Perl 6 - the future is here, just unevenly distributed

IRC log for #pr-challenge, 2015-03-28

| Channels | #pr-challenge index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:17 vroom joined #pr-challenge
00:22 tgy joined #pr-challenge
01:04 BRAD_ joined #pr-challenge
01:36 absolut_todd joined #pr-challenge
02:51 veryrusty joined #pr-challenge
03:51 absolut_todd joined #pr-challenge
05:21 punter joined #pr-challenge
05:30 ksool joined #pr-challenge
05:32 ksool joined #pr-challenge
06:24 jb360_ joined #pr-challenge
06:53 absolut_todd joined #pr-challenge
08:04 tgy joined #pr-challenge
08:06 tgy_ joined #pr-challenge
09:40 diocles joined #pr-challenge
09:45 neilb joined #pr-challenge
10:36 choroba joined #pr-challenge
10:48 kid51 joined #pr-challenge
11:51 veryrusty joined #pr-challenge
12:10 diocles joined #pr-challenge
12:16 choroba_ joined #pr-challenge
12:16 choroba I got Catalyst-View-TT-Alloy for March
12:17 choroba The metadata mention https://github.com/djzort/Catalyst-View-TT-Alloy as the github repo
12:17 choroba but it's just a fork of https://github.com/gitpan/Catalyst-View-TT-Alloy
12:18 choroba which one should I use?
12:43 kid51 choroba: My non-expert opinion:  https://github.com/gitpan/Catalyst-View-TT-Alloy says it's a "Read-only release history".  I think that means it's just mirroring what has actually been released to CPAN.
12:44 kid51 So you should probably start from the djzort repo.
12:44 kid51 The author/maintainer will probably want to apply any p.r. or patch in his/her own repo
12:46 kid51 See https://github.com/evalEmpire/gitpan/blob/master/README.md for a discussion of the purpose of gitpan
12:48 kid51 In fact, searching for one of my own distros underneath gitpan, I see that it hasn't been updated since Oct 2014, even though I've done quite a few releases to CPAN since them.
12:48 kid51 https://github.com/search?q=user%3Agitpan+List-Compare&type=Repositories
12:49 kid51 So, if someone wanted to work on List-Compare, I would *not* tell him to start from the gitpan repository, but from my own repository.
12:50 choroba ok, makes sense
12:50 choroba thanks
13:44 choroba hmm... Neil's e-mail claims the distribution has testers' failures, but I don't see any
13:45 choroba http://www.cpantesters.org/distro/C/Catalyst-View-TT-Alloy.html
13:47 Zoffix The last version was released Feb 23, 2015... Neil might have had data from an earlier period.
14:17 choroba joined #pr-challenge
15:04 neilb choroba: the CPAN Testers data was just a wee bit out of date when I sent out March’s emails — sorry
15:05 neilb it will be up-to-date when I do the april assignments
15:08 neilb honest guv!
15:08 choroba so what do I do now?
15:09 choroba was just starting to install the whole Catalyst
15:15 Zoffix There are Kwalitee errors: http://cpants.cpanauthors.org/dist/Catalyst-View-TT-Alloy/errors
15:15 Zoffix It also seems to have author/release tests in t/ : https://metacpan.org/source/DJZORT/Catalyst-View-TT-Alloy-0.00006/t
15:34 basbl joined #pr-challenge
16:58 ether Zoffix: author/release tests in t is fairly normalish - the [@Basic] dzil plugin (via [ExtraTests]) does that
16:58 ether I prefer to keep my tests in xt/, but it's an author preference
16:59 ether if dist.ini does not use [ExtraTests], it should use [RunExtraTests] which provides a 'dzil xtest' command as well as ensuring that those tests get run during dzil test --all, and also during 'dzil release'
17:58 kanashiro joined #pr-challenge
19:09 Zoffix ether, but wouldn't they run during installation?
19:09 Zoffix Oh, never mind. They have a check for release testing in them.
20:34 gansh joined #pr-challenge
20:42 veryrusty joined #pr-challenge
21:08 diocles joined #pr-challenge
21:40 punter joined #pr-challenge
21:47 athos joined #pr-challenge
22:13 gansh1 joined #pr-challenge
22:15 gansh2 joined #pr-challenge
22:57 gansh joined #pr-challenge
22:57 neilb I’ve culled the 80 people who had a january assignment but never did a PR, and who didn’t want to “Stick”
22:58 neilb blog about it: http://neilb.org/2015/03/28/culling-january.html
22:58 neilb so the candidate list has grown by 80 dists. Website updated.
22:59 genehack neilb++
23:00 neilb Have had a few emails from zombies, essentially saying “yeah, fair cop”.
23:00 * neilb slightly relieved
23:01 neilb January’s stats now: 167/173 complete = 96.5%
23:01 ether EJARGONFAIL
23:05 Zoffix .oO( fair cop?? ... )
23:05 preaction "that's fair" "fair enough"
23:06 Zoffix Where is that used? UK?
23:06 preaction Australia
23:06 Zoffix Ah
23:09 genehack i thought that was reasonably well understood across @english_speaking_areas
23:09 genehack TIL otherwise i guess. 8^p
23:10 * ether was thinking a la "cop to a plea", but that doesn't make sense in this context
23:11 genehack ether: i think it's the same usage, essentially -- http://idioms.thefreedictionary.com/It's+a+fair+cop
23:11 ether ah interesting!
23:11 neilb Generally it means “yeah, it’s a fair cop, you caught me”.
23:11 preaction like "cop to it"
23:12 genehack 'cop a plea' has a bit more of a bargaining connotation
23:12 neilb So in this context, “yeah, I hadn’t done anything and almost certainly wasn’t going to do anything. So fair enough to cull me"
23:13 ether neilb: I can't be sure because I can't see the exact algo (:D), but I suspect your scoring is giving too much weight to failing cpantesters reports and possibly some failing kwalitee metrics, and not enough to outstanding RT/github issues
23:14 neilb issues can result in 0-3 on the score. CPAN Testers can give +1, Kwalitee can give +1
23:14 ether what specifically about cpantesters?
23:15 ether a non-zero number of failures is not at all unusual if the dist has been released for a while
23:15 neilb If there are at least 50 reports and 2% or more are fails, it gets a +1
23:15 ether there's periodically bursts of failures caused by something else at a distance going awry, and there's no way to clear that out other than doing another release
23:15 ether 2%, ok
23:16 neilb well, you can clear them out via the admin interface can’t you (not that I’ve ever looked, or done it). But yeah, easier to do a release
23:16 ether it would require some custom queries to the metabase, but it would be really nifty to look for things like 100% failures on a specific OS or perl version, as opposed to a small nubmer of failures evenly distributed across everything
23:17 ether the clearing API doesn't seem to work, but that's also only applicable (IMO) if the tester themselves were at fault
23:17 ether I wouldn't kick back a report to the tester if my failure was, say, caused by a misbehaving prereq
23:17 ether there's been talk of redirecting a report to another queue, but that's not implemented
23:18 * ether would happily kick all those failures caused by LMU 0.407 back to that queue :D
23:18 ether that one caused a lot of failures on all my dists that use Moose (which is a large majority)
23:19 ether this is all very subjective, I know :)
23:19 ether and can be resolved in time as dists get cycled through the list

| Channels | #pr-challenge index | Today | | Search | Google Search | Plain-Text | summary