Parrot 2.6.0 | parrot.org Log: irclog.perlgeek.de/parrot/today | Nopaste: nopaste.snit.ch:8001 | merge html_cleanup (talk to Coke), merge gc_* branches, fix/replace/optimize hashing
Set by moderator on 10 August 2010.
Paul_the_Greek Yup, the scarcity of tuits. 00:00
00:00 Psyche^ joined
Paul_the_Greek Are there any builtin constants in PIR? 00:20
chromatic Not really, no. 00:21
Paul_the_Greek Constants is the wrong word. Does IMCC have macros? 00:22
Never mind, that's not right either. 00:23
If TRUE and FALSE were built-in, they would have to be literals.
Ah, there is .macro_const 00:26
00:27 bubaflub joined
tcurtis Also .macro. 00:29
Paul_the_Greek So I would have to code .true and .false 00:30
What's the story on IMCC vs. PIRC vs. PIRATE? 00:31
Well, I'm off to the beach. Have a happy and productive two weeks, folks. 00:34
dafrito Paul_the_Greek: have a good trip :) 00:35
Paul_the_Greek Thanks!
chromatic Oops, that isn't what I intended. 00:36
dalek rrot: r48473 | chromatic++ | branches/gc_threshold_tuning/src/pmc/fixedstringarray.pmc:
fix codetest failure - line length
rrot: r48474 | chromatic++ | branches/gc_threshold_tuning (2 files):
[TT #1738] Improve docs of PARROT_ERROR_*_FLAG Paul_the_Greek++
rrot: r48475 | chromatic++ | branches/gc_threshold_tuning/t/native_pbc (7 files):
native_pbc platform updates
rrot: r48476 | chromatic++ | branches/gc_threshold_tuning/DEPRECATED.pod:
the first release after the next supported release would be 2.10, not 2.7
rrot: r48477 | chromatic++ | branches/gc_threshold_tuning/t/op/calling.t:
add test for TT#1733
rrot: r48478 | chromatic++ | branches/gc_threshold_tuning/src/pmc/imageio.pmc:
[PMC] Fixed segfaults in ImageIO PMC's destroy().

attribute" flag is on but there's no PackFile attribute. This makes C sad. The culprit is r48447. A better approach is to use PObj_custom_destroy_SET() only when it's okay to destroy this attribute, but that fix will take longer. In the meantime, this works at the expense of a bit of extra work.
purl i already had it that way, dalek.
rrot: r48479 | chromatic++ | branches/gc_threshold_tuning/src/gc (2 files):
[GC] Made a GC callback function constant.
rrot: r48480 | chromatic++ | branches/gc_threshold_tuning/src/pmc/hash.pmc:
[PMC] Made Hash PMC's mark() skip empty hashes.
rrot: r48481 | chromatic++ | branches/gc_threshold_tuning/src/gc/alloc_resources.c:
[GC] Skipped compact_pool given full blocks.
rrot: r48482 | chromatic++ | branches/gc_threshold_tuning/src/hll.c:
[HLL] Optimized HLL typemap fetches.
kthakore hi chromatic 00:38
whiteknight NotFound: ping
kthakore hi whiteknight
whiteknight hello kthakore
kthakore whiteknight: I tried more parrotSDL today ... but I keep gettin a sefgault ...
has the NCI stuff changed again? 00:39
whiteknight what segfault? where is it happening?
kthakore I have no clue
rebooting my vm
whiteknight: I am going to try a fresh machine and parrot and sdl
:
it might be a case of old machine
or dir is not cleared 00:40
but NCI doesn't seem to have changed ...
PacoLinux whiteknight: NotFound is problably sleeping, in spain is very late in the night (02:40)
kthakore so I am thinking that
whiteknight PacoLinux: ah, thanks for the info. I'll send him some msgs then
00:49 kid51 joined
dalek rrot: r48483 | chromatic++ | trunk/src/gc (2 files):
[GC] Made a GC callback function constant.
 00:53
rrot: r48484 | chromatic++ | trunk/src/pmc/hash.pmc:
[PMC] Made Hash PMC's mark() skip empty hashes.
rrot: r48485 | chromatic++ | trunk/src/gc (4 files):
Merged gc_threshold_tuning branch into trunk.
rrot: r48486 | chromatic++ | trunk/src/hll.c:
[HLL] Optimized HLL typemap fetches.
bubaflub chromatic: are we seeing any speedups from the gc_threshold_tuning branch? 01:02
chromatic Yes.
Someone else will have to benchmark Rakudo Star against modern Parrot and modern Rakudo, but you'll notice a difference. 01:03
bubaflub sounds great. i'm a sucker for your "this improves _____ by ____%" 01:04
chromatic I had enough tunings outstanding that I'm lazy for numbers today. 01:05
sorear bubaflub: How is Parrot/RTEMS doing nowadays?
bubaflub sorear: nearing the end of GSoC
so i have to turn in what i have
which is 1) a Configure.pl file to help automatically use RTEMS tools
2) a set of patches to help Configure.pl run and build out of directory 01:06
i planned on separating system parrot from target parrot so the Makefile could handle cross-compiling, but i didn't have time
hopefully i can hack on it over the semester and have something by the end of the year
01:16 davidfetter joined 01:19 s1n joined 01:25 jsut joined
cotto notices that the meaning of "modern" seems to be shifting 01:27
or at least becoming more precise 01:28
chromatic or punny 01:29
01:41 rurban_ joined 01:42 s1n joined
dalek rrot-linear-algebra: c2cc2a9 | Whiteknight++ | t/harness:
small cleanups in the harness
 01:44
rrot-linear-algebra: b1fa723 | Whiteknight++ | t/testlib/matrixtest.nqp:
remove trampoline subs I added when Kakapo was b0rked. Add a test method to
rrot-linear-algebra: c82c8b7 | Whiteknight++ | t/testlib/matrixtest.nqp:
add short descriptions for all the common tests. Implement one stub test that
rrot-linear-algebra: 5ec5971 | Whiteknight++ | (3 files):
Add an item_at method for nummatrix2d, which obviates the need to use kakapo's

complex types) for this new method. Rewrite the elimination example to use pla.nqp, not use kakapo, and use this new method to make things easier.		 02:02
whiteknight purl msg NotFound I'm starting to play with winxed. Is there any way to install it? I don't see an "install" make target. 02:04
purl Message for notfound stored.
02:41 janus joined
dalek rrot: r48487 | jkeenan++ | trunk/src/pmc/null.pmc:
[codingstd] Insert POD 'item' so that documentor will know where to add function documentation.
 02:49
02:53 theory joined 02:55 khairul joined 03:05 brianwisti joined
dalek rrot: r48488 | jkeenan++ | trunk/src/pmc/fixedintegerarray.pmc:
Correct spelling error in POD.
 03:06
khairul cotto: ping 03:09
03:10 Chandon joined
cotto khairul, pong 03:16
dukeleto, ping 03:30
dalek rtcl-nqp: 2b66547 | Coke++ | src/ (84 files):
use more vim
 04:38
rtcl-nqp: db6793d | Coke++ | (2 files):
whitespace
purl whitespace is the ... final... frontier.
rtcl-nqp: 372e0aa | Coke++ | t/cmd_dict.t:
Add argument validation tests for [dict]
rtcl-nqp: 9ffa6a7 | Coke++ | src/Partcl/commands/dict.pm:
convert [dict] to using Austin++ style subcommands, passing new test file
purl I don't know how to convert [dict] to using Austin++ style subcommands, passing new test file.
brianwisti purl seems to be arguing with dalek. 04:42
purl brianwisti: huh?
Coke partcl-nqp: puts {opbots, names} 04:56
04:56 bubaflub joined
p6eval partcl-nqp: OUTPUT«opbots, names␤» 04:56
05:11 brianwisti joined 05:27 jrtayloriv joined 05:35 jrtayloriv left
cotto seen dukeleto 05:49
purl dukeleto was last seen on #parrot 7 hours, 50 minutes and 59 seconds ago, saying: luben: gotta go, good luck!
cotto msg khairul Make sure the documentation on the GC instrumentation includes the list of instrumentable functions and a pointer to the docs that explain them. 05:57
purl Message for khairul stored.
06:06 brianwisti joined
cotto seen fperrad 06:11
purl fperrad was last seen on #parrot 1 days, 9 hours, 42 minutes and 43 seconds ago, saying: Paul_the_Greek have you seen developer.berlios.de/projects/win32gmp/ ? [Aug 12 20:28:18 2010]
06:26 simcop238 joined
cotto Is there a better way to search through the contents of past commits than to use my Thunderbird's insane search tool? 06:37
sorear git log -p | grep 06:38
in a git-svn clone
dalek tracwiki: v9 | cotto++ | ParrotDeprecations 06:39
tracwiki: add find/store_global_s
tracwiki: trac.parrot.org/parrot/wiki/ParrotD...ction=diff
cotto That seems to have killed my pager. 06:40
but still quite helpful 06:41
thanks
dalek tracwiki: v4 | cotto++ | ParrotDeprecationsFor2.9 06:56
tracwiki: add find_global_s and store_global_s
tracwiki: trac.parrot.org/parrot/wiki/ParrotD...ction=diff
07:17 fperrad joined
mikehh All tests PASS (pre/post-config, make corevm/make coretest, test, fulltest) at r48488 - Ubuntu 10.04 amd64 (gcc) 07:17
cotto hio fperrad 07:22
07:42 Austin_away joined
Austin pmichaud, ping 07:43
Question for the back-scrollers: 07:45
In NQP (NOT perl6!) I have a grammar that includes a pattern like this: 7 $<dest>=[ <reg_id> || <target> ] 07:46
I use || instead of | because the reg-id could include an identifier - a macro - while the target rule considers a bare identifier to be an error, and .panic's. So if an identifier is a valid reg-id, it's okay - thus, an ordered check. 07:47
But when I try to access $<dest>.ast, I get Undef, despite a trace showing me that the <target> action method has run. Why is this? Is this a nqprx bug, or some kind of expected perl6 behavior, or what? 07:48
sorear Expected Perl6 behavior is that you'll need to use $<dest><target>.ast 07:58
$<dest>=[ ] effectively creates an anonymous subrule
Austin How is that? I'm defining $<dest> as one or the other?
Well, never mind. That's a bug in nqp, since the resulting Match doesn't have that structure. 08:00
But using [ <dest=a> || <dest=b> ] worked...
But using [ <dest=a> || <dest=b> ] worked... 08:01
bah. ww 08:02
NotFound msh whitenight No winxed installer yet, sorry 08:47
msg whitenight No winxed installer yet, sorry
purl Sorry, I've never seen whitenight before.
NotFound msg whiteknight No winxed installer yet, sorry
purl Message for whiteknight stored.
NotFound I'm still half-awake :o 08:48
08:50 lucian joined
moritz msg chromatic I don't know what changed, but current rakudo on parro ttrunk takes ~2G virtual mem und 1.3G RES to compile the setting... :( 09:00
purl Message for chromatic stored.
09:12 robin-gvx joined 09:24 barney joined
dalek rrot: r48489 | fperrad++ | trunk/examples/languages/squaak/setup.pir:
[squaak] fix build after r48212
 09:26
09:41 rurban_ joined 10:59 whiteknight joined
whiteknight good morning, #parrot 11:29
12:04 desertm4x joined
dalek rrot-linear-algebra: 5a36038 | Whiteknight++ | src/pmc/ (3 files):
Update PMC- and ComplexMatrix2D to use the DECLATTRs macro, for cleanliness and

again		 12:10
rrot-linear-algebra: b185b2c | Whiteknight++ | (4 files):
use PLATYPENAME macro consistently in all types. Add bounds checking to all

types
12:18 allison joined
whiteknight NotFound: ping 12:22
dalek zy-k: b86350b | bernhard++ | t/examples.t:
Elimininated an unused variable.
 12:28
zy-k: c45102f | bernhard++ | lazy.pir:
Added note why lazy-k is currently broken.
12:39 ruoso joined 12:46 Paul_the_Greek joined
Paul_the_Greek ping cotto 12:46
msg cotto I replied to your message on parrot-tickets, but it turned out I wasn't subscribed. The message is pending approval. I've subscribed. 12:47
purl Message for cotto stored.
Paul_the_Greek purl,messages
dalek tracwiki: v139 | barney++ | Languages 13:03
tracwiki: Updated status of Lazy-K and Unlambda
tracwiki: trac.parrot.org/parrot/wiki/Languag...ction=diff
13:07 bubaflub joined 13:20 kid51 joined 13:51 AndyA joined 14:23 Paul_the_Greek joined
dalek kudo: 9e1bcd4 | pmichaud++ | src/binder/bind.c:
Some refactors to array creation in the binder prior to @_ updates.
 14:40
14:58 jimk joined
jimk An interesting search result: "Sorry, 'www.google.com' does not exist or is not available." 14:58
15:01 theory joined 15:04 Paul_the_Greek joined
Paul_the_Greek Question about redoing a patch? 15:04
jimk state your ?
Paul_the_Greek I checked in a patch about a week ago, but now it conflicts with changes committed since. 15:05
What is my best approach to reworking it?
I figure I update to the latest file, then diff against my changed file, then merge my changes into the latest file.
jimk If you have an older version which is a Trac attachment, you can delete that and attach a new one. I believe there is also a replace option.
Paul_the_Greek Yes, that will work. Is my approach to merging in my changes the best one? 15:06
jimk Will you be supplying an 'svn diff' or a regular 'diff'? 15:07
Paul_the_Greek svn diff.
purl svn diff is better than cvs log, that's why I like svn more than cvs
jimk Sounds okay. The only problem -- which would be outside of your control -- would be if the file you are patching has had major alterations recently in the area you are patching. 15:09
bbial
Paul_the_Greek Yes, it has, which is why I have to start over.
My changes are mostly comment improvements, not logic changes, so not a big deal. 15:10
To prevent this from happening again, I have to push to have my changes committed quickly.
Once I resubmit the patch, that is.
15:12 chromatic joined
Paul_the_Greek Hey chromatic. 15:13
chromatic Morning.
Paul_the_Greek Well, it really is time to leave for the beach. Take care, all.
chromatic msg moritz Can you try the gc_threshold_tuning branch at r48480 to see if r48481 caused that behavior? Thanks. 15:15
purl Message for moritz stored.
arnsholt Is there a cons-list/linked list PMC? 15:26
moritz chromatic: r48480 seems pretty normal, with about ~800M memory usage during compilation
15:42 whiteknight joined 15:49 robin-gvx joined
moritz chromatic: r48481 and r48489 on that branch seem fine too... I wonder if I mis-read my memory monitor... 15:55
must have been a bad dream... never mind 16:00
Coke I am not a big fan of individually playing commits into branches and then emailing them all over again. 16:04
(but I do appreciate trying not to lose commits with bulk merging.)
moritz: bad moritz, no donut.
moritz come to think of it, I've probably swapped the memory figure with that of my browser 16:06
chromatic Coke, that was my mistake with git svn. 16:09
16:16 khairul joined 16:18 Coke joined
khairul msg cotto i think the user guide is done-ish. parrot.mangkok.com/?page_id=140 16:43
purl Message for cotto stored.
16:46 davidfetter joined 17:12 dafrito joined 17:41 rurban_ joined 17:48 plobsing joined 18:46 dafrito joined 19:03 davidfetter joined
cotto ~~ 19:17
massages
purl massages is probably wrong.
cotto no, massages is <reply>You have no massages waiting. 19:18
purl okay, cotto.
dalek rrot: r48490 | plobsing++ | trunk/src (2 files):
use PObj_custom_destroy_FLAG as marker for destroying pf in ImageIO
 19:22
19:23 Andy joined
dalek rrot: r48491 | petdance++ | trunk/t/compilers/pge/p5regex/p5rx.t:
took out a redundant vi
 19:39
19:56 tcurtis joined 19:59 kid51 joined 20:00 dha joined 20:03 dha joined
dha whois dalek 20:03
purl dalek is #parrot's spammy little rss bot or (see: dalek plugins)
cotto whiteknight, ping 20:07
20:07 Andy joined
cotto msg whiteknight Can you review the comments added by trac.parrot.org/parrot/changeset/48492 to make sure they're accurate? 20:11
purl Message for whiteknight stored.
dalek rrot: r48492 | cotto++ | trunk (7 files):
[gc] update docs in pobj.h, patch courtesy of paul_the_greek++
 20:12
20:19 dafrito joined 20:22 darbelo joined 20:23 perlite joined
darbelo cotto: ping 20:36
Nevermind, my bad. 20:37
cotto: un-ping 20:38
Although... 20:39
cotto: un-un-ping
cotto hio 20:40
darbelo, pong 20:41
20:41 preflex joined
darbelo I think something went wrong with your CREDITS update in r48492 20:42
See trac.parrot.org/parrot/changeset/48...nk/CREDITS
cotto oh my
I even have a nice alias for reviewing svn commits before I send them off. me-- for not using it 20:46
all fixed 20:48
20:51 bubaflub joined 20:53 chromatic joined
dalek rrot: r48493 | cotto++ | trunk/CREDITS:
restore some accidentally clobbered credits entries noticed by darbelo++
 21:03
dafrito In pod's, what does the N<text> formatting code mean? 21:13
plobsing perldoc seems to think it is invalid 21:16
dafrito I can't find any docs on them, either, but they're used in Parrot's docs. I'll look further into it 21:17
plobsing It also whines about the Z<> codes which seem to be associated with book-specific POD. Maybe the N<> codes are too. 21:18
dafrito currently, they only cause apparent typos, as in the "declarationor" found here: docs.parrot.org/parrot/latest/html/...s.pod.html
plobsing dafrito: svn praise might show some insight. 21:23
dafrito plobsing: Looks like they're old. The ones I've found come from when the PIR book was first added, June '09 21:27
I think they're footnotes, actually 21:32
Coke It's PsuedoPOD. 21:33
a special pod dialect used for writing books.
tcurtis From perldoc perlpod: '"Z<>" ‐‐ a null (zero‐effect) formatting code' 21:34
dafrito looks up PsuedoPOD 21:36
Yep, they're footnotes from PseudoPOD 21:39
chromatic Footnotes are N<>; Z<> are crossreference targets. 21:48
sorear PseudoPOD is nearly identical to POD 6, fwiw 21:53
which is the dialect parsed by NQP-rx
tcurtis Which of "to be ran" or "to be run" is the more correct way to say WIM in English? 21:56
sorear to be run if you want to talk about the process 21:58
to be ran if the state-of-having-been-run is important
tcurtis I would say that the register method "registers a pass to be run", then? 22:00
dafrito Should I write up a trac ticket and/or look into writing a patch to support these notes? 22:03
sorear tcurtis: yes 22:06
tcurtis sorear++ thanks. 22:07
dalek ee-optimization: dda85ef | tcurtis++ | docs/Tree/Optimizer.pod:
Start adding Tree::Optimizer docs.
 22:12
22:20 Andy joined 22:40 Andy joined
dalek TT #1743 created by dafrito++: N formatting code is unsupported in HTML output 22:46
TT #1743: trac.parrot.org/parrot/ticket/1743
rrot: r48494 | petdance++ | trunk/src/pmc/packfiledirectory.pmc:
add a couple of constants
 23:00
23:05 krunen joined
dalek rrot: r48495 | mikehh++ | trunk/include/parrot/pobj.h:
fix codetest failure - trailing whitespace
 23:50

Powered by Cro and the Raku® Programming Language.

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

Thank you!