|
Parrot 4.4.0 "Banana Fanna Fo Ferret" | parrot.org | Log: irclog.perlgeek.de/parrot | #parrotsketch meeting Tuesday 19:30 UTC Set by moderator on 15 May 2012. |
|||
| Tene | evening, at least. | 00:16 | |
| sorear | o/ Tene | 00:17 | |
| Tene | o/ sorear | ||
| whiteknight | hello Tene, sorear | 00:21 | |
| kid51 | Has everybody upgraded to Perl 5.16? :-) | ||
| whiteknight | nope | 00:25 | |
|
00:30
wagle joined
|
|||
| benabik uses system perl. | 00:55 | ||
| dalek | rrot: f7fec05 | Whiteknight++ | / (2 files): Add a new Parrot_api_set_stdhandle routine to the embedding API to set a custom IO-alike PMC as one of the standard handles for the interp |
00:57 | |
| rrot: da52d96 | Whiteknight++ | src/embed/api.c: Some improved docs for the new function |
|||
| rrot: 3d0c5f7 | Whiteknight++ | src/string/spf_ (2 files): A few codestd fixes from the sprintf cleanup work |
|||
| Heuristic branch merge: pushed 186 commits to parrot/whiteknight/gh_663 by Whiteknight | |||
| whiteknight | codetest is still telling me that src/string/spf_private.h doesn't have a coda | 00:58 | |
| bacek_at_work | aloha, humans | 01:00 | |
| whiteknight | hello bacek | 01:02 | |
| dalek | rrot: 727e515 | Whiteknight++ | src/ (2 files): Remove trailing whitespace |
||
| bacek_at_work | hello whiteknight | ||
| dalek | rrot: 68fed26 | bacek++ | src/string/spf_private.h: Update code to make codetest happy |
01:05 | |
| bacek_at_work | whiteknight, fixed :) | ||
| whiteknight | dafaq? | ||
| bacek_at_work | - * vim: expandtab shiftwidth=4: | ||
| + * vim: expandtab shiftwidth=4 cinoptions='\\:2=2' : | |||
| in the nutshell :) | |||
| whiteknight | I copied that coda from another .h file. | ||
| I copied it exactly from a file that passes | 01:06 | ||
| bacek_at_work | heh | ||
| our code test is bit fragile, isn't it? | 01:07 | ||
| whiteknight | I don't care about fragile. I wish it would output the expected value so I can just copy+paste whatever the test wants | 01:10 | |
| or, better yet, if the test could edit the file automatically, that would be great | |||
|
02:07
kid51 joined
02:51
japhb joined
03:40
alester_ joined
04:10
davidfetter joined
04:44
kurahaupo joined
05:36
kurahaupo joined
06:00
contingencyplan joined
06:13
brrt joined
06:24
benabik joined
06:44
fperrad joined
07:41
contingencyplan joined
07:47
brrt joined
08:13
brambles joined
08:17
lucian joined
08:44
kjs joined
08:54
lucian joined
09:28
kjs joined
10:19
schmooster joined
10:41
xdd joined
11:18
JimmyZ joined
11:58
benabik joined
12:01
whiteknight joined
|
|||
| whiteknight | good morning, #parrot | 12:01 | |
| moritz | \\o whiteknight | ||
|
12:03
bluescreen joined
|
|||
| whiteknight | hello moritz | 12:03 | |
|
12:05
brrt joined
|
|||
| benabik | ~~ | 12:09 | |
| whiteknight | good morning, benabik | 12:10 | |
|
12:14
JimmyZ_ joined
12:19
brrt joined
12:31
kjs joined
|
|||
| whiteknight | msg brrt I added a new Parrot_api_set_stdhandle to override one of the interpreters standard IO PMCs. For now, use a StringHandle to capture output. When the interp exits you can read all output from there. This prevents streaming, but should get you moving forward until we come up with something better | 12:43 | |
| aloha | OK. I'll deliver the message. | ||
| whiteknight | I've got over 20GB of things in my recycle bin, I think it's time to clean that out | 13:01 | |
|
13:04
PacoAir joined
|
|||
| brrt | whiteknight, i saw, awesome! | 13:04 | |
| benabik | In my department we have Mac, Ubuntu, and Windows machines. I've seen a number of people hit their quota because they never cleared out one trash or another. | 13:05 | |
|
13:07
Psyche^ joined,
bluescreen joined
|
|||
| whiteknight | brrt: for a "real" streaming solution we're either going to need a custom PMC type or NCI bindings. I'd much prefer you to get the infrastructure for the project set up first, and get basic examples working before we worry about performance or advanced features | 13:09 | |
| also, that new API is not tested at all, so use at your own risk (unless you want to add a quick test) | 13:10 | ||
| brrt | whiteknight: i imagined | 13:14 | |
| can you also do this for the input handle? | |||
| s/you/i or any other person/ | 13:15 | ||
| whiteknight | yes | ||
| brrt | nice | ||
| moritz | ooh | ||
| whiteknight | standard input 0, standard output = 1, and standard error = 2 | ||
| moritz | currently a say() call in rakudo looks up the dynamic $*OUT variable | ||
| I wonder if we can optimize that away, and instead switch to a different standard handle when a new $*OUT is declared | 13:16 | ||
| whiteknight | moritz: There is a way to override the standard output handle in the interp, if you want to avoid the lookup | ||
| brrt | well, then, first steps is going be a per-request parrot instance | ||
| with stringhandles for io | |||
| whiteknight | Yes. Rosella uses that exact mechanism in its test harness and in a few other places | ||
| moritz | otoh that optimization might be premature | ||
| whiteknight | moritz: yeah, it actually might be. Methods on the handle might get called in a nested runloop | 13:17 | |
| Here's the code Rosella uses, for your reference: github.com/Whiteknight/Rosella/blo...winxed#L10 | 13:18 | ||
| It might actually be good to assign $*OUT there too, to make sure to catch output from libraries and non-Perl6 extensions. But that's a small issue | 13:19 | ||
| brrt: I'm planning a pretty big cleanup of the IO system soon, so if you have any particular requests write them down somewhere | 13:20 | ||
| brrt | ehm, yes, but i'm honestly not at the stage that i know all the details | ||
| whiteknight | that's fine. My work won't start for a few days and will take a while to get right | ||
| so if you're hacking and think "it would be neat if..." tell me | 13:21 | ||
| brrt | will do, but i'll probably be busy just getting it running | ||
| youve seen the mod_parrot repository yet? | |||
| whiteknight | that's fine too. | ||
| yes, I've seen it. I haven't tried to play with anything yet, I figure you'll tell me when it's ready to be played with | 13:22 | ||
| brrt | it has no code right now, mod_bart has some | ||
| but all that is testing stuff | |||
| davidfetter wonders whether there's a mod_lisa | |||
| brrt | although i did manage to get a debug httpd instance running | 13:23 | |
| brrt slowly got the simpsons reference | |||
| you should try mod_bart/start_debug_apache.sh, although i'm not sure if it will work on anything not entirely standard | 13:24 | ||
| s/standard/fedora/ | |||
| whiteknight | I can give that a try at some point. I'm on a vanilla ubuntu install, which I'm sure is supported | 13:37 | |
| or, supported without too much hassle | 13:38 | ||
| brrt | you'll need a package like apache2-dev | 13:43 | |
| to get apxs | |||
| and i know debian calls apxs apxs2 | 13:44 | ||
| whiteknight | great | 13:49 | |
|
14:04
n0den1te joined
15:23
contingencyplan joined
15:27
slavorg joined
|
|||
| kjs | cotto: ping | 15:50 | |
|
15:51
Justin joined
|
|||
| Justin | good afternoon | 15:51 | |
|
15:54
d4l3k_ joined,
whiteknight_ joined,
perlite_ joined,
justin1 joined,
particle1 joined,
dmalcolm joined
15:55
Themeruta joined,
particle1 joined,
justin1 joined,
perlite_ joined,
Justin joined,
contingencyplan joined,
bluescreen joined,
Patterner joined,
brrt joined,
JimmyZ joined,
benabik joined,
schmooster joined,
brambles joined,
fperrad joined,
davidfetter joined,
japhb joined,
wagle joined,
Hunger joined,
plobsing joined,
particle joined,
perlite joined,
alester joined,
Khisanth joined,
Woodi joined,
mdupont joined,
aloha joined,
bacek_at_work joined,
pmichaud joined,
mtk joined,
alvis joined,
dalek joined,
PerlJam joined,
Coke joined,
p6eval joined,
pjcj joined,
eternaleye joined,
atrodo joined,
Maddingue joined,
simcop2387 joined,
janus joined,
elmex joined,
sorear joined,
GeJ joined,
jsut joined,
smash joined,
dngor joined,
Tene joined,
arnsholt joined,
rblackwe joined,
Kovensky joined,
NotFound joined,
wendar joined,
knewt joined,
Timbus joined,
cxreg joined,
szbalint joined,
cotto joined,
TonyC joined,
nopaste joined,
jtpalmer joined,
autark joined,
Infinoid joined,
jlaire joined,
ttbot joined,
sri joined,
awwaiid joined,
wagle_ joined,
Maddingu1 joined,
alester_ joined,
benabik_ joined,
preflex_ joined,
eternaleye joined,
JimmyZ_ joined
15:56
brrt1 joined,
Timbus joined,
_sri joined,
whiteknight__ joined,
szbalint joined,
TonyC joined,
janus joined,
nine joined
15:57
Hunger joined
16:02
eternaleye_ joined,
allison joined,
brrt joined,
whiteknight__ joined,
d4l3k_ joined,
preflex_ joined,
wagle joined,
kurahaupo joined,
nopaste joined,
Timbus joined,
tadzik joined,
alester joined,
jevin joined
16:03
sri joined
16:06
Maddingue joined
16:08
slavorg joined
16:50
PacoAir joined
17:44
lucian joined
17:57
kjs joined
|
|||
| cotto | kjs, pong | 18:17 | |
| kjs | hi cotto | 18:20 | |
| i had a few questions about the role of M0 and lorito. Perhaps you have a few mins during or after parrotsketch? | 18:21 | ||
|
18:22
crab2313 joined
|
|||
| cotto | now works. #ps isn't for another hour | 18:22 | |
| kjs | oh ok. in that case i will miss it. | 18:23 | |
| ok, some other time then. | |||
| cotto | I'm free most of today | 18:24 | |
| kjs | i'll be back an hour after #ps starts | 18:25 | |
|
18:43
particle joined
19:07
lucian joined
19:27
zby_home joined
19:34
whiteknight joined
|
|||
| whiteknight | good afternoon, #parrot | 19:37 | |
|
19:37
dmalcolm joined
|
|||
| benabik | o/ whiteknight | 19:38 | |
|
19:44
brrt joined
|
|||
| whiteknight | hello benabik | 19:48 | |
| almost done your GSOC project yet? | |||
| :) | |||
|
19:49
dmalcolm joined
|
|||
| benabik | :-P | 19:49 | |
| ... Did I just miss #ps? | 20:29 | ||
| Did everyone just miss #ps? | |||
| NotFound | Looks like we all missed it. | ||
| cotto | d'oh | 20:31 | |
| my alarm went off 45 minutes ahead. I made a note of it, continued hacking and forgot about it completely. | 20:32 | ||
| whiteknight | blah | 20:33 | |
| kjs | can we do it now? | 20:34 | |
| whiteknight | how hard would it be to setup aloha to make a countdown message about it? | ||
| "#ps meeting in 10 minutes, scrawny humans" | |||
| benabik | Aloha's just a perl bot. | 20:35 | |
| Shouldn't be too hard. | |||
| cotto | just fork, hack and ask bacek++ to pull | 20:36 | |
| kjs | is it an idea to hold #ps now? | ||
|
20:36
brrt1 joined
|
|||
| kjs | since there are several people here.. | 20:36 | |
| whiteknight | I'm fine with that | ||
| cotto | +1 | 20:37 | |
| brrt | stupid apxs | 20:58 | |
| Coke | kjs: using C worked so well attracting core developers, we now have... how many that hack on C guts on a regular basis? (I don't think dan's assumption there has stood the test of time.) | 21:07 | |
| whiteknight | our current number of core developers has very little to do with choice of language, I think | 21:08 | |
| switching over to C++ isn't going to gain us a dozen new helpers or more | |||
| kjs | Coke: you're right but implementing it in say, Eiffel, wouldn't really help :-) | 21:09 | |
| whiteknight | we've lost several core developers for a variety of reasons recently, and there's nothing really we could do about it | 21:12 | |
| cotto | nopaste? | 21:20 | |
| aloha, nopaste? | |||
| aloha | cotto: nopaste is is nopaste.snit.ch (works with the script in $_PARROT/tools/dev/nopaste.pl) | 21:21 | |
| kjs | that's it | ||
| thanks | |||
| nopaste | "kjs" at 89.101.178.50 pasted "problem with running m0 assembler" (2 lines) at nopaste.snit.ch/142354 | ||
| cotto | kjs, istr assuming perl >= 5.12.0 for the m0 assembler | 21:23 | |
| kjs | i got 5.10 | ||
| cotto | yup | 21:24 | |
| kjs | ok time to upgrade so | ||
|
22:37
lucian joined
23:07
particle joined
23:15
lucian joined
|
|||
| cotto | kjs++ | 23:58 | |