Perl 6 - the future is here, just unevenly distributed

IRC log for #onionsketch, 2014-02-19

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

All times shown according to UTC.

Time Nick Message
01:53 ether_ joined #onionsketch
01:59 kid51 joined #onionsketch
02:01 rjbs TonyC: It's good to be an oper. :)
02:01 * rjbs had to dig up his oper credentials.
02:01 TonyC wondered if you were
02:02 rjbs Good evening and welcome to the #onionsketch for 2014-02, perl 5.19.9!
02:02 rjbs How's the month been going for you?
02:03 TonyC reasonably well, though I was a bit disappointed about the perl5 core maintenance fund being low on funds
02:03 rjbs Yes, indeed.  I'm taking the very limited actions available to me to bring some attention to that.
02:04 TonyC perldelta has been easier than I remembered from last time, perhaps because I got into it sooner
02:04 rjbs What's your perldelta regimen like?
02:04 rjbs Do you use the delta tool, or just keep an eye on the history, or?
02:05 TonyC grab git log --oneline --graph for the range I haven't done and keep a tracking file, open up gitk to step through each commit
02:06 TonyC mark commits with a delta entry with d, and the commits I don't think need a delta with an x
02:06 TonyC I haven't bothered with corelist-perldelta.pl, since I step through every commit anyway
02:06 rjbs aha, and then you just keep track of where you left off? yeah, that's about what I do.
02:08 TonyC I updated perlport last night, which despite the note in the RMG, doesn't have much on known platform issues
02:08 rjbs Oh, I haven't seen that commit yet, I think, or I missed it.  I'll check it.
02:08 rjbs Something that needs doing soon for 5.20.0's delta is an updating of "chopping block" stuff.
02:09 rjbs I'm not sure you have any position on that, but there are bits we promised to excise in 5.20 that haven't been (I think) and I think some bits that we suggested were limping toward extinction.
02:10 TonyC perhaps they need tickets, and to be added to blockers
02:10 TonyC assuming anyone looks at blockers anytime other than before the run up to the next release
02:11 rjbs yeah, good idea, I will do that
02:11 rjbs I look at the blockers, at least, although not as much as I want to.
02:11 rjbs I'm trying to spend more time on that.
02:11 rjbs Also, I'd like to start setting up blead blockers.
02:11 rjbs They "can't" "really" block, but they'll set up things we really mean to do before the next release.
02:12 rjbs and we can transfer them on a month if we must
02:12 rjbs good for things like "had to add exception to pod tests to keep doing the hard work"
02:12 rjbs or "fix the stupid perldelta bug"
02:12 TonyC the problem there is if you have an internals ticket, a non-internals release manager won't be able to do much about it
02:12 TonyC unless it's just for RM issues
02:13 rjbs I think it would largely be RM issues.
02:13 TonyC even then, if you expect the RM to fix them (as opposed to following the RMG script) you might scare off RMs
02:15 rjbs Well, I'm okay with somewhat reducing the RM pool by adding a job or to two it.  We'll see.
02:15 TonyC I'm not sure how hard you've had to dig for RMs
02:15 rjbs It's not too bad.  In general, the only problems are last minutes.
02:16 rjbs Someone drops out, or I put off scheduling enough in advance.  But that's not a big deal, I just do it myself if needed, and that's good too.
02:16 rjbs but generally we have no problems
02:17 rjbs So, 5.19.9!
02:17 TonyC I still need to find an epigraph
02:17 rjbs Uh oh.  That's a blocker.
02:17 rjbs Whoops, tried to look at the delta-so-far, but had maint checked out..
02:18 rjbs The big change, of course, is signatures.
02:18 TonyC I still need to check a read-only build, the rest of the pre-reqs are on-going or done
02:18 rjbs Great rejoicing, although I think that as is often the case, very few people have actually tried using them.  So far, though, I'm pleased.
02:18 rjbs Also, bunches of locale fixes.
02:18 rjbs Sounded like Karl is still hammering on one?
02:19 TonyC yes, failures on Win32 and Darwin/PPC
02:20 rjbs Hm.  I wonder whether it's worth getting my hands on a PPC machine that I happen to know is available here..
02:20 rjbs probably not, jimk still (amazingly) works on one
02:20 TonyC and in smokes on netbsd 5.1, which Karl couldn't reproduce locally
02:20 rjbs Are you aware of any other ongoing problems?
02:20 rjbs Yes, I saw some bad smoke from netbsd, but I thought I also saw BinGOs say he was suggesting skipping some tests because of it.  Am I conflating two distinct issues?
02:21 TonyC actually, the netbsd failures aren't locale related - just timing issues
02:22 TonyC I'm not aware of any other ongoing problems
02:22 rjbs Excellent.  This week I definitely begin my bell-tolling about marked blockers, and trawl for still-open BBC tickets.
02:22 TonyC I spent a bit of time tracking down some of the long term cygwin failures (and a new one with the latest cygwin)
02:23 TonyC actually, -Uusedl fails some tests in a cpan/ module (Module-Load?)
02:23 rjbs huh
02:23 TonyC -Uusedl doesn't get smoked often
02:23 rjbs Do we have any cygwin smokers?
02:24 rjbs Yeah, not surprising.
02:24 rjbs I thought I'd seen cygwin smoking in the past, but I see nothing now.
02:24 TonyC mine, but the smoke-db bounced a report
02:24 rjbs huh
02:24 rjbs Okay.
02:24 * rjbs wonders whether Tux will be doing more work on the smoke database stuff this year at the QAH.
02:24 TonyC the smoke-db expects UTF-8 encoded JSON, but the build/test log isn't necessarily UTF-8
02:25 TonyC it might be fixed in the latest Test-Smoke code, but I need to backport some changes (distinct git-svn imports)
02:26 rjbs well, good luck with that; will be nice to see cygwin show up again
02:26 rjbs Why do you suppose we don't see a lot of smoking of (say) -Uusedl?  Just that the smoking machines are already busy enough with the more common configurations?
02:26 TonyC I might just disable smoke-db reporting for now
02:26 rjbs I really have no conception of how much time they're spending idle.
02:26 rjbs (although I know that it's a bizarre generalization to suggest they can be lumped together as if they had one profile of behavior)
02:26 TonyC I have -Uusedl as a "when there's nothing else to do" configuration
02:27 TonyC and the machine with that config is running my older smoke scheduler which prefers smoke-me over blead, so smoke-me sees a -Dusedl run more often than blead, and then not often
02:28 rjbs Is that something supported by test smoke's (notional?) scheduler?
02:28 rjbs interesting
02:28 rjbs I should really set up some smoking machines, if only to get an idea of how it all really works.
02:28 rjbs maybe I'll get some solaris smoking going, gcc and sun cc. that would be nice.
02:28 TonyC no, it just has a "run a smoke at this time of day" and "run smokes anytime there's a new commit" from memory
02:29 rjbs I saw your comment on the dtrace dependency ticket but didn't read it carefully yet; thanks.
02:29 TonyC mine looks for new smoke-me commits and bleed commits and has a default config and pool of alternate configs to run
02:29 rjbs Oh, is your smoke system something above and beyond test-smoke, then?
02:30 TonyC a wrapper around it
02:30 rjbs Huh, lib/English.t just failed for me!
02:30 rjbs So, back to 5.19.9, maybe. :)
02:30 rjbs It sounds like things look good, which is good.
02:30 rjbs Have you looked at the 5.20.0 blocker ticket lately?
02:31 rjbs https://rt.perl.org/Ticket/Display.html?id=116923
02:31 rjbs better: https://rt.perl.org/Search/Results.html?Query=DependedOnBy%20%3D%20116923
02:31 TonyC a couple of days ago, when I bisected 121259
02:31 rjbs better better: https://rt.perl.org/Search/Results.html?Format=%27%20%20%20%3Cb%3E%3Ca%20href%3D%22__WebPath__%2FTicket%2FDisplay.html%3Fid%3D__id__%22%3E__id__%3C%2Fa%3E%3C%2Fb%3E%2FTITLE%3A%23%27%2C%0A%27%3Cb%3E%3Ca%20href%3D%22__WebPath__%2FTicket%2FDisplay.html%3Fid%3D__id__%22%3E__Subject__%3C%2Fa%3E%3C%2Fb%3E%2FTITLE%3ASubject%27%2C%0A%27__Status__%27%2C%0A%27__QueueName__%27%2C%0A%27__OwnerName__%27%2C%0A%27__Priority__%27%2C%0A%27__NEWLINE__%27%2C%0A%27%27%
02:32 rjbs holy cow!
02:32 rjbs well, nevermind
02:32 rjbs sigh, RT.
02:32 rjbs I think a number of those tickets are not quite blockers.
02:32 rjbs I think some will be moved to 5.22
02:32 rjbs Do you think, though, that we're missing anything vital?  If I'm going to start droning on this week, better to have all the stuff worth droning about.
02:34 TonyC I'm not entirely sure about the behaviour in 119499
02:34 rjbs Yes, that's something that will need to be sorted out.
02:34 TonyC which changes the string form of $!
02:34 TonyC and $^E
02:35 rjbs The impression I've gotten is that the change is going to have to be at least mostly reverted, and that Karl is now working on rationalizing the code so that it can be altered more easily in the future, includign the possible reversion
02:36 rjbs but it's a hairy mess
02:37 TonyC it's a long ticket, when I last looked at it I couldn't figure out what the plans were
02:37 rjbs I'll talk it over with Karl between now and … well, hopefully before I start ringing the blocker bell.
02:37 rjbs Another blocker on there is "remove Pod::Parser"
02:38 rjbs that won't happen, btu I am hoping I can get the new Pod::Checker merged; I'll probably spend much of the QA Hackathon on that
02:38 rjbs I am *really* not excited about it.
02:38 TonyC begone cruft!
02:38 rjbs yeah
02:38 rjbs I'll feel like a champ if I get it done, anyway.
02:39 rjbs and there will be chartreuse at night to wipe away any horrors of doing it
02:39 rjbs Okay, so sounds like everything is jake.  Do you know when you plan to release the release?
02:39 TonyC tomorrow Sydney time
02:40 TonyC it's 1:40pm Feb 19 here
02:40 rjbs boy, strangest thing; my terminal program just hung for a minute :-/
02:41 rjbs yeah, I put Sydney time on my dashboard, as it's ended up being the time I most often want to check and can't do in my head
02:41 rjbs Cool, so that's "tomorrow evening" for me.  I look forward to it!
02:41 TonyC add 12 hours and subtract 1
02:41 TonyC urr, from GMT I guess
02:41 rjbs ;)
02:42 rjbs Is there anything you want me to get onto my todo list, or that I can confirm/deny/whatever?
02:44 TonyC no, I'm good thanks
02:44 rjbs Excellent.  Okay, I'll be around a bit longer this evening, but I think we've covered everything, and I should start taking care of the final things I need to do before turning in.
02:44 TonyC ok, sleep well
02:44 rjbs Thank you!  and good luck!
05:33 ether_ joined #onionsketch
05:38 ether joined #onionsketch
09:18 TonyC_ joined #onionsketch
14:57 khw joined #onionsketch

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