Parrot 2.11.0 Released | parrot.org | Log: irclog.perlgeek.de/parrot/today | Onward and upward with Google Code-In | Please test rakudo with bleeding edge parrot! | merge html_cleanup and embed_api2
Set by moderator on 21 December 2010.
mikehh Kapace_: you might want to look at search.cpan.org/~jesse/perl-5.12.2/...odspec.pod 00:08
and maybe follow some of the links there
Kapace_ ok thanks mikehh 00:09
00:18 hudnix joined 01:26 sECuRE joined 01:28 davidfetter left 01:29 whiteknight joined 01:36 sECuRE left
Kapace_ whiteknight: what do you think of gist.github.com/750697 for the make cover docs 01:40
whiteknight let me look
01:50 nwellnhof joined
whiteknight Kapace_: Looks good. There is a new "make quickcover" tool which runs fewer tests but is much faster. Can you mention that? 01:57
01:58 Nol888 joined
Kapace_ whiteknight: should I ask kid51 for more info? or just mention that ".. it's faster."? 02:00
whiteknight it is faster but it is not as complete
it doesn't run tests in other runcores
Kapace_ ok, btw, www.google-melange.com/gci/task/sho...9243786855 02:01
02:03 Yuki`N left
whiteknight approved 02:11
Kapace_ how do I generate html from the pod?
cotto_work pod2html 02:12
Kapace_ what about the fancy templates? (like on docs.parrot.org?) 02:13
dalek Some 'nwellnhof' person just gave me a 122190 byte commit packet. They need to be more careful in the future. 02:14
nwellnhof dalek: shut up 02:15
Kapace_ ahaha
sorear it works! 02:19
cotto_work sorear: feel like sharing? 02:20
sorear cotto_work: dalek's giant branch merge filter 02:21
nwellnhof but what about that 'need to be more careful' message? 02:25
02:27 kennym left
whiteknight I'm heading to bed. Kapace_: I'll take a look at your doc first thing in the morning. 02:28
Kapace_: unless you find somebody else to evaluate it before then
Kapace_ whiteknight: should I just submit a link to the gist, or add it to a branch etc? 02:29
ok
whiteknight Kapace_: a link to the gist should be fine
Kapace_ awesome good night whiteknight 02:30
whiteknight you wont get attribution for the commit, if that matters for you
goodnight
02:30 whiteknight left
Kapace_ don't really care about attribution.. maybe karma.. 02:30
aloha: karma Kapace
aloha Kapace_: Kapace has karma of 3.
nwellnhof Two of my branches could need some testing: nwellnhof/unicode_io and nwellnhof/string_unescape 02:32
Nol888 msg whiteknight did you ever get any news on any of those tasks that you said I might be interested in? 02:35
aloha OK. I'll deliver the message.
02:37 Nol888 left 03:06 bluescreen left 03:16 bluescreen joined 03:26 Kristabou left 03:32 nwellnhof left 03:58 Khisanth left 04:02 Khisanth joined
cotto ~~ 04:19
04:50 dmalcolm left 04:55 Themeruta joined, NotFound left 05:32 rurban_ joined 05:34 rurban left 05:35 rurban_ is now known as rurban 05:36 sECuRE joined 05:39 theory_ joined, theory left, theory_ is now known as theory 05:50 theory left 05:55 theory joined 06:02 theory left 06:03 theory joined 06:42 dd070 joined 06:45 dd070 left 07:13 theory left 07:42 cognominal left 08:01 fperrad joined
cotto msg fbrito task approved: go for it 09:12
aloha OK. I'll deliver the message.
09:21 rfw left
dalek Some 'mikehh' person just gave me a 41925 byte commit packet. They need to be more careful in the future. 09:27
mikehh dalek needs to handle branch merges, they tend to be fairly large - my previous one was much bigger 09:35
sorear the POST hook just hands off a block of json to dalek 09:36
there's a heuristic in place to drop large blocks
maybe the message could be changed - at the time I added that I was thinking more in terms of "AGGH I JUST REBASED MASTER BY MISTAKE"
mikehh sorear: I think I also saw the same message when nwellnhof did a push recently 09:48
sorear yes
mikehh sorear: the previous time when I brought the branch up to date the commit was over 500000 bytes 09:49
and every time master is merged into a branch it brings in all the commits from master 09:51
which can be a whole lot, not sure what happens the other way - merging into master 09:52
cotto The nqp-rx nom code is strange. I wonder how it'll work in straight M0 without PMCs to help bootstrap the process. 09:53
dalek rrot/html_cleanup: 769c83d | mikehh++ | docs/index/developer.json:
[html_cleanup] remove deleted file from list
 09:56
10:03 cognominal joined 10:16 kennym joined 12:09 bluescreen left, bluescreen joined 12:12 bluescreen_ joined 12:16 bluescreen left 12:19 contingencyplan left 12:48 jan left 12:55 mtk joined 13:06 Kovensky left 13:15 Kovensky joined 13:22 whiteknight joined
whiteknight good morning, #parrot 13:23
13:25 bluescreen_ left 13:32 rurban_ joined 13:34 rurban left, rurban_ is now known as rurban 13:36 bluescreen_ joined
dalek rrot/embed_api4: 0597887 | Whiteknight++ | docs/pdds/draft/pdd10_embedding.pod:
finish updating PDD10. If we can get architectural approval, I'll move this out of draft before the merge
 14:05
whiteknight msg cotto When you get a chance could you look at github.com/parrot/parrot/blob/embe...edding.pod . The style of this draft is a little different than other PDDs, a little more abstract and a little more explanatory, though it's probably still short on a few details 14:06
aloha OK. I'll deliver the message.
dalek rrot/embed_api4: 6465f9d | Whiteknight++ | docs/pdds/draft/pdd10_embedding.pod:
expand a little bit on the differences between extending and embedding
 14:11
rrot/embed_api4: b3baa14 | Whiteknight++ | docs/pdds/draft/pdd10_embedding.pod:
POD fix
smash hello everyone 14:22
dalek rrot/embed_api4: f974844 | Whiteknight++ | docs/pdds/draft/pdd10_embedding.pod:
fixes
 14:24
tadzik o/ 14:27
whiteknight good morning smash, tadzik 14:28
dalek rrot/embed_api4: e4a31d1 | Whiteknight++ | docs/pdds/draft/pdd10_embedding.pod:
mention the second type of string API function
 14:30
tadzik + Just In Time native call frame generation using LibFFI -- what does this actually mean? 14:31
atrodo Magic i believe 14:32
whiteknight tadzik: are you familiar with call frames at all? 14:35
or, calling conventions at the assembly level?
tadzik not really, what can I read? 14:38
tadzik lurks call stack @ wikipedia
ah, that seams clear now 14:40
14:57 Kristaba joined 15:01 Themeruta is now known as NotFound
dalek rrot: fc39f47 | (Gerd Pokorra)++ | ports/fedora/2. (4 files):
update to current fedora-ports
 15:07
15:08 dmalcolm joined
dalek rrot: 1865326 | Whiteknight++ | docs/dev/coverage.pod:
add documentation about make cover and make quickcover from Kristaba++ for GCI
 15:13
rrot: bc0a7d2 | (Gerd Pokorra)++ | docs/dev/coverage.pod:
Merge branch 'master' of github.com:parrot/parrot
 15:17
15:27 mtk0 joined 15:28 mtk0 left, mtk left 15:29 mtk joined
dalek rrot: 783a3d9 | (Gerd Pokorra)++ | README. (3 files):
change version number in README.<language> files
 15:34
15:38 bluescreen__ joined 15:41 bluescreen_ left 16:36 bluescreen__ is now known as bluescreen 16:54 theory joined
bluescreen hi, master is not compiling for me.. anybody knows how to fix: "src/runcore/cores.c:353: error: �Parrot_DynOp_core_2_10_1� undeclared (first use in this function)" 16:58
tadzik make realclean? 17:01
bluescreen let me see
tadzik: it did it.. thanks 17:03
tadzik yw
17:03 ok22 joined 17:22 Andy left 17:56 nwellnhof joined
dalek rrot: db189b2 | nwellnhof++ | src/gc/gc_ms2.c:
[gc] Don't initialize string GC twice

Fixes a memory leak when creating a second interpreter.		 18:00
18:07 nwellnhof left 18:10 macroz joined 18:32 whiteknight left 18:36 contingencyplan joined
dalek rrot: 5eef2c2 | Coke++ | MANIFEST:
re-run tools/dev/mk_manifest_and_skip.pl
 18:42
18:53 davidfetter joined 18:57 kennym left 18:58 bacek left 19:02 nwellnhof joined
nwellnhof bluescreen: ping 19:03
19:03 rfw joined
bluescreen pong 19:03
nwellnhof: looking for me? 19:04
nwellnhof bluescreen: i' wondering about TT #1765. it's strange that your patch had any effect.
bluescreen it think valgrind report in the first place is wrong
there is no memory leak
19:05 hercynium joined
bluescreen but it reports one as you have a branch in which you're not using a recently allocated ptr ( even though ptr is null ) 19:05
nwellnhof when i tried parrot trunk, i didn't get the original memory leak but another one that i fixed in db189b22.
bluescreen I've updated my copy today
also tried stashing my changes and got the same thing 19:06
meaning valgrind reporting a memory leak
the good thing is that I don't think we have a real memory leak
nwellnhof do you know where exactly the leak was reported? in allocate_interpreter?
bluescreen its just wrong report... which version of valgrind do you have? 19:07
nwellnhof i'm on another machine right now. it was the version from ubuntu 10.10.
bluescreen oh ok... I'm in 10.10 as well
I'm going to try to create a minimal c file that shows what I'm talking about 19:08
nwellnhof are you on 32bit or 64bit? 19:09
bluescreen 64 19:10
and with gcc btw
nwellnhof i'm on 32bit. that might explain the differences. 19:11
bluescreen when you get to your original computer can you try again 19:12
in anycase, we might close tt1765 one real soon :D
nwellnhof it seems that the original memory leak in allocate_interpreter was fixed by another change. 19:13
bluescreen possible... 19:14
nwellnhof anyway, t/op/interp_2.pir doesn't leak for me with the fix in db189b22. but that was another issue introduced by the GC changes. 19:15
dalek rrot/nwellnhof/string_unescape: 6541fd3 | nwellnhof++ | / (2 files):
[str] Remove deprecated string_* functions

See TT #1809		 19:23
19:23 Patterner left
19:27 Psyche^ joined, Psyche^ is now known as Patterner
dalek rrot: 8e59a00 | NotFound++ | src/pmc/exporter.pmc:
use appropiate exception types in Expporter misuses
 19:41
19:46 davidfetter_ joined 19:49 davidfetter left, davidfetter_ is now known as davidfetter 19:50 bluescreen left 19:57 mtk left 19:59 whiteknight joined, mtk joined 20:00 davidfetter_ joined 20:03 davidfetter left 20:04 davidfetter_ is now known as davidfetter 20:08 davidfetter_ joined
dalek rrot: a0cf037 | NotFound++ | t/pmc/exporter.t:
more test for Exporter PMC
 20:09
20:11 davidfetter left 20:12 davidfetter_ is now known as davidfetter
NotFound Someone want to create a task for GCI? Convert exporter.t to pir. 20:12
20:13 davidfetter_ joined
cotto_work NotFound: what's the benefit? I think we're deemphasizing converting perl tests to PIR if there's not a better reason than moving away from Perl. 20:14
NotFound Sigh 20:15
cotto_work actually, that one looks very straightforward. 20:17
20:18 davidfetter left, davidfetter joined
NotFound NM, it was just to provide a task 20:18
dalek rrot/nwellnhof/unicode_io: e60e006 | nwellnhof++ | src/io/api.c:
[io] Don't convert strings to binary in Parrot_io_putps
 20:22
20:23 davidfetter_ left
cotto_work NotFound: after thinking about it a bit, converting test isn't a bad way to get familiar with the code. For tests like that one where the conversion is straightforward, it'll give us more consistency in our test suite. 20:24
Converting tests to perl isn't an end in itself, but consistency is good and getting people familiar with the code is better. +1
NotFound :) 20:25
20:31 perlite left 20:46 kennym joined 20:48 bluescreen joined
dalek rrot: 751bc2c | cotto++ | config/auto/coverage.pm:
[configure] make coverage slightly smarter on systems without Devel::Cover
 20:53
rrot: c161e5c | cotto++ | config/ (2 files):
make coverage tests fail quickly if Devel::Cover isn't installed
20:54 ok22 left
cotto_work review and thoughts on those last two commits is appreciated 20:54
20:56 Coke joined
Coke msg cotto does c161e5c9265bff2d8b8cb74545c2b81f43e686e2 work on windows? 20:56
aloha OK. I'll deliver the message.
nopaste "bluescreen" at 192.168.1.3 pasted "Valgrind "definitelly lost"" (22 lines) at nopaste.snit.ch/27381 20:57
dalek TT #1893 closed by jkeenan++: Define a 'quickcover' make target 20:58
TT #1893: trac.parrot.org/parrot/ticket/1893
rrot: 93971ee | cotto++ | config/auto/coverage.pm:
[configure] avoid && where evitable
 21:01
cotto_work Coke: good catch. echo and exit are fine but windows doesn't seem to like &&
21:06 davidfetter_ joined 21:07 Andy joined, Andy left 21:09 davidfetter left
Coke we know they have perl. I recommend a quick-and-dirty .pl in tools 21:11
21:12 davidfetter_ left
cotto_work Why not the quicker and dirtier echo/exit? 21:12
Coke (and no need for the exit there - just declare the COVER code to invoke the new .pl unconditionally if we know they're missing the modules.)
why use the exit at all? 21:13
so, make "make cover" just run your echo if the module is missing, then.
dalek TT #1899 created by bluescreen++: Memory leak in compilers/imcc/main.c 21:14
TT #1899: trac.parrot.org/parrot/ticket/1899
TT #1765 closed by nwellnhof++: Memory leak in allocate_interpreter
TT #1765: trac.parrot.org/parrot/ticket/1765
TT #1900 created by jkeenan++: Change names of 'make' coverage targets
TT #1900: trac.parrot.org/parrot/ticket/1900
whiteknight Coke: In the embed_api4 branch, I rewrote PDD10 to talk about the new embedding API, and point out differences with extending situations 21:19
Coke: Could you give that a read-over when you get a chance and make sure it covers all the things you want it to? 21:20
21:20 mtk left
Coke I neither embed nor extend. 21:20
So, does that still matter? 21:21
I'd say ask Jeff to review it, but I think he's long gone from parrot.
dalek rrot: b479f4f | bluescreen10++ | compilers/imcc/main.c:
TT #1899: Fix memory leak in compilers/imcc/main.c
 21:22
rrot: 7fe94d2 | bluescreen10++ | config/auto/coverage.pm:
Merge branch 'master' of github.com/parrot/parrot
whiteknight It's not a big deal, I just wanted to make sure you were included in the process
you had been in the conversation yesterday when we were talking about the line between extending/embedding being blurry historically 21:23
bluescreen: We're likely going to have similar problems in the new API. I don't think main.c calls imc_cleanup at all 21:24
bluescreen do you want me to take a look in the embed_api.c branch? 21:25
embed_api4*
whiteknight bluescreen: maybe. We're probably going to want to valgrind that branch, and run coverage on it before we emrge
merge
bluescreen ok... we can run valgrind over the test cases 21:26
whiteknight We don't need 100% coverage, but if the coverage numbers are low we can assign that to GCI students
okay, awesome
bluescreen let me know as soon as they are ready play with
whiteknight bluescreen: embed_api4 is ready now
whenever you feel like doing it 21:27
bluescreen ok 21:29
whiteknight is heading home now. 21:30
Coke whiteknight: my problem has been: I use something not the in API because there is nothing IN the api to use.
whiteknight Coke: Okay. I'm heading home now. We should talk about what the solution to that problem can be going forward
you around to chat tonight/tomorrow? 21:31
Coke My takeaway from the conversation yesterday was to ask for blessed functions in the future.
dalek TT #1899 closed by bluescreen++: Memory leak in compilers/imcc/main.c
TT #1899: trac.parrot.org/parrot/ticket/1899
whiteknight That's my preferred solution, but I want to make sure we have a process in place that works for you
Coke once I get partcl working again after the last (@#&$ borkage, I'll see if there's anything else I'm using that isn't official.
whiteknight okay, really have to go now. Later
21:31 whiteknight left
Coke ... but I have no impetus to do that, as there are several outstanding bugs still. 21:32
later.
21:32 rurban_ joined
dalek rrot/embed_api4: 33cfd36 | cotto++ | docs/pdds/draft/pdd10_embedding.pod:
apostrophix and spelling
 21:33
21:34 rurban left 21:35 rurban_ is now known as rurban
Coke what commit fixed #1899? 21:36
cotto_work bluescreen: ^ 21:41
Coke is sad that whatever the commit was didn't automatically update the ticket. 21:42
cotto_work I could make that happen if enough people want it. 21:43
It does sound delightfully lazy. 21:44
dalek rrot: 5d2a92f | fperrad++ | / (2 files):
[harness.pir] add property 'Configure args'
 21:48
21:57 fperrad left 22:03 particle left 22:40 particle joined 22:49 nwellnhof left, hercynium left
cotto_work Coke: how useful is it to have commits automatically update tickets? Is that something you do at $dayjob? 22:59
23:09 dmalcolm left, dmalcolm joined 23:26 Yuki`N joined 23:41 theory left, theory joined 23:42 dmalcolm left 23:46 smash left 23:57 whiteknight joined
dalek tracwiki: v6 | cotto++ | Platforms/Windows 23:59
tracwiki: trac.parrot.org/parrot/wiki/Platfor...ction=diff
Yuki`N whiteknight, any news on those gci tasks?

Powered by Cro and the Raku® Programming Language.

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

Thank you!