Camelia, the Perl 6 bug

IRC log for #parrot, 2008-03-04

Parrot | source cross referenced

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

All times shown according to UTC.

Time Nick Message
00:29 svnbotl r26194 | cspencer++ | trunk:
00:29 svnbotl : Added a grep method that will parse and behave as expected when invoked in the form:
00:29 svnbotl :    "@list.grep({ $_ % 2 })" or "(1, 2, 3, 4, 5).grep({ $_ % 2 })"
00:29 svnbotl : Other forms (ie. "grep { $_ % 2 }, @list" and "@list.grep:{ $_ % 2 }" don't correctly parse the block and don't work as expected.
00:29 svnbotl diff: http://parrotvm.org/svn/parrot/revision/?rev=26194
00:41 rdice joined #parrot
00:56 mncharity joined #parrot
01:00 svnbotl r26195 | cspencer++ | trunk:
01:00 svnbotl : Added a first method that will work for the form:
01:00 svnbotl :   "@list.first({ $_ eq 3 })" or "(1, 2, 3, 4, 5).first({ $_ eq 3 })"
01:00 svnbotl : Other forms won't parse (see previous grep commit message) until blocks are parsed correctly.
01:00 svnbotl diff: http://parrotvm.org/svn/parrot/revision/?rev=26195
01:04 Theory joined #parrot
02:09 wknight8111 joined #parrot
02:13 * Coke stares blindly at a GC error.
02:20 Coke if I have a pointer to a struct PMC_EXT, can I get gdb to dump the contents of that pointer?
02:20 slightlyoff joined #parrot
02:42 Coke anyone have ubuntu that can address RT#51340
02:52 Coke is there a test target that does NOT include the configure stuff?
02:53 kid51 Don't we now have 'make coretest' ?  (or is it coretests?)
02:59 Coke ahhhh.
02:59 Coke danke.
03:04 svnbotl r26196 | coke++ | trunk:
03:04 svnbotl : RT #51136 -
03:04 svnbotl : avoid a segfault by commenting out this code: triggers memory leaks, but avoids segfaults.
03:04 svnbotl diff: http://parrotvm.org/svn/parrot/revision/?rev=26196
03:20 * Coke grumbles
03:23 * kid51 must sleep
03:23 purl $kid51->sleep(8 * 3600);
03:45 AndyA joined #parrot
04:25 svnbotl r26197 | chromatic++ | trunk:
04:25 svnbotl : [src] Fixed clone op not to share PMC_EXT structure between clone source and
04:25 svnbotl : its ultimate destination.  When the source PMC gets freed (during the next GC
04:25 svnbotl : run), the GC will free its PMC_EXT structure.  Sadly, because both structures
04:25 svnbotl : share the pointer, this means that it would be invalid in the clone
04:25 svnbotl diff: http://parrotvm.org/svn/parrot/revision/?rev=26197
05:33 grim_fandango joined #parrot
05:53 svnbotl joined #parrot
06:35 Andy joined #parrot
06:43 amoore joined #parrot
06:58 uniejo joined #parrot
07:26 svnbotl r26199 | petdance++ | trunk:
07:26 svnbotl : Noting a bug with an XXX
07:26 svnbotl diff: http://parrotvm.org/svn/parrot/revision/?rev=26199
07:45 svnbotl r26200 | chromatic++ | trunk:
07:45 svnbotl : [src] Minor cleanup and tidying.
07:45 svnbotl diff: http://parrotvm.org/svn/parrot/revision/?rev=26200
08:12 iblechbot joined #parrot
08:35 svnbotl r26201 | fperrad++ | trunk:
08:35 svnbotl : [Lua]
08:35 svnbotl : - 64bits support for bitlib
08:35 svnbotl diff: http://parrotvm.org/svn/parrot/revision/?rev=26201
08:35 hugues23 left #parrot
08:42 svnbotl r26202 | allison++ | trunk:
08:42 svnbotl : [test] Appropriately skipping the rest of the file.
08:42 svnbotl r26203 | allison++ | pdd17pmc:
08:42 svnbotl : [pdd17pmc] Cleaning up and porting new code (implementing does in objects)
08:42 svnbotl : brought in by the merge from trunk.
08:42 svnbotl diff: http://parrotvm.org/svn/parrot/revision/?rev=26203
08:42 contingencyplan joined #parrot
09:01 svnbotl r26204 | allison++ | trunk:
09:01 svnbotl : [oo] Code was moved over from src/pmc/class.pmc, but documentation was left
09:01 svnbotl : behind. Fixing.
09:01 svnbotl diff: http://parrotvm.org/svn/parrot/revision/?rev=26204
09:26 svnbotl r26205 | allison++ | pdd17pmc:
09:26 svnbotl : [pdd17pmc] Include the Class PMC's generated header file, since oo_private.h no
09:26 svnbotl : longer contains the macros for core struct access.
09:26 svnbotl diff: http://parrotvm.org/svn/parrot/revision/?rev=26205
10:00 AndyA joined #parrot
10:35 IllvilJa joined #parrot
10:53 rdice joined #parrot
11:07 Khisanth joined #parrot
11:20 slightlyoff joined #parrot
11:21 avar joined #parrot
11:40 avar joined #parrot
11:43 avar joined #parrot
12:50 svnbotl r26206 | fperrad++ | trunk:
12:50 svnbotl : [codingstd]
12:50 svnbotl : fix line length
12:50 svnbotl diff: http://parrotvm.org/svn/parrot/revision/?rev=26206
13:31 Coke joined #parrot
13:31 sjansen joined #parrot
13:48 iblechbot joined #parrot
14:20 ask_ joined #parrot
14:24 gryphon joined #parrot
14:25 rdice joined #parrot
14:46 wknight8111 joined #parrot
15:33 rblackwe joined #parrot
15:44 pmichaud I may miss #parrotsketch today ... Paula is having an all-day chemotherapy session
15:56 AndyAway Sheesh, doesn't the hospital have wifi?
15:57 rjbs left #parrot
15:57 rblackwe joined #parrot
15:58 pmichaud I'm at the hospital now, yes
15:59 pmichaud (and on its wifi)
15:59 pmichaud but at the time of parrotsketch (2.5 hrs from now), we're likely to be in the midst of the procedure
16:03 buildbot joined #parrot
16:08 Coke best of luck.
16:08 pmichaud thanks
16:09 pmichaud this is supposed to be the particularly grueling round (which we were supposed to have done on Feb-07, but illnesses and other things have delayed us to today)
16:09 particle if paula asks for more gruel, you know she's okay :)
16:09 pmichaud heh
16:10 pmichaud she brought a _ton_ of food with her today
16:13 wknight8111 it will be good if she feels like eating it
16:14 pmichaud indeed
16:14 wknight8111 the chemo can make you so nautious sometimes, it's hard to eat and keep your strength up
16:14 pmichaud correct
16:15 pmichaud after the first round (IV chemo) she didn't have too much trouble with eating/strength
16:15 pmichaud but this round is "IP chemo" (direct into the abdomen), and it's supposed to be a lot worse
16:17 wknight8111 oh yeah, it will be. Chemo is the quintessential "things get worse before they get better" drug
16:20 wknight8111 but with support and determination, they do get better
16:21 pmichaud well, Paula's a very strong person (both physically and emotionally), so we're very optimistic :-)
16:23 particle pmichaud:  if you have any ideas for rakudo/parrot SoC projects, update the wiki
16:23 particle http://www.perlfoundation.o​rg/perl5/index.cgi?gsoc2008
16:25 Theory joined #parrot
16:27 cognominal I am still reading docs and sources, but I can't still understand if (or how) one can distinguish the Perl 6 Hash from the Parrot Hash
16:27 cognominal ....so that the rakudo Hash.pir works as expected
16:32 TimToady surely if things are set up right methods should simply dispatch to the correct code?
16:33 TimToady why would you need to test?
16:34 cognominal my vtable method override is not called in Hash.Pir, that means that the compile creates Parrot Hash and not Perl6 Hash
16:35 cognominal rakudo
16:35 cognominal my %a; say %a
16:35 cognominal Hash[0x1a22c70]
16:36 cognominal someone coded get_string before me, it is  not called either
16:37 cognominal I am not proficient enough to know how to fix that problem
16:42 cognominal apparrently oo_get_class_str searches in the HLL before digging in native pmc and creating a proxy. So It should get the Perl Hash if the namespace is set
16:43 TimToady I'm just channeling something I don't understand here, but I've seen various comments that rakudo doesn't properly do HLL yet
16:43 TimToady whatever that is...
16:44 particle it's not using it's own language namespace, so pmc class names clash
16:44 cognominal ok, I will dig it further
16:44 pmichaud we need to create a Perl6Hash that subclasses Parrot's Hash
16:44 particle cognominal: $work is pretty busy now, otherwise i'd help
16:44 particle ah, look. i've been rescued :)
16:45 pmichaud the tricky part will be getting the classname mappings working properly
16:45 pmichaud then Perl6Hash can override get_string
16:48 pmichaud maybe I'll do some class conversions this morning... or at least get started on them
17:11 * Coke thinks chromatic has fixed all the GC related segfaults I'm seeing in tcl.
17:11 Coke (just the GC ones, though. =-)
17:21 autarch joined #parrot
17:30 autarch left #parrot
17:31 pmichaud afk for a while
17:56 cognominal it seems that .HLL_map pir directive may part of the solution to my problem
17:56 cognominal s/may/may be/
18:13 Coke depends on what you're trying to do, but yes. =-)
18:28 * PerlJam wanders over to spy on some sketchy characters
18:29 Coke is that now or in 60?
18:29 chromatic joined #parrot
18:30 chromatic It's #ps time
18:30 Coke <blue's clues>Open the sketch and see what's inside!
18:40 hugues23 joined #parrot
18:45 chromatic rgrjr... is not on this channel.
18:45 rgrjr joined #parrot
18:46 rgrjr chromatic:  you rang?
18:46 Coke ha ha!
18:46 davidfetter joined #parrot
18:46 chromatic I'd like to hear more about these problems you're still seeing.
18:48 rgrjr The only new one a this point involves the whole system (runtime+compiler, which is huge), and I haven't even started to track it down.  Don't worry; you'll be hearing from me.  ;-}
18:49 chromatic Do you use the clone opcode?
18:49 Coke chromatic: I think you've killed all the GC related segfaults I'm seeing in tcl. Thank you.
18:49 chromatic You're welcome.
18:49 Coke (I still have several other non GC segfaults, I'll try to isolate them.)
18:49 Tene chromatic++
18:49 chromatic I saw a few strange errors when I ran the tests, but I don't know what they are.
18:49 rgrjr chromatic:  Let me check...
18:50 Tene Every once in a while I look at the number of bug fixes chromatic has contributed, and wonder how doomed Parrot would be without him.
18:51 chromatic A significant chunk of those are bugs I created.
18:51 rgrjr Tene:  And not just the easy ones . . .
18:52 PerlJam chromatic: do you have Munchausen syndrome?  :)
18:52 chromatic That was Dan!
18:53 PerlJam I guess it would be Munchausen by proxy since you're giving parrot the "illness" and then curing it.
18:54 chromatic It's only Munchausen if it's deliberate, even if it's unconscious.
18:57 Ron joined #parrot
18:57 Coke chromatic: so you're just a bad parent, then? ^_-
18:58 chromatic Have you ever met my kids?
18:58 * PerlJam is shocked to find out that you have kids  ;>
18:58 chromatic I don't.  That's why.
18:59 PerlJam chromatic: come on ... you're one of parrot's parents
18:59 particle he has grown up cats. maybe some day he'll be ready for baby goats.
19:00 rgrjr chromatic: Kea-CL does not use clone.
19:01 chromatic rgrjr, then my most recent checkins from last night probably won't fix any of the other bugs you're seeing.
19:01 chromatic Is anyone else seeing one of the IMCC tests fail an assertion for valp->s in imcc.l?
19:03 Coke chromatic: you said clone, do you mean copy?
19:03 Coke (in your commit message. and maybe here.)
19:06 chromatic Yes, I do mean copy.
19:06 chromatic rgrjr, do you use the copy op?
19:07 Tene I need to find a local coffee shop to caffinate myself for an evening of hacking tonight.
19:08 rgrjr chromatic: ah, but I can still check for regressions.
19:09 chromatic Do you have your own PMCs in C?
19:09 stephen left #parrot
19:10 rgrjr Just one, the aforementioned Character that "extends scalar."
19:11 chromatic Hm, so probably not a problem there either.
19:12 Coke Tene: just avoid this: http://www.shamusyoung.com/twentysidedtale/?p=1552
19:12 rgrjr Think so.  But the full system is 140K lines of PIR, so I'm probably stress-testing something ordinary.
19:12 chromatic Tell me you generated most of that.
19:13 rgrjr All but 2.3K.  (I'm not *that* crazy.)
19:14 chromatic Much better.
19:14 Coke why does that sound like someone calling me crazy for tcl? =-)
19:14 rgrjr uh, because, uh, tcl isn't really a compiler?  or something like that.
19:14 Tene Coke: you're crazy for tcl.
19:14 davidfetter tcl: because you don't actually stick bricks together with glue in real life ;)
19:15 davidfetter o/` crazy // crazy for feelin' so TCLish o/`
19:15 chromatic At least Tcl I can run individual programs from the languages/tcl/ directory.
19:15 davidfetter (with apologies to willie nelson)
19:16 trammell joined #parrot
19:16 Coke chromatic: btw, I tried to shorten down that mess of tcl for the test case, but it stopped segfaulting when I pulled out one tcl test from the file.
19:18 chromatic Yeah, it depended upon the GC kicking in at a certain place.
19:18 Coke and when I turned on gcdebug, i had the same issue you had the other way. =-)
19:19 arcady joined #parrot
19:19 rgrjr left #parrot
19:19 Coke just didn't want you to think i was ignoring your earlier request.
19:19 Tene I'm looking for a simple language to implement during my 'Implementing languages with parrot' talk.
19:20 chromatic languages/abc/
19:20 purl languages/abc/ is well documented
19:20 chromatic In theory, IO is easy to implement.
19:20 chromatic Not input/output but the language called IO.
19:20 particle iolanguage.org or something
19:21 * davidfetter queues up, "outland"
19:21 Tene I might just go with a basic stripped-down Generic Imperitive Language.
19:21 * Coke is reminded he has a PL/1 manual at his house.
19:21 particle ooh. logo.
19:21 Tene math, variables, functions, that's it.
19:22 particle bind it to sdl, and let's get turtlebasic!
19:24 * Coke wanders off to get free coffee.
19:25 svnbotl r26207 | chromatic++ | trunk:
19:25 svnbotl : [src] Removed trailing whitespace.
19:25 svnbotl diff: http://parrotvm.org/svn/parrot/revision/?rev=26207
19:25 davidfetter mmm...coffee
19:25 Ademan joined #parrot
19:28 parrot-poke joined #parrot
20:11 pmichaud aiigh, about to lose battery power, no ac outlet available
20:11 pmichaud bbl
21:15 grim_fandango joined #parrot
21:20 peeps[work] joined #parrot
21:45 allison joined #parrot
21:45 allison particle: ping
22:05 davidfetter joined #parrot
22:21 x___ joined #parrot
22:25 skv_____ joined #parrot
22:50 svnbotl r26208 | allison++ | pdd17pmc:
22:50 svnbotl : [pdd17pmc] Bringing the pdd17pmc branch up-to-date with trunk r26207.
22:50 svnbotl diff: http://parrotvm.org/svn/parrot/revision/?rev=26208
22:50 Limbic_Region joined #parrot
23:01 skv_____ joined #parrot
23:14 rdice joined #parrot
23:15 skv_____ joined #parrot
23:21 kid51 joined #parrot
23:27 rdice joined #parrot

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

Parrot | source cross referenced