Camelia, the Perl 6 bug

IRC log for #onionsketch, 2013-03-11

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

All times shown according to UTC.

Time Nick Message
14:40 khw joined #onionsketch
19:35 trs joined #onionsketch
19:42 bluescreen joined #onionsketch
19:43 Corion joined #onionsketch
19:50 mmcleric joined #onionsketch
19:54 rjbs Back and present momentarily!
19:57 rjbs ...and then I knocked over my pitcher of water.
19:57 * Corion now has a mouth
19:57 Corion Whoops
19:57 rjbs But it's cleaned up now, at least enough that I can ignore it
19:57 rjbs Fortunately, it went in the other direction from my power brick.
19:58 Corion Zzzzap
19:58 rjbs Okay, so!
19:58 Corion I almost did something like that when presenting the Frankfurt YAPC bid in Riga...
19:59 Corion Shall we start?
19:59 Topic for #onionsketch is now The State of the Onion | 5.17.10 on March 21
19:59 rjbs Let us!
19:59 rjbs I have no old business to carry over.  5.17.9 was released without issues.
20:00 Corion You've already removed one point from my list - the release is planned for 21st
20:00 rjbs I also know you should have no "first time release" issues to deal with: you've done this before.
20:00 Corion Yep, at least that's what I hope for. Of course, there's the second system effect and overconfidence, but I hope everything goes as well as last time
20:00 Corion Windows builds clean currently, which gives me hope for next week :)
20:01 rjbs Woo!
20:01 Corion I still don't have a quote for the post, but if that's the only problem, that's OK :)
20:01 rjbs So, 5.17.10 is scheduled to be "full code freeze"
20:01 rjbs The code freeze is a bit of a muddle right now because we had a few fairly important big changes still coming in for a landing, this month.
20:01 rjbs I believe all are now landed, except for 1 big one and maybe two smallers ones.
20:01 Corion Oh - which changes should still land? I didn't notice...
20:02 rjbs 1:  Yves Orton is likely to finalize some changes to hashes.
20:02 Corion Yeah, more hash randomization
20:02 rjbs This will make hash traversal less predictable.
20:02 rjbs (to grossly simplify)
20:02 rjbs We were hoping for an EUMM release before that, but I think we'll just be patching in blead instead.
20:03 rjbs 2: Interpolation of array literals in regex seems to have regressed since 5.14
20:03 Corion Has a test been added for that?
20:03 Corion (so we know it's landed and fixed)
20:03 rjbs 3: Regexp::Grammars is broken (and probably should be broken) but the "this should fix it" didn't work, so there's evidence of a bug there.
20:04 rjbs I don't believe we have a test *committed* for the array interpolation issue, but there *is* a test in the RT.
20:04 Corion Ah. Maybe I'll add that test, but if it produces black smoke, what should happen?
20:04 rjbs Wait, which test are you considering?
20:05 rjbs The tests for #2 and #3 are guaranteed to fail, since the bugs are not yet fixed, even in branches.
20:05 Corion The array interpolation-in-regexes one
20:05 Corion Ah, OK
20:05 rjbs My hope is that they *will* be between now and the 21st. :)
20:05 Corion But there are tests for that in blead. That's what I wanted to know.
20:05 Corion So there are known failures currently.
20:05 Corion This reduces my expected amount of panic :)
20:06 rjbs We don't add failing tests for known bugs like that.  We don't leave failing tests on blead, ever, on purpose.
20:06 rjbs Instead, we have the test in the RT and we would expect that when the bug is fixed, the test is put into the test suite.
20:06 Corion Ah, OK. So I shouldn't see failures now, and shouldn't on the 21st either.
20:06 rjbs Correct!
20:06 Corion Right. Failing tests would break bisecting.
20:06 rjbs Exactly so.
20:07 rjbs So, you may see a bunch more "churn" this month than we'd have hoped, given its nearness to the hoped-for date of 5.18.0...
20:07 rjbs ...but still it should be a nice calm cycle for you, I hope.
20:07 rjbs I'm hoping for an RC1 in April.
20:08 Corion I hope so too :) I'll do a couple of smokes during the next weekend I think, and a test run on Monday or Tuesday
20:08 rjbs You have the URLs for the smoke farm results?
20:08 Corion No
20:08 rjbs http://perl.develop-help.com/?b=blead
20:08 Corion Cool, bookmarked!
20:08 rjbs http://doc.procura.nl/smoke/index.html
20:09 rjbs Those are good for checking through the month.
20:09 Corion I have to get this ARM box set up for smoking - it has one weird signal failure, but so far, the box isn't stable enough yet...
20:09 rjbs Cool. :)
20:10 rjbs I don't think I have much else to say, really.  Perhaps I can get a few stakeholders for the remaining issues to show up for our next meeting to review the state of things to get them hashed out before RC1...
20:10 Corion Cool. My shell account on dromedary still works.
20:10 rjbs ...but for now, you know how to do a release, and have used the RMG before.
20:10 rjbs Good to know!
20:10 Corion I don't have any questions/comments/issues either
20:10 Corion The panic will set in later
20:11 Corion But so far, the sky looks clear
20:11 rjbs One thing that came up in recent history was git use:  when you do your final merge, make sure that you've tagged the actual commit you released, and that it ends up in the history of the blead branch.
20:11 rjbs (A few releases ago, it was rebased away and things got a bit confused.
20:11 rjbs )
20:12 Corion Ah. I think that's a change since the last time - I do the release from a release branch, which is later merged, instead of doing the release from blead, right?
20:12 Corion Hopefully nobody else commits to blead while I work on the release branch, so we keep the nice/clean history
20:12 rjbs Right.  In my experience (and if you look at the history of blead) it ends up being linear.
20:12 rjbs Only once or twice did we end up with a few commits on both sides of a merge.
20:13 rjbs but if that happens, merge , don't rebase, or we lose track of what was really released! :)
20:13 Corion Yeah - the only time the rest of the world has to rebase, not me ;)
20:13 rjbs I don't know how cozy you are with git.  I will probably be around when the release is going on, so just let me know (or ask #p5p) if you hit any issues.
20:15 Corion Not totally cozy, so if I have to stray from the beaten path, I'll come asking :)
20:15 rjbs Okay.  I've got nothing more.  You've got nothing?  If so, I declare our business concluded and will go write some code so I can buy more bacon.
20:15 Corion Yep, nothing more. Business concludes as usual.
20:15 rjbs Great.  Thanks!  Talk to you on the 19th at the same time of day?
20:16 Corion Yep, marked in my calendar
20:16 rjbs and it looks like your time zone won't change until Mar 31, so no worries there
20:16 rjbs (I was just sent a link to a petition for whitehouse.gov to push to abolish daylight saving.  Good luck, petition!)
21:48 khw joined #onionsketch
22:58 SvenDowideit joined #onionsketch
23:39 ether_ joined #onionsketch

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