Perl 6 - the future is here, just unevenly distributed

IRC log for #phasers, 2011-08-30

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

All times shown according to UTC.

Time Nick Message
13:22 [Coke] joined #phasers
17:13 mberends joined #phasers
17:22 masak joined #phasers
17:28 benabik joined #phasers
18:22 pmichaud my #phasers report:  Haven't had properly shaped tuits for p6 hacking; participated in some #perl6 and #parrot discussions, though.  I'm expecting things to clear up a bit tonight and/or tomorrow.
18:24 masak my report: blogged some. expect to blog some more in the next few days. I have a macros-in-Rakudo Hague grant application in the works. EOR
18:28 Util Pre-report:
18:28 Util Nothing done, due to $WORK.
18:28 Util Plan to play with Niecza-9, and give talk on RosettaCode to Atlanta.pm on Thursday.
18:28 Util EOR
18:29 moritz what I did:
18:30 moritz * aligned names of backtraces classes with latest spec
18:30 moritz * removed 'mu' from backtraces
18:30 moritz * applied memory leak patches by mls++
18:30 moritz * fixed a bug that prevented List::Utils from compiling (still fails at run time)
18:30 moritz * various small fixes
18:31 moritz * blogged about my progress
18:31 moritz * blogged about why we need NQP (metacircularity)
18:31 moritz Will do: revive nom-exceptions branch
18:31 moritz EOR
18:32 jnthn report: mostly took a week offline-ish, to visit friends in the UK and try to relax a bit. Next few days will be a bit of a blur with $dayjob, then things settle down and I can get back to more actively working on Perl 6 stuff. EOR
18:32 jnthn moritz++
18:33 masak moritz++
18:34 mberends moritz++ jnthn++
18:35 masak (getting karma for slacking off? well, I never!) :P
18:35 jnthn huh, what, wow...I should slack off more :P
18:35 mberends no!
18:35 jnthn :)
18:35 masak mberends: look what you did! :P
18:35 mberends *facepalm*
18:41 mberends done:
18:41 mberends * added categorize() to nom and nice tests for it to roast
18:41 mberends * added the roast-tested part of sprintf() to niecza
18:41 mberends * rewrote niecza doc/compiler.pod
18:41 mberends plan:
18:41 mberends * more work on niecza sprintf() and its roast tests with diakopter
18:41 mberends * get back to 6model/c (currently on memory management/GC)
18:41 mberends EOR
18:42 * masak categorizes mberends as mberends++
18:46 moritz mberends++ indeed
18:50 colomon report: did nothing much except use rakudo master & the ABC module to generate sheet music.  did verify that ABC module cannot yet run on nom.
18:51 colomon plan: complex number issues blog post.  possibly try ABC on niecza again.
18:51 colomon EOR
18:58 * Util is keen to see that blog post from colomon
18:59 colomon Util: I'm keen to write it, too, it just seems like there is always something more important for me to be doing each day until I'm too tired to think straight...
19:00 Util I'm right there with you
19:00 moritz \o
19:00 jnthn o/
19:01 moritz it's time to start
19:01 moritz anybody got another report to submit?
19:02 birdwindupbird joined #phasers
19:03 moritz seems like a "no"
19:03 moritz pmichaud: what's the plan for the August release?
19:07 pmichaud moritz: it kind of depends on my tuit availability.  I could do one tomorrow, if I get tuits tonight (60% chance)
19:11 colomon is that an R* release or nom as new master or ... ?
19:11 Util Do we expect the September release (or rather, the release after the one under discussion) to be Nom or Master-based?
19:12 mberends certainly nom, there have been almost no commits to master in three months now
19:17 pmichaud we don't plan any more releases based on the current master branch.
19:17 Util ok
19:17 moritz pmichaud: is there anything I could do to help you prepare the nom release?
19:20 pmichaud the biggest thing is to decide what blockers we have for a nom release and make sure they're removed to whatever extent we can
19:20 pmichaud I know that regexes are an issue
19:20 pmichaud I feel like the slow compilation speed is also a big issue... but not sure what to do about that.
19:22 jnthn I think we mostly block on understanding why it's slow.
19:22 pmichaud right, but is slowness a release blocker?
19:23 pmichaud or do we release a nom-based compiler even knowing that it's really slow and has memory issues?
19:23 jnthn For compiler or distro?
19:23 jnthn For distro, I'd really like a shot at tracking this down.
19:23 jnthn But realistically I won't have tuits until the weekend for that. And the weekend is in September.
19:24 jnthn Of course, @other are free to beat me to it. :)
19:24 moritz well, I think we can forget basing a distro on an August nom release
19:24 pmichaud well, I think for distro we have to have that fixed.
19:24 pmichaud I'm asking and only looking at compiler atm.
19:24 moritz there are just too many things that stop modules from workin on nom
19:24 jnthn OK. For compiler I'm less bothered. Distro needs higher standards.
19:28 birdwindupbird joined #phasers
19:29 mberends the situation is exceptional enough that I think we should drop the "time based release" principle until nom almost catches up to master in our judgement.
19:31 benabik joined #phasers
19:34 pmichaud I'll agree to that if there's a consensus agreement on that point... but I'm a little uncomfortable with it myself.
19:35 jnthn If we drop it, it removes a target date to aim towards
19:35 moritz agreed
19:35 pmichaud well, we've already missed our dates.
19:36 masak not having time-based releases would feel a bit demoralizing in itself, methinks.
19:36 pmichaud we can still make an August release, but it'll have to be made clear that it's development-only release
19:36 pmichaud I don't have a problem with that
19:36 moritz +1
19:36 masak +1
19:37 mberends +1 :)
19:39 pmichaud okay, that's what will happen.  I'll make an August release by tomorrow midnight, with whatever we have by then, and put lots of announcement text in place for it.
19:39 pmichaud I should have a draft announcement in the repo tonight.
19:40 pmichaud (so that people can review/wordsmith/improve prior to the release)
19:40 moritz +2
19:40 pmichaud within the next 24 hours, I will also do the branch renames.
19:41 pmichaud thus "nom" will become known as "master", and "master" will be renamed to 'ng"
19:41 pmichaud ..."ng"
19:41 jnthn +1
19:41 pmichaud I'll likely do that late tonight, when most of you aren't typically hacking on things :)
19:41 * moritz doesn't like 'ng'
19:41 * mberends really dislikes the name "ng"
19:41 moritz I'd prefer "beta"
19:41 jnthn Is "beta" better? :)
19:41 masak +1
19:41 mberends aye
19:42 jnthn I guess it comes after alpha :)
19:42 pmichaud I don't really want to establish an alpha-beta-gamma-naming sequence
19:42 jnthn I don't really want to have another such major refactor ;)
19:42 moritz "aleph" then :-)
19:42 pmichaud in particular, I don't want the current master to be known as a "beta" release.
19:42 masak "A"
19:42 jnthn pmichaud: ah, that's a good point.
19:42 masak indeed.
19:42 mberends "nqprx"
19:42 moritz if we run out of sccripts to chose the first letter from, we know we have to stop rewriting :-)
19:42 jnthn "centauri" :)
19:43 pmichaud We can always just refer to it as 2011.07, fwiw.
19:43 masak "alfalfa"
19:43 pmichaud it doesn't necessarily need a name beyond that.
19:43 moritz pmichaud: if we never plan to commit anything to it, that's fine
19:43 masak it's nice to be able to refer to it, though.
19:43 moritz (that's the difference between tag and branch)
19:44 masak 2011.07 is... somewhat clinical.
19:44 pmichaud I think that refering to it as 2011.07 is more accurate than anything else.
19:44 mberends the name should work nicely as directory name as well
19:44 * colomon kind of likes     alfalfa
19:44 masak fair enough.
19:44 pmichaud Instead of saying "are you using the 'beta' release"  or "are you using the 'alfalfa' release"  we can say "are you using the 2011.07 release?"
19:44 mberends I currently have alpha, master and nom directories
19:44 masak colomon: the naming sequence is scary to imagine, though :)
19:44 pmichaud which is really what we want to be asking and pointing to anyway
19:44 masak nod
19:45 moritz masak: you can also refer to it as Beijing if you prefer
19:45 pmichaud Q: "I'm looking for a stable version of rakudo."   A: "For now, you want the 2011.07 tag"
19:45 masak moritz: good point.
19:45 mberends moritz: +1
19:45 pmichaud or yes, Beijing also.
19:45 masak "alpha", "beijing". wfm.
19:45 pmichaud I'm fine with a "beijing" branch if we want that.
19:45 pmichaud although there's already a "Bejing" tag.
19:45 masak \o/
19:45 pmichaud er, "Beijing" tag.  so there might be some confusion there.
19:46 masak nah :)
19:46 pmichaud anyway, the features matrix purposely refers to that branch as "2011.07"
19:46 pmichaud and not "master"
19:46 moritz renaming it in the feature matrix is not a big deal, if need arises
19:47 pmichaud anyway, given the discussion above, I'm inclined to go ahead and use "ng" as the branch name, and hopefully we'll all refer to it by things like "Beijing" and/or "2011.07" when we need to talk about it.
19:48 moritz or maybe don't use a branch for now, since the head of the branch is identical to the tag right now
19:49 masak +1
19:49 moritz we can still create a branch when we want a change on top of it
19:49 pmichaud if we can do that, I'm fine with not creating a branch at all at this point.
19:49 moritz sure we can
19:49 pmichaud I was thinking it might be good to have a branch already created for future-proofing
19:49 moritz oh wait, not identiacl after all
19:49 moritz $ git rev-parse Beijing
19:49 moritz 6d4edbb2ed2da189fbbdfe08e431ba4f60775f00
19:49 moritz $ git rev-parse master
19:49 moritz 9225003fe4d3b0231f76a46f66e92628d06f4948
19:50 moritz but 'git log Beijing' and 'git log master' show the same things
19:50 * moritz confused
19:50 pmichaud "Beijing" like refers to a tag commit
19:50 pmichaud *likely
19:51 pmichaud there is no 6d4edbb commit in "git log" that I can find.
19:52 moritz aye
19:52 pmichaud anyway, can I just migrate existing "master" to "ng", as a safety?  We don't ever have to refer to it as such; we can use the tags for now.  And if we find we do need to refer to it and that "ng" is too bletcherous, we can always rename that branch again to something we like.
19:53 pmichaud personally, I plan to continue to refer to it as 2011.07, since that's really what it is and where it theoretically ends.
19:53 moritz ok
19:54 * moritz -> sleep
19:59 mberends ok
20:09 tadzik ooops, late /o\
20:14 mberends 166 hours early for the  next one ;)
20:15 tadzik oh, okay then :)
20:23 benabik left #phasers
21:01 mberends left #phasers
22:54 jlaire joined #phasers
22:54 jnthn joined #phasers
22:54 TimToady joined #phasers
22:54 tylercurtis joined #phasers
23:05 jlaire joined #phasers
23:05 jnthn joined #phasers
23:05 TimToady joined #phasers
23:05 tylercurtis joined #phasers
23:17 [Coke] joined #phasers
23:24 [Coke] joined #phasers
23:31 [Coke] joined #phasers
23:38 [Coke] joined #phasers
23:44 [Coke] joined #phasers
23:50 [Coke] joined #phasers
23:56 [Coke] joined #phasers

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