Camelia, the Perl 6 bug

IRC log for #parrot, 2008-08-25

Parrot | source cross referenced

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

All times shown according to UTC.

Time Nick Message
00:18 bacek joined #parrot
00:38 Andy joined #parrot
00:48 dalek r30526 | jkeenan++ | trunk:
00:48 dalek : A bit more cleanup a la RT 58296.
00:48 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=30526
01:14 dmknopp joined #parrot
01:17 rurban_ joined #parrot
01:34 dalek r30527 | jkeenan++ | trunk:
01:34 dalek : Follow-up on http://rt.perl.org/rt3/Tic​ket/Display.html?id=57884.  Delete some
01:34 dalek : duplicated and/or unnecessary code in Compiler.pm.  Add test for previously
01:34 dalek : unreached branches in 034-step.t.
01:34 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=30527
02:04 gmansi joined #parrot
02:06 tetragon joined #parrot
02:36 s1n joined #parrot
03:02 s1n joined #parrot
03:21 petdance joined #parrot
03:46 tetragon joined #parrot
04:53 tetragon joined #parrot
05:29 Psyche^ joined #parrot
05:35 tetragon joined #parrot
06:15 uniejo joined #parrot
06:19 tetragon joined #parrot
06:21 rurban_ joined #parrot
06:22 mj41_ joined #parrot
06:23 viklund joined #parrot
06:51 Zaba joined #parrot
06:54 gmansi joined #parrot
06:58 dalek r30528 | julianalbo++ | trunk:
06:58 dalek : [PDB] some more refactoring
06:58 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=30528
07:22 masak joined #parrot
07:31 Zaba_ joined #parrot
07:59 tetragon joined #parrot
08:42 dalek r30529 | allison++ | pdd27mmd:
08:42 dalek : [pdd27mmd] Bringing the pdd27mmd branch up-to-date with trunk r30528.
08:42 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=30529
08:50 Zaba joined #parrot
08:55 megamic joined #parrot
08:56 iblechbot joined #parrot
08:58 cognominal_ joined #parrot
09:06 rba joined #parrot
09:14 tetragon joined #parrot
09:17 rurban_ joined #parrot
09:38 tetragon joined #parrot
09:44 Whiteknight joined #parrot
09:52 barney joined #parrot
10:39 megamic joined #parrot
10:40 megamic joined #parrot
11:04 bacek joined #parrot
11:07 bacek g'localtime
11:10 moritz hi bacek ;)
11:11 bacek moritz: hi
11:12 bacek moritz: looks like I've missed first diff from #58308 ;)
11:15 dalek r30530 | julianalbo++ | trunk:
11:15 dalek : [PDB] some fixing of breakpoint's conditions
11:15 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=30530
11:18 Whiteknight joined #parrot
11:47 gmansi joined #parrot
11:54 tetragon joined #parrot
12:15 iblechbot joined #parrot
12:21 dalek r30531 | jkeenan++ | trunk:
12:21 dalek : Eliminate import of init::defaults; not needed for tests in this file.
12:21 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=30531
12:23 moritz how can I mark a bug as a duplicate of another?
12:24 moritz ah, "link" -> "merge into"
12:24 masak ah, so _that's_ how you do it.
12:24 moritz it's not exactly "marked as duplicate", but it's good enough for now
12:28 masak sooth.
12:29 tetragon joined #parrot
12:31 * moritz wonders if PAST::OP() introduces some kind of scope
12:37 moritz masak: I found a workaround
12:37 masak yay!
12:37 masak to what?
12:37 purl it has been said that to is Tonga Islands
12:37 moritz my $m := $str ~~ /.../; if $M { ... }
12:37 moritz to not being able to use $/ in an if statement
12:37 masak ah.
12:37 masak does it work for while loops too?
12:38 masak your commit 30503 made them susceptible, too
12:38 moritz the whole thing is inside a while loop
12:38 masak I mean,
12:38 masak does it work in a case where the while condition is a regex match?
12:39 moritz dunno, I have to try
12:41 dalek r30532 | allison++ | pdd27mmd:
12:41 dalek : [pdd27mmd] Special-case selecting a type for the 'DEFAULT' variant,
12:41 dalek : since it isn't a standard PMC type.
12:41 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=30532
12:49 TonyC joined #parrot
12:49 nopaste joined #parrot
13:01 dalek will@coleda.com | pdd27mmd_tasklist:
13:01 dalek link: http://www.perlfoundation.org/pa​rrot/index.cgi?pdd27mmd_tasklist
13:05 gryphon joined #parrot
13:20 dalek r30533 | julianalbo++ | trunk:
13:20 dalek : [PDB] more src/debug.c refactoring, solves RT#46121
13:20 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=30533
13:27 dalek r30534 | allison++ | pdd27mmd:
13:27 dalek : [pdd27mmd] Strip '*' off the type too, in case the PMC file uses 'PMC* name'
13:27 dalek : declarations instead of 'PMC *name' declarations.
13:27 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=30534
13:37 dalek r30535 | allison++ | pdd27mmd:
13:37 dalek : [pdd27mmd] Add a simple test case for multiple dispatching
13:37 dalek : two Integer types on 'add'.
13:37 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=30535
13:40 rdice joined #parrot
13:50 AndyA joined #parrot
14:01 teknomunk joined #parrot
14:04 jhorwitz joined #parrot
14:13 moritz bacek: I applied both patches, and still get an error from ld
14:18 moritz I'll try again with ccache, if that causes problems
14:20 moritz I also get warnings
14:20 moritz src/ops/string.ops:292: warning: implicit declaration of function ‘string_str_rindex’
14:20 moritz src/ops/string.ops:292: warning: nested extern declaration of ‘string_str_rindex’
14:44 pmichaud rindex in parrot:  that decision needs to be made by allison
14:45 pmichaud in general I wouldn't think that having an rindex opcode is very important.
14:56 grim_fandango joined #parrot
15:01 jhorwitz pmichaud: you here?  :)
15:03 sjansen joined #parrot
15:03 pmichaud yes.
15:03 jhorwitz sweet.
15:04 jhorwitz question about %*ENV -- why is it readonly right now in rakudo?
15:04 pmichaud afaik rakudo isn't doing anything special to make it readonly
15:05 pmichaud so my guess is that it's readonly in Parrot
15:05 jhorwitz no, the Env PMC is writable
15:05 particle Env pmc is not readonly
15:05 dalek r30536 | pmichaud++ | trunk:
15:05 dalek : [rakudo]: spectest-progress.csv update: 127 files, 2278 passing tests
15:05 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=30536
15:05 pmichaud here's the code that sets up %*ENV
15:05 pmichaud env = new 'Env'
15:05 pmichaud set_hll_global '%ENV', env
15:06 jhorwitz i figured there was a reason for it right now, but if not, i can open a ticket
15:06 pmichaud there's no reason for it to be readonly, now.
15:06 pmichaud s/now/no.
15:07 pmichaud yes, if you can set up a test or ticket that demonstrates it as readonly, that'd be great.
15:07 jhorwitz ok, will do.
15:07 pmichaud (the code for %*ENV is at src/builtins/globals.pir:19 )
15:08 jhorwitz yeah, after reading the code, it made no sense why it didn't work
15:08 pmichaud how are you determining that it's not working?
15:08 jhorwitz i write to it, then read from it
15:08 * pmichaud tries that.
15:13 pmichaud looks like a parrotbug (nopaste coming)
15:13 nopaste "pmichaud" at 76.183.97.54 pasted "bug with Parrot Env" (28 lines) at http://nopaste.snit.ch/13869
15:16 pmichaud I'm guessing that ENV only changes the environment in response to the set_*_keyed opcode
15:16 pmichaud s/ENV/Env/
15:18 jhorwitz possibly, cuz this works for me...
15:18 nopaste "jhorwitz" at 96.245.16.64 pasted "env working" (10 lines) at http://nopaste.snit.ch/13870
15:18 pmichaud sure, but that's actually a different opcode :-)
15:19 jhorwitz right
15:20 jhorwitz i'm just saying that it does work.  sort of.  :)
15:20 * jhorwitz goes to open a ticket
15:20 dalek r30537 | julianalbo++ | trunk:
15:20 dalek : [PDB] fix for RT#58358
15:20 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=30537
15:21 nopaste "pmichaud" at 76.183.97.54 pasted "Env FAIL" (23 lines) at http://nopaste.snit.ch/13871
15:22 * jhorwitz understands now
15:23 pmichaud another case where "set" has been confused with "assign" in Parrot :-)
15:24 jhorwitz and probably not the last...
15:45 kj joined #parrot
15:51 Ademan joined #parrot
16:23 davidfetter joined #parrot
16:25 rurban I have tiny Configure patch. Should I paste it?
16:27 nopaste "rurban" at 212.183.63.231 pasted "Init MANIFEST.configure.generated" (15 lines) at http://nopaste.snit.ch/13872
16:29 moritz why is that needed?
16:45 rurban I wrote it at the header
16:45 rurban Without this patch, MANIFEST.configure.generated grows and grows, it should be reset.
16:46 rurban requiring make clean is lame.
16:52 moritz this is so silly - why don't you have a commit bit?
16:52 moritz did somebody offer you one?
17:01 dalek r30538 | moritz++ | trunk:
17:01 dalek : [config] reset MANIFEST.configure.generated before writing to it.
17:01 dalek : Patch courtesy by Reini Urban
17:01 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=30538
17:17 rurban_ joined #parrot
17:17 rurban_ Thanks Moritz
17:17 moritz oh, I forgot the obligatory rurban++ ;-)
17:17 rurban_ I'll get one, sooner or later.
17:17 moritz you're welcome
17:18 moritz I'll suggest it tomorrow at #parrotsketch
17:18 rurban_ I sent the paperwork to the foundation about a week ago.
17:18 moritz very good, that would have been my next question ;)
17:18 rurban_ Now I just need a meta-committer
17:18 rurban_ Then I want to finish my branch. I just finished the new complicated recursive lisp style parser for makefiles :)
17:19 rurban_ 57548-CONDITIONED_LINE_enh.patch
17:19 moritz I think DietCoke is a meta-commiter
17:19 rurban_ Then we can specify all the platforms quirks in the makefile template and don#t have to code around it in a config method
17:20 NotFound rurban_: remember that mail is slow without e- ;-)
17:20 rurban snail
17:21 rurban anyway, there's no hurry now. The relase is out.
17:21 rurban But I have to work on clisp also in the next weeks.
17:21 moritz yes, no hurry, but I think it's sad if you have to do more work to keep your patches up-to-date
17:22 dalek r30539 | julianalbo++ | trunk:
17:22 dalek : [PDB] 'watch' command now works
17:22 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=30539
17:25 rurban Who is the master of JIT here? The new tracing-loop JIT trick by TraceMonky is nice and probably quite easy to add.
17:26 rurban The detect loops (for...) at runtime, and inline it aggressively
17:27 rurban Which is nice, because the types will not change for the good case. There are checks nevertheless, which break out of the fast loop if anything is different
17:27 rurban oops: They detect loops
17:29 rurban http://ejohn.org/blog/tracemonkey/
17:30 dalek r30540 | coke++ | trunk:
17:30 dalek : [docs] Remove reference to old file which this document superseded.
17:30 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=30540
17:30 rurban It reminds me on graphic cards fooling benchmarks
17:30 rurban They just optimize loops with some high loop-count, because only then it pays off.
17:32 rurban And they wrote the JIT interpreter in forth, and compile that with python to C++.
17:33 rurban not JIT interpreter, the JIT'ting interpreter.
17:44 Debolaz joined #parrot
17:48 dalek r30541 | moritz++ | trunk:
17:48 dalek : [config] simplification suggested by DietCoke++
17:48 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=30541
18:01 rurban ok, why not.
18:02 rurban I sent now the rest of my branch code to parrot-porters
18:04 dalek r30542 | julianalbo++ | trunk:
18:04 dalek : [PDB] change output from C stderr and debugee stderr to debugger stderr in a lot of places of src/debug.c
18:04 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=30542
18:05 Copas joined #parrot
18:12 Ron joined #parrot
18:20 moritz rurban: I just applied your cygwin patches to the cygwin070patches branch before noticing the revised patch
18:20 dalek r30543 | moritz++ | cygwin070patches:
18:20 dalek : [config] more cygwin patches by rurban++
18:20 dalek :  * removed spurious TODO comments
18:20 dalek :  * more parsing capabilities for Parrot::Configure::Compiler
18:20 dalek :  * updated some languages/*/config/makefiles/root.in
18:20 dalek :  * fixed some POD
18:20 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=30543
18:20 rurban thanks, no problem. just detected some more minor issues. will send a new one shortly.
18:21 moritz if the difference between the two diffs isn't big, I can also revert the old one and apply the new
18:21 * moritz looks how big the diff is...
18:21 rurban no. i just send a new one.
18:21 moritz ok
18:22 moritz as you wish
18:22 dalek r30544 | allison++ | pdd27mmd:
18:22 dalek : [pdd27mmd] Also capture returns into CallSignature.
18:22 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=30544
18:24 rurban moritz: just sent it
18:25 rurban There's also a ports/cygwin patch sitting in the inbox from 09:53 today. this removes some files and adds some files
18:28 rurban Is it known that a multicore machine with JIT enables produces much more excpetions than a single core? single core none, multi-core several
18:29 rurban test still pass though (cygwin / i686 / 2x core2)
18:29 moritz rurban: is that ports/cygwin patch for the branch, or for trunk?
18:29 rurban that's for trunk
18:29 rurban it's just the update for the recent release
18:30 rurban I update it with every official release and delete the old ones.
18:32 dalek r30545 | moritz++ | trunk:
18:32 dalek : ports/cygwin for 0.7.0-1, patch courtesy by Reini Urban
18:32 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=30545
18:34 moritz I'm confused, why do I have local modifications in my local copy of the cygwin070patches branch? :/
18:35 dalek r30546 | julianalbo++ | trunk:
18:35 dalek : [PDB] more refactoring in src/debug.c
18:35 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=30546
18:35 rurban which files?
18:35 purl which files are burnt at the stake.
18:35 rurban I just found some hard tabs. Now I have to fix xemacs first
18:35 rurban then find my detab script.
18:36 moritz http://sial.org/pbot/31988
18:36 moritz have I been in the wrong dir when I applied r30543 ?
18:37 rurban hmm, all these files were changed by me recently.
18:37 moritz oh wait, I first applied the port-cygwin patch in the branch - maybe that's the cause?
18:37 rurban it looks like a combination of the svn rm stuff + the latest small patch
18:37 moritz anyway, if your patch is against r30543 I'll simply revert those
18:37 rurban ok
18:38 pmichaud grrr... rakudo.org only lets me do one blog post per day?!
18:38 pmichaud or there's a time limit between posts?
18:39 rurban maybe the blog SW needs a unique day for the autogenerated title?
18:39 rurban which would be lame
18:39 pmichaud it's movable type
18:39 pmichaud afaict, the titles aren't date sensitive
18:40 pmichaud I want to post a blog entry about November, but it's not letting me do it.  I'll try again in a few hours.
18:40 dalek r30547 | moritz++ | cygwin070patches:
18:40 dalek : fix cuddled else and some beautification. rurban++
18:40 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=30547
18:42 moritz pmichaud: regarding november, today I hacked up a grammar for HTML::Template, which November uses. It was quite fun ;-) (maybe you want to mention that also in the blog)
18:42 pmichaud I'm primarily pointing people to your use.perl journal entries, which currently don't appear in the perl 6 / rakudo feeds.
18:43 pmichaud they should probably be mentioned on perlbuzz, too :-)
18:43 rurban Anyone knows how to get of hard tabs in emacs?
18:44 rurban get rid of hard tabs, sorry
18:44 rurban found it.
18:44 moritz they are on the perlbuzz front page ;)
18:45 rurban (setq-default indent-tabs-mode nil)
18:45 moritz under the title "How cool Perl 6 really is"
18:45 moritz (which brought about 350 hits so far; not bad ;-)
18:46 pmichaud ah, that primarily talks about the perl 6 posts (also cool) but doesn't mention november
18:47 moritz right
18:47 pmichaud I think november deserves special mention, since it's a working application :-)
18:48 moritz aye ;)
18:48 moritz I quite enjoyed the video from the yapc lighting talk about november ;)
18:48 tewk november?
18:48 purl i heard november was the 'dead' month....all rain and darkness
18:49 pmichaud (see?  if tewk hasn't heard of november yet, that's why we need a blog post :-)
18:49 moritz tewk: a wiki engine writen in Perl 6
18:49 tewk post! post! post! :)
18:49 tewk Is there a webpage, link, code?
18:49 pmichaud I'd like to post, but rakudo.org isn't letting me :-|
18:50 moritz http://github.com/viklund/november/tree/master
18:50 pmichaud tewk:  http://use.perl.org/~masak/journal/
18:50 tewk thanx
18:51 ejs joined #parrot
18:51 moritz purl, november is also a perl 6 wiki, see http://use.perl.org/~masak/journal/37212
18:51 purl okay, moritz.
18:53 pmichaud ooooooh, my lightning talk is in video, too :-)
18:53 moritz aye ;)
18:53 moritz that was also quite fun to watch
18:53 pmichaud paula will be very happy
18:54 pmichaud (paula == my wife)
18:56 Zaba_ joined #parrot
19:01 rurban pmichaud: video url?
19:03 cotto_work rurban: yapc.tv
19:03 cotto_work it's both awesome and perverse
19:05 moritz aye
19:05 cotto_work I wonder how something like that would work for porting libraries between normal HLLs
19:10 rurban M-x untabify ! oh my
19:26 paco joined #parrot
20:13 rurban Moritz, I have one more patch for my branch
20:14 moritz rurban: go ahead
20:14 Infinoid joined #parrot
20:15 rurban Allison doesn't like lisp, hmm.
20:15 rurban Well, I can try to use stupic ALGOL syntax also...
20:16 rurban "var1 or var2 and var3 and not var4" what are the precedence rules for that?
20:17 rurban compared to "(or var1 (and var2 var3 (not (var4)))"
20:17 moritz same as in perl
20:17 rurban "var1 or (var2 and var3 and not var4)"
20:17 moritz so it's var1 or (var2 and var3 and (not var4))
20:17 particle just use parens to disambiguate
20:18 rurban I dnot like to add an AND between all possibilities. This make a bad list
20:18 rurban or an OR. just stupid.
20:18 rurban well, oh well
20:19 moritz some people like lisp, and some people like syntax
20:19 moritz that's the way the world is
20:20 rurban (cygwin|win32&(!has_glut)) would be perl then
20:21 rurban ok, something like that maybe
20:21 rurban binding rules... how stupid
20:22 moritz some people like to be able to write stuff like 'for @list Z 0..* -> $item, $index { ... }', which is why they invented Perl 6, with all its precedence levels and all it's associativities
20:22 moritz ;-)
20:23 rurban well oh well. almost bf
20:24 rurban I'll write an assembler now in perl5, to reach a higher sanity level
20:24 rurban I need it to test JIT'ting.
20:24 dalek r30548 | moritz++ | cygwin070patches:
20:24 dalek : fixed cuddled 'else' and tabs, rurban++. Patch coutesy Reini Urban.
20:24 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=30548
20:24 moritz thankfully there's more than one programming language, so we don't have to agree on one syntax for everybody ;)
20:25 rurban assemblers in lisp are the easiest to write :)
20:25 rurban perl is also fine.
20:25 purl okay, rurban.
20:31 rurban ok, I'll go to bed now. tomorrow I'll implement a perl-like sytnax for makefile template macros.
21:00 dalek allison@perl.org | pdd27mmd_tasklist:
21:00 dalek link: http://www.perlfoundation.org/pa​rrot/index.cgi?pdd27mmd_tasklist
21:02 dalek allison@perl.org | pdd27mmd_tasklist:
21:02 dalek link: http://www.perlfoundation.org/pa​rrot/index.cgi?pdd27mmd_tasklist
21:04 dalek allison@perl.org | pdd27mmd_tasklist:
21:04 dalek link: http://www.perlfoundation.org/pa​rrot/index.cgi?pdd27mmd_tasklist
21:11 peepsalot joined #parrot
21:16 Zaba joined #parrot
21:21 Auzon joined #parrot
21:51 Whiteknight my laptop is crashing, can't develop parrot :(
21:55 Debolaz joined #parrot
21:56 Debolaz joined #parrot
21:59 Whiteknight yay! I uncrashed my laptop and can get back to Parrot
22:03 Limbic_Region joined #parrot
22:06 Tene Whiteknight: aw. :(
22:07 dalek r30549 | Whiteknight++ | pdd09gc:
22:07 dalek : [pdd09gc] update to trunk from r30475
22:07 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=30549
22:08 Tene Whiteknight: http://pleasedieinafire.net​/~tene/parrotexceptionnotes
22:20 AndyA joined #parrot
22:21 Whiteknight thanks Tene, I'm reading it now
23:12 dalek r30550 | Whiteknight++ | trunk:
23:12 dalek : [Core] add some function-level documentation to src/pic_jit.c. I can't document all these functions, however, because i do not know what they all do.
23:12 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=30550
23:35 tetragon joined #parrot
23:41 dalek r30551 | Whiteknight++ | trunk:
23:41 dalek : [Core] add some function-level documentation to src/nci_test.c.
23:42 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=30551
23:55 dalek r30552 | Whiteknight++ | trunk:
23:55 dalek : [Core] add some function-level documentation to src/global.c.
23:55 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=30552
23:57 davidfetter joined #parrot

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

Parrot | source cross referenced