Camelia, the Perl 6 bug

IRC log for #onionsketch, 2013-01-18

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

All times shown according to UTC.

Time Nick Message
04:15 SvenDowideit joined #onionsketch
08:12 arc joined #onionsketch
16:05 ether_ joined #onionsketch
16:12 ether joined #onionsketch
16:21 ether joined #onionsketch
16:54 ether_ joined #onionsketch
17:08 arc joined #onionsketch
18:57 * rjbs is here, thanks to Calendar.app!
18:58 * arc too
18:58 arc (well, thanks to the reminder on my phone)
18:58 rjbs How has your period of mandatory bonus duty been treating you?
18:59 arc mandatory bonus duty?
19:00 rjbs Sorry, irrelevant Paranoia RPG reference.
19:00 arc ah
19:00 rjbs The release shepherding
19:00 arc i regret i've been a little absent this week, but today and the whole of the weekend have been allocated to the task
19:02 rjbs Oh, right, the release is Sunday and not tomorrow.  Excellent.
19:02 rjbs I hope you will find that it will not take the whole time you have allocated :)
19:02 rjbs voice BinGOs
19:02 rjbs oops
19:03 arc my plan is to update perldelta and do a practice release run on a local branch tomorrow
19:03 * BinGOs waves
19:03 arc leaving sunday for the real thing
19:03 rjbs arc: Cool, that's what I'd advocate doing.
19:03 rjbs arc: The two bits that, ime, are the fiddliest and more tedious are perldelta and Module::CoreList.
19:03 arc yeah, that sounds plausible
19:04 arc perldelta looks like it's had quite a lot of the big stuff added already, fortunately
19:04 rjbs Yeah, the perldelta situation hasn't been too bad for a while.
19:04 arc and i've been keeping an eye on commits, and they don't look surprising
19:04 rjbs I'm having a quick look at smokes now...
19:05 arc today i've been (among other things) trying to reproduce some of the smoke failures
19:05 arc and failing, sadly :-(
19:06 rjbs More red than I'd expect.  I need to find myself one best view for reviewing smokes more regularly.
19:06 rjbs I used to do it more fastidiously.
19:06 rjbs Well, at any rate, my recent smoke-mes suggested to me that things are not too bad at present, if not perfect.
19:07 rjbs The SAW_AMPERSAND problem, I think, remains unaddressed.
19:07 rjbs That should probably get a "known problem' entry.
19:07 arc the speed problem? yes, i agree
19:08 rjbs Other things mentioned last time:  (?[ .. ]) is now in.
19:08 rjbs I believe that got a delta already?
19:08 arc yes, it did
19:08 arc i'm delighted to be releasing that work; khw++
19:09 rjbs The "once" warnings were not addressed, I think. :(  I'll chase that up.
19:09 rjbs Karl is a treasure who must be chained up and kept.
19:09 arc lack of "used only once" warnings in imported code, right?
19:10 rjbs Yeah, I believe the concise description is "'once' warnings only affect main compilation unit."
19:10 arc right
19:10 rjbs Probably fixing that will turn all of CPAN into a warning-squawking machine.
19:10 arc i fear so
19:10 rjbs So I'd suggested maybe not landing that until 5.19.0.  We'll of course have to have a patch first to consider where to apply it! :)
19:11 rjbs I really don't think I have anything else.  We've seen a lot of version bumps this release, mostly to work on eliminating "low line versions."
19:11 rjbs The deprecation of Text::Soundex and Pod::LaTeX in core should not affect many people. :)
19:11 arc no, probably not :-)
19:12 rjbs BinGOs: You had said you were going to have a look at the effects of 'use deprecate' on CPANPLUS.
19:12 rjbs BinGOs: Did that happen?
19:12 rjbs I'm ready to start sending out patches for the other modules in our list, but if you've found that it will be a travesty, maybe not.
19:13 BinGOs just adding use deprecate to all of the CPANPLUS modules didn't cause any warnings when using 'cpanp' to do stuff.
19:13 rjbs That's a bit surprising, actually.
19:13 BinGOs mainly because a lot of the CPANPLUS only has use strict
19:13 rjbs Ahh.
19:14 rjbs Yes, I was mildly surprised to learn that "use deprecate" was not "mandatory," but then I realized that I don't think we have a simple way to say "mandatory unless suppressed" for perllexwarn.
19:14 BinGOs forcing with perl -w cpanp did generate a lot of deprecation warnings
19:14 rjbs BinGOs: How do you feel about that?  I mean: would you rather that a warning is issued, to get users warned?
19:15 BinGOs I am slightly uneasy about the word 'deprecated'
19:15 rjbs Yeah, one sec.
19:15 rjbs The actual warning is:
19:15 rjbs $package will be removed from the Perl core distribution in the next major release. Please install it from CPAN. It is being used at $call_file, line $call_line.
19:15 BinGOs I might lose all three of the users I currently have >:)_
19:16 rjbs So, "deprecate" is never shown the user.
19:16 rjbs I agree that it's a confusing choice of word, but it's what the pragma's called now, and I'm not keen to work on renaming it.
19:16 BinGOs its cool though, the will of the people and all that.
19:17 rjbs CPANPLUS is a lot of modules.  I suppose it would be best to send you a patch that patches each module for 'use deprecate'?
19:18 BinGOs Now this was another thing I was thinking about ...
19:18 * rjbs listens.
19:18 BinGOs I see that Switch has the use deprecate line in it in the CPAN edition
19:19 rjbs Yeah, also Pod::LaTex now.
19:20 rjbs I was going to ask whether you'd rather we only put a "use deprecate" in a final core version for CPANPLUS.
19:20 rjbs after which we won't need to upgrade it in core, and will remove it in 5.19.0
19:20 BinGOs That was what I was thinking.
19:20 rjbs I think that's quite reasonable, also, as long as you're careful not to reuse the same version number. :-)
19:21 rjbs If you like, I can make the requires patches in a topic branch.
19:22 BinGOs I hadn't made CPANPLUS install into site yet
19:22 rjbs That can be your first release with a version higher than the one that adds deprecate.pm! :)
19:22 rjbs I'm really glad I remembered to add that change when I submitted deprecate.pm patches, myself.
19:23 rjbs It would've been embarrassing to have to submit a second one.
19:23 BinGOs I've had some irrational fears about sitebin and versiononly perls that have made me not do that
19:24 rjbs Mm.  Something you could test to make yourself feel less anxious?
19:24 BinGOs I didn't know for sure whether a non-dev perl could be versiononly before I tried it the other day.
19:25 * rjbs thinks we use that in production, actually.
19:25 BinGOs but I already deal with that in the CPANPLUS Makefile.PL
19:27 BinGOs I am probably just being over-cautious with the sitebin concerns. More often than now sitebin is the same as normal bin
19:27 BinGOs s/now/not/
19:27 rjbs Well, I'll prepare the branch, which should be pretty easy.  Meantime, you can investigate any last possible avenues toward disaster.
19:27 BinGOs Yeah. Its all cool.
19:27 rjbs Sound good?
19:27 BinGOs Yeah. Its all cool.
19:27 rjbs \o/
19:27 rjbs Okay, and I'll try to get the other patches out ASAP.
19:28 rjbs arc: Any other business from your end?
19:28 arc i think i'm all good
19:28 rjbs BinGOs: Anything else?
19:30 BinGOs No, thanks. I'll be around as usual with deal with Module::CoreList release-age
19:30 rjbs Thanks, always appreciated.
19:30 rjbs arc: Oh, and: I don't know at all where I'll be Sunday, but if you need anything, I'll give you an email address that pages me.
19:30 BinGOs I may even be able to construct coherent sentences.
19:30 arc cool, thanks
19:31 rjbs Okay, then I think we are done.  Thanks very  much for your time!
19:31 rjbs arc: ...and good luck with the release!
19:31 arc rjbs: thanks!
19:31 * BinGOs salutes and signs off.
20:12 arc1 joined #onionsketch

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