parrot.org/ - clean up those smolders for the release!
Set by moderator on 20 October 2008.
00:10 AndyA joined
chromatic make testj: Error: PARROT_JIT_FLAG is set, but interpreter is not JIT_CAPABLE! 00:19
NotFound chromatic: the -j option set that flag unconditionally 00:28
chromatic Hm. I'm not sure what I was testing for then. 00:29
NotFound Maybe an option error in parseflags is more reasonable that let if fail later. 00:31
chromatic I configured with --jitcapable=0, thinking that that would force JIT NCI calls, but I don't know why I thought that. 00:32
NotFound After halloween effects? ;) 00:33
chromatic I only slept ten hours. That must be it.
00:42 Theory joined 02:02 allison joined 02:19 allison joined 02:23 allison joined 03:51 Psyche^ joined 03:58 TiMBuS joined 05:40 mberends joined 05:53 tetragon joined 06:14 tetragon joined 06:36 Bzek joined 07:09 mberends joined 08:01 Ademan joined 09:00 baest joined
cotto it's nice to be in the single day of the year when daylight savings time does something useful 09:11
it's also nice to have most of my clocks update themselves 09:12
dalek r32279 | fperrad++ | trunk: 09:26
: [Lua] LPeg
: now aligned with version 0.9 09:27
diff: www.parrotvm.org/svn/parrot/revision?rev=32279
09:29 iblechbot joined 10:38 masak joined
masak 'is export' is in place in Rakudo now, right? 10:47
I'm asking because rt.perl.org/rt3/Ticket/Display.html?id=59104 is blocking a server upgrade for us in the November project right now. 10:49
there's no real hurry, but I'm curious as to what the next step is to getting precompiled PIR modules to work again. 10:50
10:52 contingencyplan joined 10:55 bacek joined 10:56 contingencyplan joined
NotFound Is a shame that November is blocked in november. 10:58
masak nah -- there'll be more Novembers 11:00
besides, it's still early in the month
and we're not _blocked_ blocked, we just can't upgrade the server without getting incredibly slow again, like we were in August
NotFound "There'll be more Novembers"... looks like a great title for a romantic comedy 11:01
masak someone probably already wrote it. :)
cognominal the hunt for red November? 11:02
masak cognominal: ooh, that's a great idea for a skin for November, thank you!
NotFound Be careful, boys, someone in the US can think we are a communist faction X-) 11:05
In soviet Russia, November skins you 11:06
masak NotFound: it wouldn't be the default skin. :) and the communist side of it can be toned down appropriately 11:07
cognominal NotFound, I thought US was going socialist
masak I'm thinkking of emulating CSS Zen garden in some very small way in November
cognominal in France, socialism and communism parties are in decay, the most popular figure in the left wing is at the left of the left. 11:10
en.wikipedia.org/wiki/Olivier_Besancenot
masak Sweden got a right-wing government in the last electon, first time in at least a decade 11:12
cognominal and he got more popular recently because the Taser french company put a tail on him and he figured out. I think the prosecution os ongoing.
our right wing president had US as a model when he was elected :) 11:13
but people forget so fast...
jonathan masak: I'll put fixing the issues that came up in pre-compiled PIR modules on my task list. 11:15
masak \\o/
jonathan Would be good to have some tests for that stuff to, if there aren't any written yet.
Want to make sure I fix the regression without breaking the thing I fixed that introduced the regression in the first place...
...if I could remember what that was. :-) 11:16
masak jonathan: rt.perl.org/rt3/Ticket/Display.html?id=59104
hooray for rt!
jonathan Aha, thanks. 11:17
Hinrik perlcabal.org/syn/ has a link to a nonexistent Formats apocalypse 11:18
masak Hinrik: aye, it does. I know I've seen that apocalypse somewhere, though. 11:19
and ISTR that the module it describes is on the CPAN already.
Hinrik masak: really? the preceding Apocalypse says there will be no Formats Apocalypse
masak hm.
Hinrik there's just the Exegesis 11:20
jonathan Hey, the Rakudo queue isn't as big as I'd feared. 11:21
I expected we'd have passed 200.
masak Hinrik: that sounds familiar, now that you mention it.
jonathan: I've been a bit busy, sorry :P 11:22
jonathan :-P
masak btw, there's an infinite loop when doing `say Bool.list` in Rakudo, or any other class name instead of Bool 11:23
it's because of the method on line 576 in src/classes/Object.pir
I have a patch for fixing that problem, hold on
nopaste "masak" at 130.238.45.242 pasted "patch for fixing the problem with infinite looping on `Bool.list`" (13 lines) at nopaste.snit.ch/14445 11:24
masak if it looks ok, I can apply it
jonathan I'd suspect maybe the intention was to call the list built-in, but instead we call this method. 11:26
masak yes.
jonathan To call that do
$P0 = get_hll_global 'list'
.return $P0(self)
masak oki, will try that
jonathan I think that is preferable to directly dealing the the List class.
Hinrik what does hll stand for? 11:27
masak Hinrik: High Level Language.
jonathan: aye.
Hinrik ok
jonathan sees lots to be doing in the RT queue. 11:31
dalek r32280 | masak++ | trunk: 11:41
: [rakudo] Object.list used to call itself and loop indefinitely, now DTRT instead
diff: www.parrotvm.org/svn/parrot/revision?rev=32280
11:53 cotto joined 12:50 baest joined 13:44 kj joined
dalek r32281 | kjs++ | trunk: 13:50
: [pirc/new] fix register allocator + some other stuff.
: + fix sorting/inserting of list
: + handle :unique_reg symbols properly
: + disable local fixup stuff; needs more work.
: + implement resizing for macro buffer if necessary.
: + add commandline option to specify initial macro buffer size.
diff: www.parrotvm.org/svn/parrot/revision?rev=32281
13:53 masak joined 15:06 davidfetter joined 15:14 davidfetter joined 15:16 jan joined 15:45 grim_fandango joined 15:48 kj joined
tewk japhb: if you can provide a small test case for opengl examples failure, I would appreciate it. 16:11
add it to nci.t
The obvious tests pass, even the sigs I captured from running opengl examples. 16:12
NotFound I added a 'say now' in .sub idle in triangle.pit and it says 'nan' 16:46
16:48 jq joined
NotFound And commentig out all variables, and letting the idle does just: glRotatef(0, 0, 1, 0) it paints an static triangle. 16:50
So the problems seems to be in the parrot environment during the callback. 16:51
16:54 allison joined 17:23 notbenh joined
japhb bak 17:34
NotFound: can you nopaste your changes? 17:35
NotFound Now is not working, I'm not sure what I changed... 17:43
reverting and rebuilding... 17:44
japhb is creating a cut-down version of triangle.pir. 17:45
Rebuilding to test both ways, then will commit
17:50 Theory joined
japhb bak 18:34
Yep, cut down example still shows black window with NCI JIT, works normally without. 18:35
18:42 notbenh joined
dalek r32282 | japhb++ | trunk: 18:43
: [OpenGL] New cut-down OpenGL example for NCI debugging
: * Add examples/opengl/static-triangle.pir, a cut-down copy of
: triangle.pir to help with debugging NCI.
: * Still shows same "black window" bug when NCI JIT is enabled.
diff: www.parrotvm.org/svn/parrot/revision?rev=32282
r32283 | japhb++ | trunk:
: [OpenGL] Add examples/opengl/static-triangle.pir to MANIFEST
diff: www.parrotvm.org/svn/parrot/revision?rev=32283
NotFound japhb: sorry, I tested a lot of things at the same time and don't know what was the combination that showed something.
japhb NotFound: no problem, hopefully static-triangle.pir will give us a common base to work from. We can keep cutting things out as needed from there. 18:44
OK, svn props now set 18:46
(Man, I wish git-svn just made that work, rather than needing two checkouts.) 18:47
dalek r32284 | japhb++ | trunk:
: [OpenGL] Add svn props to static-triangle.pir
diff: www.parrotvm.org/svn/parrot/revision?rev=32284
japhb tewk: is examples/opengl/static-triangle.pir simplified enough for you?
kj moritz: ping 19:19
moritz kj: pong 19:20
kj hi, any news on your box?
eh, being able to reach svn.perl
moritz kj: aye. It was sending up to one request each 2 seconds to svn.perl.org, and has been firewalled... 19:21
kj: so I need to fix the setup, and then I'll get unblocked
kj oh ok
every 2 seconds... what happened?
moritz dunno, I suspect broken IRC bots
I plan to redo them with plagger, that's tested and should work 19:22
kj plagger?
purl plagger is plagger.org/ or available at CPAN, search.cpan.org/dist/Plagger/ or Pluggable RSS/Atom Aggregator
kj ok, great, thanks
moritz now if only I could find that mail where the admins told me where the rss feeds are... 19:28
kj Parrot::OpsFile: Unrecognized line: 'inline op end() :base_core :check_event :flow { 19:33
'!
make: *** [lib/Parrot/OpLib/core.pm] Fehler 9
moritz: I get this error (just downloaded a fresh parrot through winscp) 19:34
moritz isn't familiar with this one 19:35
kj I saw it's perl 5.8.8
is that ok?
moritz in general, yes 19:36
did you try a 'make realclean' first?
kj well, it's a complete fresh one
just downloaded it on my own pc, then uploaded a zipped version onto your box 19:37
moritz is your own PC a windows box?
kj yes
moritz maybe wrong line endings then?
svn converts those
kj ah yes
that may be it
moritz I can rsync a copy from my own local machine, just a sec... 19:38
kj: parrot copy is in ~moritz/parrot2/ 19:50
kj thanks!
moritz bah, plagger is non-trivial :( 20:09
20:20 rindolf joined
rindolf Hi all. 20:20
masak rindolf: hello :) 20:21
rindolf masak: hi.
21:09 cognominal joined
dalek r32285 | allison++ | pdd22io: 21:18
: [pdd22io] Convert the core public I/O API to the new architecture.
diff: www.parrotvm.org/svn/parrot/revision?rev=32285
21:30 Zaba joined 21:37 Zaba joined
dalek r32286 | allison++ | pdd22io: 21:40
: [pdd22io] Linked the platform I/O headers into the main header. Generated
: prototypes with 'make headerizer'.
diff: www.parrotvm.org/svn/parrot/revision?rev=32286
r32287 | allison++ | pdd22io: 21:43
: [pdd22io] Added new I/O C files and CONST_STRING files to the Makefile
: template.
diff: www.parrotvm.org/svn/parrot/revision?rev=32287
r32288 | allison++ | pdd22io: 21:47
: [pdd22io] Cleanups in the platform and common utility functions found while
: running the headerizer or compiling.
diff: www.parrotvm.org/svn/parrot/revision?rev=32288
moritz kj: timtowtdi.org is now unblocked, ie svn shoudl work again. 22:04
kj ok, great 22:07
22:09 chromatic joined 22:18 rdice joined
kj are PIR macros supposed to work in PASM mode? 22:22
(general question to anybody who thinks s/he knows :-) ) 22:23
kj thinks it's a PIR feature
Tene I would be very surprised if they did. 22:25
kj i was more wondering whether they should..
moritz thinks that PASM is too low-level to have such convenience features as macros 22:28
dalek r32289 | kjs++ | trunk: 22:31
: [pirc/new] integrate PASM grammar/lexer into PIRC's lexer/parser.
: + use a neat trick to decide whether to scan PASM or PIR
: + fix a typo in a PARROT_ASSERTion
: + PASM code is not handled yet, but parses, mostly -- needs some extra token definitions.
: + remove some old code.
diff: www.parrotvm.org/svn/parrot/revision?rev=32289
jonathan wonders what the neat trick is 22:32
kj hehe
I initialize the lexer in a special start state, which reads the first character. Then, based on a flag set in the lexer data structure, it switches to PASM state or "normal" (INITIAL) state. First, it gives back the first character 22:33
I looked at how IMCC does it, but as far as I understand that code, it executes a particular block (at the top of the lexer spec. ) for EACH token (I think, didn't try). If that's actually happening, that may cost quite some speed. 22:34
well, that was fun (hacking) :-) Time to catch the last bus home. Good night 22:36
chromatic The sooner we replace IMCC, the better. 22:38
Tene I've got a peach here we could replace it with. It doesn't compile PIR, but it does taste delicious. 22:39
allison Tene++ 22:40
chromatic It's a start. 22:41
Tene Ack, nm. The girlfriend is eating it now. 22:42
jonathan allison: Any idea when the travel grants get reviewed for the summit? 22:45
allison Tene: "Hey, that peach is our next-generation assembly language compiler!" I guess that's one benefit of revision control. No excuses about the girlfriend eating your homework. ;) 22:47
jonathan: we did one last week, will ping for another
jonathan allison: OK. Would be good to know if I'm going to be there or not, for scheduling other work. 22:48
Plus its a bit of a pre-req to seeing if I can room share with anyone... :-) 22:52
Woo. Someone has signed up to the hackathon I'm hosting at my place after the Twin City Perl Workshop! 22:55
23:07 kid51 joined 23:25 dmknopp joined

Powered by Cro and the Raku® Programming Language.

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

Thank you!