
17:10 pmichaud anyway, yapc::na submissions are due 15-mar-2008, according to the website
17:11 pmichaud hmmm, they're also doing workshop-style tracks
17:11 pmichaud I should probably ask jmcadams about those
17:12 particle good idea
17:12 purl particle: Good Idea: Drinking fresh milk from the carton. Bad Idea: Drinking fresh milk from the cow.
17:15 cognominal apparently self can't be used by itself in pir :  rakudo generates      $P57 = self  and pir barks :(
17:15 pmichaud self can be used as long as it's in a :method
17:15 pmichaud if the sub isn't a :method, then self isn't defined
17:16 pmichaud that *does* have some pretty nasty implications for nested closures, however
17:16 pmichaud er, nested blocks
17:16 cognominal indeed
17:17 pmichaud I think that argues that 'self' needs to be lexical or in the symbol table as opposed to using 'self' directly in the PIR
17:17 cognominal can someone review my previous patch . I will get more interest in fixing rakudo
17:17 pmichaud which patch?
17:20 cognominal see #51350
17:20 cognominal in rt
17:22 pmichaud should Bool::True return 'True' for .perl() ?
17:22 pmichaud seems like that might not be correct if there's a locally defined version of 'True'
17:24 cognominal it's for the Bool class
17:24 pmichaud yes
17:24 pmichaud but if I do    my $a = Bool::True;   say $a.perl();
17:25 pmichaud it outputs "True"?
17:25 pmichaud also, the 'perl' method refers to $P0, but $P0 isn't initialized
17:25 pmichaud should probably be self
17:26 pmichaud and the correct way to test a PMC for 'truthiness' is to do either
17:26 pmichaud $I0 = istrue $P0
17:26 pmichaud or
17:26 pmichaud if $P0 goto ...
17:27 pmichaud doing a simple  $I0 = $P0  does an integer conversion
17:28 cognominal ok, the Pair code is more tested. In fact jonathan code was plain wrong.
17:28 pmichaud are the various capabilities related somehow?  It might be easier to review them as separate patches
17:28 pmichaud i.e., one patch for Pairs, one for boolean .perl, one for the :a, :!a and :$a
17:30 cognominal well :a end :!a makes the path Bool unused.
17:30 cognominal ...patch
17:33 jq joined #parrot
17:34 iblechbot joined #parrot
17:36 cognominal :a , :!a and :$a are pairs btw
17:36 cognominal postcircumfixed pairs are not yet implemented
17:48 pmichaud that's not too hard to add, though -- we have postcircumfix pairs in NQP
17:50 cognominal yes. but the patch will be bigger :)
17:50 pmichaud ah, I see
17:51 pmichaud according to pugs,   say True.perl();  should output "Bool::True"
17:52 pmichaud anyway, I have to go fetch lunch -- I'll review/apply the patch in more detail as soon as I get back
17:52 cognominal thx
18:07 gryphon joined #parrot
18:11 katie_d I'm trying to walk myself through the initialization of the interpreter, and I wanted to check to make sure that I hadn't misunderstood something. As far as Parrot_set_config_hash(), is this function just loading the serialized config data from disk into a string, and then this string is not deserialized until later during init_world()? Why, then, is Parrot_set_config_hash() the first thing called, if it won't be used until so much late
18:13 Psyche^ joined #parrot
18:22 Theory joined #parrot
19:04 jq joined #parrot
19:22 davidfetter joined #parrot
19:33 x___ joined #parrot
19:47 DarkWolf84 joined #parrot
19:49 allison joined #parrot
20:14 Coke seen barney?
20:14 purl barney was last seen on #parrot 1 day and 2 hours ago, saying: Sounds like a mess, I hope that everybody involved is sane. At least they will hear about Parrot :=)  [Mar  6 09:58:06 2008]
20:21 parrot-poke joined #parrot
20:33 cotto_ anything relating to phparray?
20:53 allison joined #parrot
21:13 Dave joined #parrot
21:18 cotto_ excuse
21:18 purl cotto_: because of network lag due to too many people playing deathmatch
21:33 svnbotl r26262 | pmichaud++ | trunk:
21:33 svnbotl : [pge]:
21:33 svnbotl : * Convert various pge tests to use Perl6Regex instead of
21:33 svnbotl :   deprecated P6Regex (RT #48028).
21:33 svnbotl diff: http://parrotvm.org/svn/parrot/revision/?rev=26262
21:40 svnbotl r26263 | pmichaud++ | trunk:
21:40 svnbotl : [tcl]:
21:40 svnbotl : * Convert deprecated PGE::P6Regex to PGE::Perl6Regex (RT#48028)
21:40 svnbotl : * 'make test' produces same results before and after this change.
21:40 svnbotl diff: http://parrotvm.org/svn/parrot/revision/?rev=26263
21:43 cotto_ joined #parrot
21:47 svnbotl r26264 | pmichaud++ | trunk:
21:47 svnbotl : [m4]:
21:47 svnbotl : * Convert deprecated PGE::P6Regex to PGE::Perl6Regex (RT#48028).
21:47 svnbotl : * "make test" produces same results both before and after this change.
21:47 svnbotl diff: http://parrotvm.org/svn/parrot/revision/?rev=26264
21:47 allison joined #parrot
21:49 cotto_ what's the ETA for the pdd17pmc merge?
21:49 Tene soon
21:57 cotto_ nopaste?
21:57 purl nopaste is, like, at http://nopaste.snit.ch/ (ask TonyC for new channels) or http://rafb.net/paste or http://paste.husk.org/ or http://poundperl.pastebin.com/ or http://paste.scsys.co.uk/ or don't bother me while I'm eating
21:57 pmichaud "saturday", iirc
21:58 Coke joined #parrot
22:00 cotto_ sounds good
22:01 cotto_ would someone mind applying the patch at http://mksig.org/parrot/phparray_remove.patch to pdd17pmc to remove phparrays from that branch
22:01 cotto_ Barney already took them out of trunk, but I forgot to ask him about pdd17pmc
22:04 cotto_ details are at http://rt.perl.org/rt3/Publi​c/Bug/Display.html?id=51466
22:27 Theory joined #parrot
22:29 cotto_ no takers with commit bits?
22:42 Theory_ joined #parrot
22:44 Theory joined #parrot
23:05 petdance joined #parrot
23:20 Ademan joined #parrot
23:30 Theory joined #parrot
23:58 teknomunk joined #parrot

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

Parrot | source cross referenced