|
Parrot 2.9.1 Released | parrot.org | Log: irclog.perlgeek.de/parrot/today | Nopaste: nopaste.snit.ch:8001 | remove deprecations | volunteer for embedding or Lorito teams | Set by moderator on 26 October 2010. |
|||
|
00:02
simon_ joined
00:09
dngor left
00:27
theory left
|
|||
| cotto | plobsing, ping | 00:33 | |
| nopaste | "cotto" at 192.168.1.3 pasted "add op map output to pbc_dump -d" (30 lines) at nopaste.snit.ch/24974 | 00:38 | |
|
00:38
dmalcolm left
|
|||
| cotto | warnock means... | 00:55 | |
| commit | |||
| chromatic | Do it. | 00:56 | |
| bacek_at_work | cotto, grep -i warnock CREDITS :) | ||
| cotto | bacek_at_work, yes. We have a proud legacy of famous contributors. | 01:04 | |
| nopaste | "cotto" at 192.168.1.3 pasted "cleanup recommendations?" (31 lines) at nopaste.snit.ch/24975 | ||
| cotto | Hm. I guess that casting business at the beginning isn't too bad. | 01:05 | |
|
01:17
davidfetter left
|
|||
| dalek | rrot: r49705 | cotto++ | trunk/src/pbc_dump.c: [tools] add op map output to pbc_dump's disassembly output |
01:21 | |
| chromatic | Looks like a few places that could replace (&something)->something with something.something | ||
| but my eyes were already crossing before I read C code today | |||
|
01:33
kid51 joined
01:35
theory joined
01:36
theory left
|
|||
| kid51 | dukeleto ping | 01:43 | |
|
01:48
theory joined,
theory left
|
|||
| dalek | rrot: r49706 | cotto++ | trunk/src/pbc_dump.c: [tools] try to make code easier to follow |
01:52 | |
|
02:02
Khisanth left
|
|||
| dalek | rrot: r49707 | cotto++ | trunk/src/pbc_dump.c: [tools] add library version information to pbc_dump's disassembly ouput |
02:07 | |
|
02:08
Khisanth joined
|
|||
| kid51 | msg dukeleto Am recruiting a new smoke tester for Darwin/Intel (OS X 10.6). Is it currently possible to perform end-to-end smolder tests with 'git' -- or do I first have to train him on Subversion? | 02:13 | |
| aloha | OK. I'll deliver the message. | ||
| plobsing | cotto++ # mapping printing | 02:17 | |
|
02:22
kid51_ joined
02:26
kid51 left
|
|||
| kid51_ | msg nwellnhof remove_charset_ops branch: make smoke: PASS at r49701: linux/i386. See smolder.parrot.org/app/projects/rep...etails/825 | 02:27 | |
| aloha | OK. I'll deliver the message. | ||
| kid51_ | plobsing: In gsoc_nci branch, I had two failures in perl Configure.pl --test: t/steps/auto/libffi-01.t and t/steps/gen/libffi-01.t. Will try to examine further tomorrow. Non-critical. | 02:37 | |
| dalek | rrot: r49708 | plobsing++ | branches/gsoc_nci (2 files): kill dead code |
02:38 | |
| atrodo | chromatic: pong? | 02:41 | |
|
02:43
kid51_ left
|
|||
| chromatic | I'm here to encourage you to post Lorito implementation thoughts early and often! | 02:44 | |
| atrodo | Consider me encouraged! | ||
| cotto too | |||
| cotto wonders if plobsing's commit message is redundant | |||
| atrodo | I've been trying to, but things keep coming up. It's on the shortlist for sure | ||
| chromatic | It doesn't have to be complete; any thoughts you have now are valuable even if they're not finished. | 02:45 | |
| atrodo | actually, if i can figure out why MooseX::SemiAffordanceAccessor has stopped working when I moved to a new box, I'll try and do something tonight | 02:46 | |
| cotto | atrodo, +1 to what chromatic said. Even a couple 50-word posts somewhere would be great. | ||
|
03:13
dngor joined
03:22
patspam left
|
|||
| atrodo | chromatic, cotto> gist.github.com/650578 | 03:41 | |
| cotto | atrodo, by vmethods do you mean VTABLE functions? | 03:49 | |
| atrodo | yes | ||
| chromatic | I think of "goto" as "change the pc to something other than the next opcode". | 03:50 | |
| cotto | Thanks for the thoughts. My brain's a bit mushy at the moment, but you raise some good questions in need of answers. | ||
|
03:51
bluescreen left
|
|||
| atrodo | chromatic> right. if you only have a pc that can change to a new block, that means all of your ops are addressable in a flat segment | 03:52 | |
| chromatic | That depends on what the pc is, I suppose. | ||
| If it's merely an address, sure. | 03:53 | ||
| Though I do see your point, now that you make me think about it. | |||
| Yes there should be a difference between local goto and switch to new segment and goto a pc offset within it. | |||
| I suppose that any strong optimizer should flatten segments. | 03:54 | ||
| atrodo | call me old fashion, or crazy, but I liked the concept that the i386 affords the ability to use different segments in the address space | 03:55 | |
| but no one uses that concept any more. everyone has a flat segment address space | |||
| chromatic | Depends how much we want to relocate when we load more bytecode. | 03:56 | |
| If we never have to do runtime fixups.... | |||
| plobsing | I like the concept of more general chunk-of-memory PMCs. Right now arrays and hashes live in a sort of semi-pmc-limbo. | ||
| chromatic | I agree, with the caveat that it's very nice to be able to keep precision as to what's a PMC and what isn't. | 03:57 | |
| cotto | Is there a tradeoff for obviating runtime fixups by using segments? | 03:59 | |
| other than a lot of adding on the part of the cpu | |||
| dalek | rrot: r49709 | cotto++ | branches/opmap_aware_pmcs/src/pmc (2 files): [pmc] move get_pointer to PackfileBytecodeSegment, finish (untested) implementation |
04:56 | |
| rrot: r49710 | cotto++ | branches/opmap_aware_pmcs/src/pmc/packfilerawsegment.pmc: [pmc] minor cleanup in PackfileRawSegment; no functional changes |
04:57 | ||
| cotto | atrodo++ | ||
|
05:11
theory joined
05:30
theory left
|
|||
| chromatic | Assume that in the "run really fast" mode, we'll rewrite hot sections of code (including inlining calls) into flat segments. | 05:39 | |
| dukeleto assumes | 06:04 | ||
|
06:05
plobsing left
|
|||
| chromatic | Now clap your hands to bring Tinkerbell back to life. | 06:19 | |
|
06:38
cotto left,
silug left,
ascent left,
nopaste left,
PerlJam left,
Infinoid left,
autark left,
sECuRE left,
aloha left,
bacek_at_work left,
he left,
pmichaud left,
jnthn_ left,
moritz left,
Tene left,
Maddingue left,
dukeleto left,
szabgab left,
confound left,
integral left,
KatrinaTheLamia left,
lucian left,
particle left,
jsut_ left,
mikehh left,
raek left,
Hunger left,
dip left,
slavorgn left,
simcop2387 left,
GeJ left,
dzoe left,
Util left,
frodwith left,
ingy left,
atrodo left,
mikegrb left
06:55
chromatic left
06:58
wagle_ joined
07:00
wagle left
08:08
cotto joined,
fperrad joined,
lucian joined,
silug joined,
ascent joined,
nopaste joined,
particle joined,
PerlJam joined,
jsut_ joined,
mikehh joined,
raek joined,
Hunger joined,
dip joined,
Infinoid joined,
confound joined,
autark joined,
sECuRE joined,
moritz joined,
KatrinaTheLamia joined,
Tene joined,
pmichaud joined,
integral joined,
dukeleto joined,
Maddingue joined,
aloha joined,
bacek_at_work joined,
jnthn_ joined,
he joined,
szabgab joined,
slavorgn joined,
simcop2387 joined,
GeJ joined,
dzoe joined,
Util joined,
frodwith joined,
ingy joined,
atrodo joined,
mikegrb joined
|
|||
| cotto | 'night | 08:44 | |
| cotto is slightly puzzled by his new early bedtime of 0145 | |||
|
09:36
AzureSto_ joined
09:39
AzureStone left
09:57
contingencyplan left
11:14
dngor left,
dngor joined
11:24
dngor_ joined
11:26
dngor left
11:44
patspam joined,
patspam left
12:16
whiteknight joined
|
|||
| whiteknight | good morning, #parrot | 12:23 | |
|
12:24
smash joined
|
|||
| smash | hello everyone | 12:24 | |
| whiteknight | hello smash | 12:43 | |
|
12:52
mikehh left
12:59
dngor_ is now known as dngor
13:42
mikehh joined
|
|||
| dalek | rrot: r49711 | pmichaud++ | trunk/compilers/pct/src/PAST/Compiler.pir: Fix arguments to .lineof() call in PAST::Compiler. |
13:43 | |
|
13:47
plobsing joined
14:13
hudnix left
14:14
dmalcolm joined
14:16
hudnix joined
14:34
bluescreen joined,
whiteknight left
14:51
whiteknight joined
15:04
theory joined
15:53
brianwisti joined
|
|||
| whiteknight | msg dukeleto some parrot bindings and a parrot CLI client for libgit2 would be quite nice, methinks | 16:13 | |
| aloha | OK. I'll deliver the message. | ||
| whiteknight | I don't know who has the time or energy to write up those bindings, but it would still be nice | 16:27 | |
| cotto | ~~ | 16:35 | |
|
16:41
contingencyplan joined
|
|||
| dukeleto waves hello from SJC | 16:43 | ||
| whiteknight | SJC? | 16:44 | |
| dukeleto | whiteknight: San Jose Airport | 16:47 | |
| whiteknight: you don't have all the airport codes of the US memorized yet? | |||
| whiteknight: so you are interesting in libgit2+parrot? | 16:48 | ||
| interested, rather | |||
| dukeleto is severly sleep-deprived | |||
| aloha, msg kid51 smoking from git works fine | 16:49 | ||
| aloha | dukeleto: OK. I'll deliver the message. | ||
|
17:06
kurahaupo joined
17:07
mikehh left
|
|||
| whiteknight | dukeleto: yeah, bindings for libgit2 would be pretty sweet to have | 17:12 | |
| I don't know whether we would want to do NCI bindings or write like a PMC wrapper for it | 17:13 | ||
| cotto | How complete is libgit2? | 17:14 | |
| whiteknight | reasonably | 17:17 | |
| it's still under pretty active development, I think | 17:19 | ||
| but it does appear to be missing features | |||
| cotto | sounds promising. It made me sad to see that the original implementation of Git didn't bother using a shared library. | 17:26 | |
|
17:28
smash left
|
|||
| whiteknight | yeah, that is pretty weird considering the people who are writing it | 17:28 | |
| you'd figure there would be more care put into that kind of organization | |||
| atrodo | the original implemtnation of git was a series of low level commands. There was no plan | ||
| whiteknight | atrodo: that may be. But the people involved should be aware of how to make a proper library, and the importance of doing it correctly | 17:36 | |
| PerlJam | whiteknight: just like the people writing parrot, right? | 17:38 | |
| ;) | |||
| whiteknight | We do have a library. We have a libparrot. the API is lousy, but it's usable for everything that Parrot does | 17:48 | |
| and I'm not exactly Linus Torvalds | |||
| he probably wrote much of the damned library loader | |||
| atrodo | i've always wondered if the dyanmic linker was a glibc thing or a kernel thing | 17:57 | |
| whiteknight | blah. I can't understand why parrot-dev list gets so much hot lesbian xxx spam | 18:02 | |
| and healthcare provider spam | |||
| they've got us pegged as a group of horney, yet unhealthy voyeurs | 18:03 | ||
| cotto | atrodo, it's open source. I bet you could figure it out with a few minutes of digging. | ||
| whiteknight, is this something that gets filtered out before mere subscribers would see it? | 18:04 | ||
| atrodo | cotto> true. But i'm awfully lazy | 18:06 | |
| cotto | So you'll evaluate it when a result is needed. | 18:07 | |
| atrodo | exactly | 18:18 | |
| whiteknight | cotto: yeah, I'm filtering it all out | 18:19 | |
| I'm the only thing standing between you guys and hot, filthy, lewd lesbian XXX madness | 18:20 | ||
| thank me later | |||
| cotto | whiteknight++ | 18:25 | |
| How much of that stuff gets sent to parrot-dev? | 18:26 | ||
|
18:30
kurahaupo left
18:31
kurahaupo joined
|
|||
| whiteknight | a few per day | 18:38 | |
| well, to all the lists | |||
| cotto | ok. So it's nothing that needs parallelization atm. | 18:39 | |
| whiteknight | kid51 does it too | 18:42 | |
|
18:44
kurahaupo left
18:45
brianwisti left
18:49
kurahaupo joined
|
|||
| particle | parrot.org is again free of comment spam. | 19:10 | |
| cotto | particle++ | 19:12 | |
|
19:18
wagle_ is now known as wagle
19:48
fperrad left
20:03
kurahaupo left
20:09
tadzik joined
|
|||
| dalek | rrot: r49712 | cotto++ | branches/opmap_aware_pmcs/src/pmc/oplib.pmc: [pmc] use one method to get version informtion instead of 3 |
20:12 | |
|
20:15
kurahaupo joined
20:24
whiteknight left
|
|||
| dalek | rrot: r49713 | cotto++ | branches/opmap_aware_pmcs/src/pmc/packfilebytecodesegment.pmc: [pmc] implement push_pmc, use oplib's version method |
20:28 | |
| rrot: r49714 | cotto++ | branches/opmap_aware_pmcs/src/pmc/packfileopmap.pmc: [pmc] various cleanups and bugfixes, still doesn't work though |
|||
|
20:53
mikehh joined
|
|||
| GeJ | Bonjour everyone. | 20:56 | |
|
20:58
perlite_ joined
21:02
perlite left
21:24
bluescreen left
21:25
confound left
21:28
tadzik left,
confound joined
|
|||
| dalek | rrot: r49715 | NotFound++ | trunk/src/pmc/pmcproxy.pmc: use appropiate exception type in PMCProxy init_int |
21:29 | |
| rrot: r49716 | NotFound++ | trunk/t/pmc/pmcproxy.t: test PMCProxy init_int with invalid type |
|||
| GeJ | It is my understanding that nqp-rx is imported from external source. Is it ok if I commit a codingstd fix locally and report it upstream? Or should I wait that it gets applied there before including it in parrot? | 21:35 | |
| cotto | GeJ, it won't hurt anything, but it'll also get clobbered next time someone updates Parrot's copy of nqp in trunk. | 21:37 | |
| GeJ | cotto: I'll run a fulltest just to make sure I didn't botch anything. If successful, I'll commit the patch and report it. | 21:40 | |
| Tene | GeJ: with only a codingstd issue, it's probably fine. | 21:41 | |
| cotto | Tene, sure. Committing any other fix to Parrot's copy would be both difficult and ill-advised. | 21:42 | |
| Tene | cotto: We've received patch submissions to it, actually. :) | 21:43 | |
| iirc | |||
|
21:55
jsut joined,
jsut_ left
|
|||
| dalek | rrot: r49717 | geraud++ | trunk/ext/nqp-rx/src/stage0/HLL-s0.pir: [codingstd] Fix a couple of =over and =back issues in POD. |
22:00 | |
| rrot: r49718 | NotFound++ | trunk/t/pmc/fixedbooleanarray.t: ensure coverage of FixedBooleanArray destroy |
|||
| GeJ | msg pmichaud I commited a minor POD fix in r49717. I'll send a pull request on github later today (probably tomorrow for you) from home unless someone beats me to it. | 22:01 | |
| aloha | OK. I'll deliver the message. | ||
|
22:07
Hunger left
|
|||
| GeJ | aloha: clock? | 22:20 | |
| aloha | GeJ: GeJ: LAX: Thu, 15:20 PDT / CHI: Thu, 17:20 CDT / NYC: Thu, 18:20 EDT / UTC: Thu, 22:20 UTC / LON: Thu, 23:20 BST / BER: Fri, 00:20 CEST / TOK: Fri, 07:20 JST / SYD: Fri, 09:20 EST | ||
|
22:32
dngor_ joined
22:33
dngor left
|
|||
| bacek_at_work | aloha, humans | 22:54 | |
| cotto | good morning, bacek | 22:55 | |
| bacek_at_work | hio cotto | ||
| GeJ | G'Day bacek | 22:56 | |
| bacek_at_work | GeJ, G'Day | ||
|
23:02
kurahaupo left
23:04
kurahaupo_mobi joined,
dngor_ is now known as dngor
23:06
bluescreen joined
23:10
kid51 joined,
dmalcolm left
|
|||
| kid51 | ~~\\ | 23:12 | |
| GeJ | Good evening kid51. | 23:22 | |
| dukeleto | hola. | 23:23 | |
|
23:30
lucian_ joined
23:33
lucian left
23:37
kid51 is now known as kid51_at-dinner
23:43
cognominal joined
23:46
nwellnhof joined
23:48
kurahaupo joined
23:53
kurahaupo_mobi left
|
|||