Camelia, the Perl 6 bug

IRC log for #onionsketch, 2013-10-02

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

All times shown according to UTC.

Time Nick Message
03:39 khw joined #onionsketch
07:57 ribasushi joined #onionsketch
14:02 rjbs_ joined #onionsketch
14:04 rjbs_ joined #onionsketch
14:05 rjbs Foo.
14:05 rjbs Ridiculous.
14:05 rjbs I had ops (and opers) here on my electret.
14:05 rjbs but looks like everyone else is out of sync
14:05 rjbs so I had to connect as rjbs_ on astral and op myself from there
14:05 rjbs voice ribasushi
14:06 rjbs ribasushi: Can you talk now?
14:06 rjbs You show voiced here.
14:06 ribasushi voice, you are doing it wrong
14:06 rjbs Good.
14:06 ribasushi :)
14:06 ribasushi alright so, we actually spoke with rjbs a little earlier this morning, but to recap
14:07 * ribasushi followed through the RMG and made a bunch of trial-runs of the avrious utilities, all the way to making a tarball, everything seems sane, no questions on procedures etc
14:08 ribasushi I still need to chase ANDK for a pumpkin bit
14:08 * rjbs can take care of that.
14:09 ribasushi on the point of obtaining a rt queue admin rights - the RMG was not clear whether point relmanagers actually are expected to triage any particular classes of bugs ([META something])
14:09 ribasushi (I will add a clarification to the rmg once this is clarified)
14:09 ribasushi clearly... ;)
14:09 ribasushi rjbs: ^^
14:10 rjbs I'll get RT sorted out, also.  As for expectations:
14:10 rjbs The best thing is to notice tickets about regressions and "BBC" notices.  (Blead Breaks CPAN)
14:11 rjbs and, having noticed them, to try to respond with something of use: "I can't reproduce this" or "I bisected this" or "I bet this is because X"
14:11 rjbs after that, just having an eye on new tickets during the month is useful
14:12 rjbs in theory there are a few dozen people doing the same, but when you're employee of the month, it's a good time to be extra vigilant for bugs that might be trivial or important to fix right now
14:12 ribasushi understood, do we have *any* text related to what you just said in any of the Porting/ docs?
14:12 ribasushi if not I will add a paragraph or two
14:12 rjbs I doubt we do.
14:13 ribasushi aye
14:13 rjbs Right now the RMG is mostly a procedural checklist.
14:13 rjbs A section, even a stubby one, on "stuff you'll be doing all month" would be good.
14:13 ribasushi I also am thinking to reformat the RMG to actually start with a section that links towards the rest of the document below
14:13 ribasushi right, gmta then
14:13 rjbs You get that if you use the ...
14:14 rjbs Porting/make-rmg-checklist
14:14 ribasushi the little utility mentioned yes (something checkilist)
14:14 rjbs yeah
14:14 ribasushi right good point that we have a dynamically generated thing, I forgot about it ;)
14:15 rjbs I really like that thing, and would like to keep making it better.
14:15 rjbs Its "skip if X" logic is useful, but imperfect.
14:15 ribasushi right, running it now again, I will ponder how to make it more n00b-friendly
14:16 rjbs What's your rt.perl.org username?
14:16 rjbs (when ou're logged in, what does it say on the upper right corner?)
14:16 ribasushi sec
14:16 ribasushi actually bitcard is the only supported thing right?
14:16 ribasushi not like the cpan rt
14:16 rjbs yeah
14:16 rjbs right
14:17 rjbs for now :)
14:17 ribasushi Logged in as ribasushi
14:17 rjbs hoping to see that fixed in the v4 upgrade, which is still a'movin'
14:17 rjbs okay
14:17 ribasushi aye ;)
14:17 rjbs So, we've wandered through most of my usual agenda already, good.  More things:
14:18 ribasushi another question (that is not answered seemingly anywhere in Porting/) is what is the general workflow for releng related changes (i.e. fixes/clarifications to the RMG)
14:18 ribasushi (finishing up q)
14:18 rjbs I suggest just committing them and posting a note to p5p that you have done so, unless you think they are controversial.
14:18 rjbs Otherwise, put them in a branch and ask for review.
14:19 rjbs (they're just like any other change in this regard, but fewer people will care :)
14:19 ribasushi ok fair enough, this particular bit wasn't listed anywhere, will add an RMG note on "editing this and related documents"
14:19 rjbs For sure.
14:19 rjbs There's at least one likely interesting change to land soon: experimental postfix dereferencing.
14:19 ribasushi right, I come from more "bondage oriented" projects, so it's hard to establish the right footing
14:20 rjbs Our policies are tightening up little by little, but it's very slow, so yeah.  There's still a lot of "do what thou wilt."
14:20 ribasushi right - so on deref
14:20 ribasushi I am relatively uptodate on the final design, of course we have not really heard from CPAN yet
14:20 rjbs I don't expect to see many problem exposed with the feature landing, but I'll be running a week of smoking first, probably starting today.
14:21 rjbs and then once it's merged, we'll see whether BBC reports show up
14:21 rjbs so there should be time to figure out whether we need to do much about it
14:21 ribasushi right, I assume even if it turns out utterly unworkable, the back-out commit will happen sometime in Nov.
14:22 ribasushi to allow time for the customary choice of paint
14:22 rjbs Yeah, although I doubt that will happen.  I've converted a fwe hunks of code to it w/o issue.  But time will tell!
14:22 rjbs phew, and it rebases atop blead nicely
14:22 ribasushi \o/
14:22 rjbs I don't *think* we have other major changes coming.
14:22 ribasushi um
14:22 ribasushi what about P?
14:22 * ribasushi runs
14:23 rjbs I'm working on an alternate implementation.  We'll get П.pm in the core soon.
14:23 rjbs Anyway!
14:23 ribasushi my name is Peter Rabbitson, and I approve this change.
14:24 ribasushi anyway!
14:24 rjbs So, there's one more agenda item, and then a meta-item.
14:24 ribasushi it looks like I have the rest of my bases covered, and do not have outstanding questions at this moment
14:24 ribasushi go ahead
14:24 rjbs I will once Calendar.app opens. :-/
14:24 ribasushi I heard you like vm's in your vm...
14:25 rjbs I think my computer is angry because I'm compiling.
14:25 rjbs Well, I got it on iOS.
14:25 rjbs I have our next meeting for the same time, Thursday, Oct 17th.
14:25 ribasushi this is correct
14:25 rjbs Okay, good.
14:25 rjbs The meta-item is:
14:26 rjbs Earlier you were not sure you'd have the time to carry on this work.  I'm not able to answer that, and I'm not even sure I can give a good estimate of how much work it will take to get the job done well enough (as opposed to heroically).
14:26 rjbs The main time-consumers that *must* occur are perldelta wrangling and corelist wrangling.
14:26 rjbs After that and testing, most things are "optional but highly encouraged."
14:26 rjbs You should let me know (ASAP) if you think you'll need to bow out this time, and we can rope you back in another time.
14:27 rjbs In my experience, it's a pretty easy task, but my experience is clouded by having done it quite a lot.
14:27 ribasushi aye - to clarify for the bystanders
14:28 ribasushi given that the RMG is a well written but *large* body of text, I augmented my understanding of it by observing the actions of the past two rel engineers (Steve and David)
14:29 ribasushi I will actually put a note about this (what is expected and what is heroic) at the top of the RMG
14:29 ribasushi I am still distilling how to word it correctly to not sound discouraging while at the same time not to encourage NMP attitudes
14:30 ribasushi from what we talked about so far I do not see a problem with me continuing onwards and having a "I am all good to go" chat on the 17th
14:30 rjbs Consider starting withs omething like the priority list I mentioned in privmsg.
14:30 rjbs It may be a good start to better formally defining the releng job in print.
14:30 ribasushi right, this is the basis for it
14:31 ribasushi tbh I dove into the pot specifically to get a feel of "large hairy project" releng procedure
14:31 ribasushi so most of my work will be concentrated on improving the verbiage around it
14:31 rjbs Cool.
14:32 rjbs Yeah, the release engineering of perl5.git is not an extremely regimented process.  The history is:
14:32 rjbs it used to be that the technical-master-pumpking was the only one who did it, and it was sort of "I know the stuff to do, and I do it, and then we have a release"
14:32 rjbs so it was demystified and distilled into the RMG
14:32 rjbs and from there has been more and more automated and clarified
14:33 rjbs I think it's finally getting to the state where "follow these steps and you are good" is about right, but they're not all prioritized correctly, exceptional cases aren't spelled out,e tc.
14:33 rjbs so it sometimes may seem like "this must happen no matter what" when in reality it's "this is kinda good too"
14:33 ribasushi right, which is what tripped me to go for "maybe next time"
14:34 rjbs nod
14:34 rjbs okay; if we're done, I should go, work is calling :)
14:34 rjbs but you know where to find me with any future question!
14:35 ribasushi I am not going to have access to hack-time until late Sun (which is why we have the sketch so early), but expect the first draft of "prioritizations" to come your way early next week if not even earlier
14:35 ribasushi yep, no further questions, and I know where you live
14:35 rjbs o/
14:35 ribasushi o/
14:35 rjbs cool, thanks, ttyl
14:43 khw joined #onionsketch
14:50 arcanez joined #onionsketch
14:54 arcanez left #onionsketch
17:29 perigrin joined #onionsketch

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