|
Post closed tickets in your report. | Note: This channel is for our weekly status meetings (Tuesdays at 19:30 UTC); you probably want #parrot instead. | irclog: irclog.perlgeek.de/ Set by moderator on 18 February 2014. |
|||
|
00:13
rurban1 joined
01:14
rurban1 joined
01:39
rurban1 joined
07:28
basiliscos joined
13:22
kshannon joined
13:42
bluescreen joined
14:26
rurban1 joined
16:42
Chirag joined
17:17
basiliscos joined
18:29
rurban1 joined
|
|||
| Util | Pre-report: | 19:30 | |
| # Done/Doing: RosettaCode matrix tasks; no commits yet. | |||
| .end | |||
|
19:30
rurban1 joined,
rurban joined
|
|||
| rurban | Done: added GSOC task 1 for Chirag (#1069) | 19:31 | |
| released 6.3 | |||
| Util | Hello! Who all is in attendance? | ||
| Chirag | Me :) | 19:32 | |
| Util | Welcome, Chirag ! | ||
| rurban | and me, I already started | ||
| Chirag | Thank you! | ||
| rurban | we do it like in a scrum standup. done, work+todo, blockers | 19:33 | |
| Util | Congrats to Chirag and rurban for GSoC acceptance | ||
| rurban | yes, we were lucky | ||
| Chirag | Thank you for accepting our proposal! | ||
| rurban | you have to thank the silent mass of perl devs who voted | 19:34 | |
| Chirag | How to reach them? google group? | ||
| rurban | I believe only mentors can vote | 19:35 | |
| and tpf admins | |||
| Util | I am handling the next Parrot release. | ||
| Am I correct that the GSoC work has 0 chance of merging into master branch in the next 3 weeks? | |||
| (So I don't have to worry about that) | |||
| rurban | well, maybe we can get the GC WB done. looks easy enough. let's say 50%. | 19:36 | |
| testcases are not easy. you really need to do code review and try parallel test cases (threaded ones) | |||
| but the current wrong code used to SEGV on me, so the current testcases seem to be okay | 19:37 | ||
| Util | OK | ||
| rurban | well, not really threaded code, more like GC torture tests | 19:38 | |
| like t/stress/gc.t | |||
| Util | Was a backup mentor ever assigned to this GSoC task? I requested it, but I see no evidence that I have been assigned that role. | ||
| rurban | t/examples/shootout.t used to crash on me | 19:39 | |
| you need to ask mark keating to add you | |||
| Util | Will do | ||
| rurban | got one feedback from rakudo, that 6.3 did not build for them. but this was not repro (auto/expect was missing) | 19:40 | |
| they are now starting parallelizing hyper ops, which would be a good point for nqp-p (this is were we would shine over moarvm) | 19:41 | ||
| Util | "auto/expect was missing" - I don't know what that meant. I did understand "not reproducable". | 19:42 | |
| rurban | they told me the error was related to auto/expect.pm was missing during configure | ||
| Util | Oh, OK | ||
| rurban | and they couldn't repro it (well, froggs only) | ||
| so it looks like a git pull problem to me, because auto/expect was added with 6.2 and there should be no issues | 19:43 | ||
| Util | Right | 19:44 | |
| rurban | will wait for issues when they are reporting it | ||
| that's all from my side. pretty busy with perl5 hashes and cpan stuff still. | 19:45 | ||
| smolders still horrible, and my buildbot fails on 50% of the archs due to arch specific setup problems | |||
| (ppc, mingw, solaris, bsd quirks) | 19:46 | ||
| but parrot is still better than nqp, rakudo or p5 in this regard | |||
| rperl with its stricter C++ and libstdc++ requirements found even more issues, which parrot already solved | 19:47 | ||
| Util | Any blockers for either of you two, or anything that I should liaison with the Perl 6 team? | 19:49 | |
| rurban: are you doing anything on rperl, besides smolder/buildbot? | |||
| rurban | I would want to know if they plan hyperops for nqp-p also or if we should do it | 19:50 | |
| I'm fixing buildproblems for rperl and fixing Inline issues, yes | |||
| Inline::Filter, Inline::C and Inline::CPP | |||
| My buildbot btw: perl514.cpanel.net/build/builders | 19:51 | ||
| Util | I will ask about the hyperops. | ||
|
19:52
rurban1 joined
|
|||
| Util | rurban++ # for helping his "competition" :^) | 19:52 | |
| rurban | And I was working a lot with symbolic annotations lately to work with klee, frama-c, and similar solver based backends. openssl, hashes, security stuff | 19:53 | |
| which is good to create testcases from failures, formally verifiy documented behavior, and so on | 19:55 | ||
| Util | Cool stuff | 19:56 | |
| rurban | and create meaningful fuzzing input (like taint input vars, and do dataflow analysis on them) | ||
| but I see no usecase for us yet. more with p5 (their hashes are broken, and they have no idea) | 19:57 | ||
| Crypt::SSLeay / Net::SSLeay is also a problem | |||
| Coke is reminded to bump his smolder replacement up the queue slightly. | 19:58 | ||
| Coke has waaaay too many plates in the air at the moment. | |||
| rurban | smolder is not that important. they blame their 32bit VMs. you need to kick it a few times to warm up | ||
| Util | Noooo! Juggle balls, not plates; plates break if you miss! :) | 19:59 | |
| Anything else before we adjourn? From anyone? | 20:03 | ||
| OK, then. Meeting adjourned. Thanks! | 20:07 | ||
|
23:24
rurban1 joined
23:26
rurban2 joined
|
|||