Camelia, the Perl 6 bug

IRC log for #onionsketch, 2012-06-04

| Channels | #onionsketch index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
07:00 hoelzro left #onionsketch
10:16 Zefram joined #onionsketch
10:41 Mithaldu joined #onionsketch
11:43 Mithaldu joined #onionsketch
11:46 tomboh_ joined #onionsketch
16:01 toddr joined #onionsketch
16:02 toddr joined #onionsketch
17:56 doy joined #onionsketch
17:59 rjbs Hi!  I'll be ready in just 30s.
18:00 doy okay
18:02 rjbs Okay!  I was in the middle of an "add -p" :)
18:02 doy wouldn't want to interrupt that
18:02 doy (:
18:02 rjbs I'd totally forget what the heck I was doing.  I finished the first set of commits and can finish the rest later...
18:03 rjbs ANYWAY!
18:03 rjbs Hello, and welcome to #onionsketch for June 4.
18:03 Topic for #onionsketch is now The State of the Onion | 5.17.1 due June 20th
18:03 rjbs So, as Seveas just said in #p5p, you already had a commit bit!  Who knew?  I guess I did...
18:04 rjbs I wonder why you do not appear at http://perl5.git.perl.org/committers.cgi -- I'll ask Seveas.
18:04 rjbs And you have your "pumpking bit" on PAUSE now.
18:04 doy yeah, saw that too
18:04 rjbs So, technically speaking, you should be able to do the release.
18:04 rjbs Have you tried pushing a head to perl5.git yet, just to make sure it works?  (Something you can delete afterward :)
18:05 doy i have not, i didn't realize i had a commit bit until about 5 minutes ago
18:05 rjbs Yeah.  Well, so, there's an action item for you. :)  Make sure the bit works before the 19th. :)
18:05 doy error: insufficient permission for adding an object to repository database ./objects
18:05 doy i suppose that should go to #p5p
18:05 rjbs Okay.  I will continue to bug Dennis until this is sorted.
18:06 rjbs I'm guessing my question just now about the committer list will be relevant.
18:06 doy yeah, probably
18:06 rjbs (Using gitolite might fix this, but Dennis has indicated that it would be a big PITA for him to switch right now.  Maybe we'll get to it later this year...)
18:07 rjbs Okay, sorry, just updating my little list of to-do-after-meeting. :)
18:07 rjbs I don't think we have much old business from last release, except for the version bump time.
18:08 rjbs Zefram: If you're here:  Do you recall any loose ends from last meeting or 5.17.0?
18:08 Zefram there was nothing explicitly to discuss now
18:08 Zefram one thing came up during release that is slightly relevant
18:09 Zefram some of the CPAN comparison code broke due to CPAN modules now having their .meta in JSON
18:09 Zefram I coded around it, so should be fine for now, but if [rt.cpan.org #77447] gets fixed then it should be recoded more cleanly
18:10 rjbs Yes.  You know, we wrote something to autodetect based on content in CPAN::Meta.  I wonder if we decided not to keep it, before the first version was published!
18:10 rjbs I also encountered that bug, though, doing 5.16.0 and some other work.
18:10 rjbs I'll see if we can't get that fixed before June 20.
18:10 rjbs Thanks.
18:11 Zefram oh, and I said I'd work on a separate release branch, but in the end there were no conflicting commits, so it's as if there was an actual freeze on blead.  so no real experience to report there
18:11 rjbs Right, so.
18:11 rjbs doy: I need to update a few things in the RMG that won't affect you any time soon, but let me give you the very very short version…
18:12 rjbs (1) Rather than say "we have frozen for 24h" or anything like that, just make a branch to do the release work in as late as you like, then in the event that blead moved while you worked, you can merge it back together when you're done.
18:12 rjbs This, I realize, is one case where we will allow non-rebased merges.
18:12 doy okay
18:13 Zefram if there were conflicting commits, that merge will be extra exciting wrt perldelta
18:13 rjbs (2) I need to add a "send an email to xyz" step or two; basically, don't sweat it, but remember to check for changes to the RMG between "whenever you make your checklist" and Jun 18 or so. :-)
18:14 rjbs doy: Have you reviewed the RMG steps yet?
18:14 doy yeah, i read through them over the weekend
18:15 rjbs Any questions or problems?
18:16 doy not at the moment (other than the version number thing that was discussed earlier)
18:16 doy i'll probably run through everything sometime this week
18:16 doy to make sure it all makes sense
18:16 rjbs Yeah.  For now, let's assume that any change to version numbers will happen /after/ your release.
18:16 doy okay
18:17 rjbs I think the answer there is going to be "the releaser of X is repsonsible as one of his last acts for bumping to X+1"
18:17 rjbs and the question of the "never released" version number will be "we will do that once someone gets the process built and tested"
18:17 rjbs Zefram: HINT HINT :-)
18:18 rjbs doy: You will want to keep an eye on smoke color just in case everything goes horribly wrong just before release and there's a chance to back out one commit to make the dev snapshot work a zillion more places.
18:18 Zefram ok
18:18 rjbs In the 2nd release, it's not the end of the world if one or two platforms need to go back to the blead version, but if avoidable, nice to avoid.
18:18 rjbs http://perl5.test-smoke.org/ <-- useful
18:19 doy oh, one thing i did notice actually, we need release managers for the next few months
18:19 rjbs also useful: http://perl.develop-help.com/?b=blead
18:19 doy i think i'm the last one scheduled
18:19 rjbs Yes, we do!  Don't worry about it, I'm working on it.
18:19 rjbs I know it won't be a problem, but it's good to keep me bugged about it.
18:19 doy ah, i see a couple more were added in the past couple days
18:21 rjbs Right now, the builds look pretty good.  There are a few big branches in flight that might get merged between now and then, but we'll see what happens.
18:21 doy i guess i should stop hiding the smoke report emails
18:22 doy (:
18:22 rjbs Also, consider running through a few of the before-you-release steps to see which ones are problematic or fail-y and whether you can fix the automation.
18:22 rjbs One of the most valuable things you can do as rel eng is to improve the automation for the next guy.
18:22 rjbs (So I'll really have a go at that CPAN::Meta bug. :)
18:22 doy yeah
18:23 rjbs Yes, unhide the smoke at least for June. :)  And keep an eye on new RT submissions.  There are a decent number of eyes on bugs these days, but one more pair for the month helps.
18:23 rjbs Other than that, I don't have anything to mention.  So far, nothing is particularly hinky in blead.
18:24 rjbs I'd suggest we meet again on the 18th, same time.
18:24 doy do all new rt submissions end up on p5p?
18:24 rjbs We can go over anything that came up and make sure all systems are go.
18:24 rjbs All new RT submissions end up on p5p, subject to moderation.
18:24 * rjbs is a moderator and tries to moderate regularly.  Worst case delay is usually "a day."
18:24 rjbs So: reading p5p is sufficient.
18:25 doy okay
18:25 rjbs Anything else you need from me/us/anyone?
18:27 doy not that i can think of
18:27 rjbs Cool.  Then we will meet again on the 18th, if that works for you, and of course I'll be on IRC continually if anything comes up.
18:27 doy and also at yapc
18:27 doy (:
18:27 rjbs Indeed!
18:28 rjbs doy: Oh, you know, one patch you could look at, because it might be nice to get merged, would be Brian F.'s require change.
18:28 rjbs I'm guessing you've got some knowledge of the area of code being touched.
18:28 doy i will keep it in mind
18:28 rjbs I will also try to have a look at it, but haven't so far.
18:28 rjbs Cool.  It's not urgent, but it'd be nice to see what we break. :-)
18:28 rjbs Thanks very much for your time, talk to you later!
18:53 Mithaldu joined #onionsketch

| Channels | #onionsketch index | Today | | Search | Google Search | Plain-Text | summary