weekly Rakudo status meetings with phase transitions: Tue 19:00 UTC | IR clogs at irclog.perlgeek.de/phasers/today
Set by moderator on 5 December 2010.
08:47 lue left, spinclad left, PerlJam left, TimToady left, moritz_ left 08:48 eternaleye left, [particle] left, sorear left, eternaleye joined, sorear joined, [particle] joined, spinclad joined, lue joined, TimToady joined, PerlJam joined, moritz_ joined 11:23 Util joined 13:25 [Coke] joined 15:12 colomon joined
sorear will miss #phasers. 15:16
done: translated a lot of stuff to Perl 6 in preparation for Niecza self-hosting. added pmurias to the team. 15:17
blockers: $reallife
15:38 PerlPilot joined 16:11 PerlJam left 17:34 PerlPilot is now known as PerlJam 18:40 pmichaud joined 18:48 takadonet joined
Util Pre-report: Made first cut of Rakudo Star OS X .dmg binary, Started pushing (no code yet) updates to Blue_Tiger. Caught bad cold or mild flu. EOR 18:54
moritz_ didn't do much Perl 6 related work, just some parrot bashing and answering questions 18:58
colomon Util++
PerlJam What's Blue_Tiger? 19:00
TimToady lurks harder 19:01
Util PerlJam: (on GitHub) my tinkertoy Perl5-to-Perl6 translator. Dusty code from 3yrs to 1.5 years ago, that I am slowly feeding into Git.
pmichaud report: I issued the 2010.12 star release, investigated moving rakudo.org content to a wordpress host, worked on setting up a new blog for myself, and provided some feedback to parrot devs about rakudo breakages 19:02
19:03 diakopter joined
diakopter I finished porting nearly all of the old perlesque functionality to the new perlesque. ready to start writing a viv backend to desugar STD/Cursor to perlesque. partially blocked on TimToady's STD_P5 19:05
(eventually) ;) 19:06
actually to be honest, Cursor should be translated first..
colomon -- Got the latest IO::Prompter and Testing modules from Damian, added them to the p6 ecosystem, added a few tests, and fixed a few minor issues. 19:07
-- Produced working solutions for all 5 of masak's problems.
-- Tested jnthn++'s caching patch, determined it caused new errors, and reverted it.
diakopter also maybe it's a good time for me to transition back to nqpclr (given I can get 15-30 minutes of pmichaud chat time at that time). EOR
colomon: sry :) 19:08
colomon I should be apologizing to you! :)
diakopter beat you to it! 19:09
moritz_ so, do we need to discuss anything? 19:11
TimToady what do people most need from me this week/month/year? besides the usual insanity... 19:12
moritz_ TimToady: Programming Perl 6, the book 19:13
PerlJam TimToady: a finished Perl 6 spec :)
I guess I should have said "finished"
TimToady okay, beside omnipotence and omniscience, what do you want? 19:14
moritz_ TimToady: I think the spec is progressing nicely in response to compiler writer questions 19:16
... and users, of course
Util Was the LTM implementation issue ever resolved? Is Rakudo using STD.pm to parse? Is this something that needs attention, or explaination so others can work on it? 19:17
moritz_ not sure what else we need from you, except the usual community hacking :-)
colomon If anything, we need to provide more implementation so that TimToady has something to iterate against...
Rakudo is not using STD.pm to parse.
moritz_ Util: sorear implemented LTM in niecza
rakudo "just" needs to steal that
TimToady rakudo could perhaps...yeah that
PerlJam TimToady: your continued presence on #perl6 seems quite useful. 19:18
TimToady diakopter++ requested STD_P5; anything else on that scale?
pmichaud given the latest discussion on perl6-users, I'm curious what we'd want to do to declare a "production" Rakudo release
or, alternatively, an "official" Perl 6 release 19:19
(besides yapsi, which is already official :)
Util TimToady: Your (and everyone's) work on RosettaCode has been quite helpful w.r.t. new-user introduction.
TimToady I think I prefer to think in terms of agreement by more than one implementation on what constitutes "possible to implement"
PerlJam pmichaud: I think having nice reference material prior to or concurrent with an "official release" is a must. 19:20
pmichaud PerlJam: I suppose. I'm not sure that's historically consistent.
TimToady when multiple implementations agree on that, even if they don't actually all implement it quite yet, then we can perhaps nail down some of the loose flounders
PerlJam pmichaud: times have changed. expectations have changed. 19:21
IMHO of course
pmichaud anyway, having some set of goals/targets for an "official" or "production" release might be good for Rakudo, along the lines of the ones we had in place for R*
TimToady I'd be really good at delegating that management task to someone else :) 19:22
pmichaud oh, I think I can manage it; it's just that this sort of thing really wants a lot of TimToady++ input, I think
TimToady I'm good at being responsive, but lousy at anything resembling instigation 19:23
colomon pmichaud: What sort of goals are you thinking of? Other than the obvious "it goes a lot faster". 19:25
pmichaud colomon: mostly, "What prevents us from calling it 'ready' now?"
TimToady well, speed, mostly... :)
plus a few little bugs like &foo not referring to the (possibly cloned) proto :) 19:26
and everything else...
pmichaud PerlJam suggests we also need better docs / reference material, which is reasonable
so, if we put together a list of those current "blockers", that'd be a good task list for moving to the next phase of development (hey, this *is* #phasers, after all) 19:27
TimToady I guess we need to figure out how 6.0 can be a complete subset of 6.1, so that people don't accidentally use 6.0 misfeatures that will break in 6.1
which they will...
PerlJam I'd like to see a general roadmap that fleshes out the possibilities between the release mile markers. 19:28
(i.e., we may have X by release A or release A+1)
pmichaud that sounds awfully.... vague
at this point I'm more interested in the candidate X's
TimToady to what extent is parrot refactoring the Big Blocker for rakudo?
pmichaud I don't know. I don't have the feeling that it's the Big Blocker 19:29
TimToady even on the speed issue?
pmichaud or, to the extent that it is -- well, jnthn++ is working on the solution, as opposed to us depending on others
ah yes, speed is a big of a critical issue
*bit
diakopter speaking of, pmichaud, do you think it's possible to wedge an LTM solution into a direct port of your nqp-rx engine in nqpclr (that I've made)? 19:30
pmichaud I need to understand the LTM solution, first.
but the answer should almost certainly be "yes, it's possible"
TimToady until the speed issue is resolved, we have to stay low-key on PR, or lots of people will take it as over-promising and under-performing
PerlJam indeed
pmichaud okay, then perhaps I'll table any thoughts of official/production until we have a good handle on speed 19:31
(and focus a bit more on speed, which is what I've been leaning towards working on anyway)
PerlJam that sounds like "official" is a year or so away to me.
pmichaud oh, I don't know if it's that far
TimToady 11 months and 21 days, maybe :) 19:32
PerlJam heh
colomon ;)
pmichaud I have another item for brief discussion, if we're done with that one
colomon Should we maybe try to come up with some goals that aren't speed to shoot for in an April-ish release?
Just to provide focus?
pmichaud if someone can come up with a list of the more major issues, we can see about assigning priorities to them 19:33
where "major issues" is from the "user perspective" as opposed to the implementors' perspectives
TimToady++ has already been providing a nice list
diakopter donde 19:34
pmichaud via comments in irc logs
diakopter oh :)
PerlJam The one thing that seems to irk me all the time is Parrot leaking through.
diakopter pmichaud: sorry if this is noise, but you saw the concat change in PIR? I think that might need some more nqp-rx tweaking than what cotto & bacek did 19:35
pmichaud okay, "reduce Parrot leakages"
diakopter: I think that nqp-rx is pretty well set to avoid the mutable concat... if not, it shouldn't be hard to fix 19:36
diakopter oh
pmichaud (I'm assuming that's what the concat change is.)
as a first cut, I'd expect to replace concat $S0, $S1 with concat $S0, $S0, $S1 19:37
diakopter assign only
pmichaud but there really shouldn't be many of the first form anyway
...assign only?
diakopter irclog.perlgeek.de/perl6/2011-01-04#i_3149194 19:38
I don't know, maybe the other one is allowed too
seems utterly silly to me, since one form is trivially converted to the other by the compiler. 19:39
pmichaud oh
the changes bacek made are what I just said
except he did $S0 = concat $S0, $S1 instead of concat $S0, $S0, $S1 (the two are exactly equivalent) 19:40
diakopter as I postulated above, also.
but the deprecation still seems silly to me.
oh well. 19:41
pmichaud anyway, I just tried building/running nqp-rx with parrot trunk and didn't encounter a concat issue.
so I think it's okay.
diakopter cotto was saying he didn't have trouble building, but he did have trouble bootstrapping
pmichaud I didn't try that. There are other bigger problems in parrot trunk that are causing issues atm. 19:43
diakopter TimToady: could you perhaps distill the irc comments to which pmichaud referred into a list in some git repo? 19:44
or someone else who might know where they are 19:45
19:45 smash joined
diakopter so we don't all have to search for them 19:45
pmichaud maybe search for RT tickets that contain TimToady :-)
diakopter but... ok. 19:46
pmichaud on another topic -- last week I started investigating moving rakudo.org content to wordpress (more) 19:47
the current rakudo.org website doesn't offer an RSS feed, and has been a horrible haven for spam users and comments
and we're basically bottlenecked by alester's tuits (more) 19:48
so, I started putting that together, then alester++ cleaned up rakudo.org
(although afaik we still don't have an rss feed)
19:48 Tene joined
pmichaud so, since alester++ seems to be progressing on rakudo.org again, I'm planning to hold off on any move to wordpress. but if things fall behind again, I might pick it back up. 19:48
any comments?
PerlJam +1 to moving especially if we can increase the bus/truck number. 19:49
pmichaud oh, undoubtedly we can increase that 19:50
one challenge might be that alester owns the rakudo.org domain -- so we might have to use a different domain if we move
(I have rakudoperl.(com|org|net) so those are available) 19:51
diakopter he's not willing to point it elsewhere?
pmichaud I don't know if he is or not.
when I asked about it in the (longer ago past) I detected some resistance
TimToady btw, there was some indication that wordpress might be blocked by the GCF 19:54
pmichaud GCF?
TimToady or is it the GFC?
the Bamboo Curtain of the internet... 19:55
pmichaud you mean the great firewall?
TimToady someone was concerned about it shortly after you announced the wordpress idea 19:56
diakopter wordpress blog software? or wordpress.com/org ?
pmichaud wordpress.com, apparently. latest notes seem to indicate that it's not blocked 19:57
still, it could end up being that way
oh, reported blocked again as of nov 20 19:59
anyway, I'll keep that in mind -- good to know.
diakopter (but you can host your own wordpress software, afaik) 20:00
pmichaud ah, one report on oct 15 says it's actually the domain *name* (.wordpress.com) that is blocked. Other domains hosted on wordpress.com servers still seem to get through okay, which would be the case for us 20:01
TimToady ah, yes, it was JimmyZ at irclog.perlgeek.de/perl6/2010-12-31#i_3135264
pmichaud well, if the choice is between "full of spam and no rss feed" and "not accessible in china due to political constraints" -- I'm likely to go for the latter. (But this is likely a false choice anyway -- we may have other options.) 20:03
anyway, we'll see how rakudo.org fares over the next few weeks
PerlJam pmichaud: so, what are you doing rakudo-wise between now and April-ish? Is there anything you need from the rest of us? 20:07
pmichaud 1. benchmarking/speed
2. list documentation (grant)
3. support jnthn++ and others working on nqp ports 20:08
4. review/fix bug tickets
eol
I should also note (somewhat depressingly) that my wife and I suspect we'll have another health-related interruption sometime in the first half of 2011
colomon :( 20:09
pmichaud indeed, there were a few days in december where we thought we might have problems before christmas (but everything is looking better again since then)
she has another test scheduled for friday, which should give us a lot more information 20:10
(results won't be back until early next week)
anyway, my primary goals are to show progress again on reported bugs and on speed
oh, and 5. work on cleaning up perl6 book
PerlJam your #5 is always on my list but I never seem to get around to doing it. 20:12
TimToady well, just don't get into the frame of mind where whatever you're doing, you feel like you're cheating someone else 20:13
colomon I've very willing to do anything I can to help with benchmarking/speed, especially if that will help take some of the load off of you.
pmichaud colomon: I'll likely start putting together a set of benchmarks over the next few weeks; having additional tests and tools for summarizing results are likely where I can use help 20:17
20:19 tylercurtis joined
Util The #parrotsketch meeting looms for some of us. Any final issues before we adjorn here? 20:24
colomon pmichaud: Just let me know what you need, and I'll do my best to help. 20:25
pmichaud will do 20:26
I have no further issues
Util Thanks, everyone! We are adjourned. 20:29
20:32 diakopter left 21:05 dukeleto left, dukeleto joined 21:08 takadonet left 21:25 lue left 21:49 tylercurtis left

Powered by Cro and the Raku® Programming Language.

Please report any issues / comments / feature requests as an issue on App::Raku::Log.

Thank you!