Priority for this week: Review GSoC submissions | Post closed tickets in your report. | Note: This channel is for our Tuesday status meetings (at 20:30 UTC); you probably want #parrot instead. | irclog: irclog.perlgeek.de/
Set by moderator on 5 April 2011.
00:03 spinclad left 00:33 whiteknight left 01:10 spinclad joined 02:06 ShaneC left 06:39 contingencyplan left 06:46 contingencyplan joined 09:34 contingencyplan left 12:15 Coke joined 12:40 bluescreen joined 14:30 lucian joined 14:33 lucian_ joined 14:36 lucian left 15:42 lucian joined 15:46 lucian_ left 15:54 mikehh joined 16:07 contingencyplan joined 16:11 atrodo joined 17:16 lucian_ joined 17:19 lucian left 17:33 ShaneC joined 19:12 NotFound joined 19:33 benabik joined 19:38 kid51 joined
kid51 kid51's report 19:38
* DONE
** At bacek's request, opened, implemented, closed trac.parrot.org/parrot/ticket/2089 (llvm) 19:39
** Working on trac.parrot.org/parrot/ticket/2092 (llvm -- optin or optout?) 19:40
** Tested on various machines. Reported (and just now closed) trac.parrot.org/parrot/ticket/2091 19:42
** Commented on GSOC proposals.
* WILL DO
** Will be making Trac tickets for a number of test failures that I/we have been observing on some boxes but not others.
EOR 19:43
cotto_work *did: 19:54
-- helped soh_cah_toa get a patch submitted
-- started a rough outline for my State of Parrot talk at LinuxFestNW
* M0 roadmap progress:
-- thinking
- none
*eor
*will do: 19:55
- more M0 dynop prototyping, updating the spec as I go
*blockers:
- none
*eor (for real this time)
Util Report: Nothing done. MacBook died; Attempting to re-animate. 19:57
7-day ticket report:
3 closed: fixed
7 new
.end
NotFound What I did (last week forgot to report, this cover two weeks): 19:59
-parrot
* Testing
* Minor fixes
-winxed
* Several fixes, refactors, and code cleaning.
* Backwards incompatible change: call constructor in new with empty argument lis
(not the same as no argument list).
* New syntax for calling functions with multiple return.
* New syntax for declaring classes without defining them.
* Allow multiple declarations in 'using static' statement.
* Disallowed direct usages of stages >= 1 from command line,
update the non installable driver to work that way.
* Allow argument modifiers in stage 0.
* New statement 'switch without expresion'.
What I will do:
No plan
EOR
20:09 soh_cah_toa joined 20:11 lucian_ left 20:13 bluescreen left 20:16 plobsing joined, lucian joined 20:24 nwellnhof joined
mikehh What I did since my last report: 20:28
* building and testing parrot on amd64/i386, with gcc/g++
* some fixes
* testing Rakudo and Winxed on latest parrot
What I intend to do in the next week:
* testing and fixing
.eor
atrodo .done 20:30
* sadly, nothing because of $dayjob deadline
.todo
* alorito hacks
* m0 help if possible
* isparrotfastyet refactors
.eor
allison Got several new interested contributors to Parrot Debian packaging. Working on spinning that up into a more active Parrot packaging group for Debian.
cotto_work any more last-minute reports?
allison EOR
kid51 Good day/evening.
Util Hello 20:31
allison hi
soh_cah_toa hello
cotto_work hello
mikehh hi there
atrodo heyo
cotto_work How's we do this week?
GSoC prep seems to be in full swing. 20:32
soh_cah_toa good 20:33
NotFound Hola
cotto_work any thoughts on the past week? 20:34
We're one week from 3.3 with whiteknight as the release manager.
I believe he was planning an rc. 20:35
mikehh rc?
cotto_work release candidate
kid51 I believe whiteknight is working on a couple of bugs that showed up after imcc_compreg merge, particularly when run thru rakudo
We also have a couple of failing tests where we have to decide whether they're really failing or not. 20:36
atrodo Is rakudo still segfaulting?
cotto_work They appear to be triggered with gms, which won't be the default for 3.3
kid51 Other than that, I don't see anything big being merged into master before 3.3
cotto_work (of course, we should still fix them since Rakudo uses gms by default)
mikehh rakudo tested ok for me on i386 - I think there are problems with amd64 and without --optimize 20:37
kid51 Does that mean we have to 'switch back' from GC x to GC y this week ? And then switch again after release?
cotto_work It also seems to be related to the version of Perl used by Configure.pl
mikehh I am using 5.12.3 up to date with CPAN 20:38
on Ubuntu 11.04 beta
(which still has 5.10.1 20:39
20:39 bubaflub joined
cotto_work kid51: we're still on ms2 for default in master 20:39
NotFound What impact can have the version of perl? Default cflags?
cotto_work whiteknight and moritz were working on it this morning. I don't know if they found that out. 20:40
This isn't a release blocker (unless whiteknight decides otherwise), but it's worthwhile to make it a priority. 20:42
mikehh I am getting some test failures with t/examples/pod.t because of the chane in behavior of print after the merge 20:44
#change
cotto_work mikehh: can you file a bug about that?
mikehh have done TT #2088 I think 20:45
kid51 whiteknight will have to look at TT #2088; that was one of what I was referring to as "is it a bug or was it wrong before?" 20:46
cotto_work ok 20:48
any other thoughts?
questions? goals for the week? 20:49
mikehh it seems that print with :immediate and ;postcomp actually outputs, which it did not before (just compiled before)
kid51 Goal: Plan for Parrot Developer Summit weekend of Apr 30/May 01. Formulate roadmap goals (or start thinking about them)
cotto_work We need to make sure that HLLs get tested well
kid51: that's the weekend of LinuxFestNW. It might be tricky for dukeleto and me. 20:50
mikehh review GSOC proposals
bacek ~~ 20:51
kid51 cotto_work: We set this schedule in Dec 2010. If we start varying from it we'll never be consistent and PDSes will grow weaker. 20:52
cotto_work: Each and every weekend we choose, there will be a conflict for someone. 20:53
allison kid51: a one week delay to accomodate the schedules of the leads on the architecture and community management teams seems reasonable 20:55
kid51: especially since they'll be there putting in a good word for Parrot in a public forum 20:56
kid51 allison: We had a discussion about this at our PDS in December.
< 3 weeks before PDS is not a good time to bring up scheduling conflicts
cotto_work I would strongly prefer a delay. I suspect that talk prep will take most of my tuits as LinuxFestNW approaches, since it'll be the first time I'll be giving that talk. 20:57
allison LinuxFestNW isn't known for announcing dates early
cotto_work I found out I'd be talking a couple weeks ago.
if that
allison but, it is an important forum, gets quite a few significant speakers and attendees from the wider open source world
cotto_work Whenever the PDS happens, I'll do my best to be there. I can't promise anything if it's that weekend though. 20:58
allison I can guarantee I won't be there, as I'll be on a flight to Hungary, but I consider cotto and dukeleto more important 20:59
cotto_work kid51: I'll leave this to your discretion. My preference is clear. 21:01
In any event, we should start the scheduling process sooner than later to figure out the exact time. 21:03
kid51: do you mind sending out a doodle request in the next few days? 21:04
He must have been called away. Any other questions? 21:06
mikehh well whiteknight said we should discuss the issues brought up by TT #2088, but he is not here 21:07
cotto_work mikehh: any idea what he wanted to discuss about it? 21:08
mikehh changes in behavior in IMCC afdter the branch merge 21:09
after
print with :immediate and ;postcomp actually outputs, which it did not before (just compiled before) and possibly other issues 21:11
plobsing also mentioned something about that
plobsing yeah. imcc did funky stuff before. now it is consistent.
cotto_work plobsing: so this is something that should have been failing for a long time but only started after the merge? 21:12
mikehh I am not sure how to handle that change in t/examples/pod.t - it does not expect output from passing stuff to imcc 21:13
other than error output
plobsing cotto_work: basically, yes. the new behaviour is correct IMHO. 21:14
cotto_work Then it sounds like we should fix the tests.
mikehh at the moment the test passes snippets to imcc to check for correctness
cotto_work and hilarity results if those snippets output anything 21:15
plobsing so don't print in those snippets. do some math in stead or something. 21:16
mikehh the test loses its sequence (reported in the summary)
cotto_work wfm
That fix is straightforward. Any other questions? 21:17
mikehh dukeleto needs to look at g++ failures in t/src/extend_vtable.t 21:18
I tried a couple of additions - they did not work - need to try more 21:19
benabik I get failures there using just gcc. (Which I noted on TT#2084)
cotto_work any volunteers for TT #2088? 21:20
mikehh yeah - passes gcc --optimize for me, one failure withh gcc, all fail with g++
the simplest solution ATM is just to pull the snippet (one of 613 or something) but I will think on it 21:22
cotto_work mikehh I'm fixing it. 21:23
21:24 kid51 left
cotto_work Here are the goals I have: 21:24
GOAL 1: potential release blocker - --optimize --gc=gms segfaults Rakudo (whiteknight, all)
GOAL 2: HLL testing leading up to the 3.3 release (all)
GOAL 3: Formulate roadmap goals (team leads)
GOAL 4: Review GSoC proposals, help students as needed (gsoc mentors)
GOAL 5: fix g++ failures in extend_vtable tests (dukeleto)
any additions or other questions? 21:25
soh_cah_toa sounds good to me 21:26
mikehh we should have a bug-fest etc before the release 21:27
cotto_work mikehh: not a bad idea
mikehh usuall on the weekend
usually 21:28
cotto_work no reason not to start earlier
mikehh of course
cotto_work Is there a number of bugs you want to try to close? 21:29
21:30 plobsing left
cotto_work We're at 506 according to trac.parrot.org/parrot/report/1 21:30
mikehh let's try to get below 500 at least 21:31
cotto_work wfm. We can always go lower.
GOAL 6: bugfest this weekend - get below 500 open tickets
Let's call it a wrap. 21:32
21:39 lucian left 21:53 NotFound left 22:08 benabik left 22:13 nwellnhof left 22:28 whiteknight joined 22:35 bubaflub left 23:50 soh_cah_toa left