% rdice has joined #parrotsketch
% pmichaud has joined #parrotsketch
% kjs has joined #parrotsketch
% chromatic has joined #parrotsketch
% rdice has left rdice!~richard_d@CPE0014bfafbbd5-CM0011e6ecf48a.cpe.net.cable.rogers.com
% Coke has joined #parrotsketch
% barney has joined #parrotsketch
<Coke> Ding.
<chromatic> You're a minute early.
<Coke> Should we be expecting allison? 
<chromatic> I dunno what time it is in Capetown.
<chromatic> eleventy four hundred hours or something?
<pmichaud> it should be about 7:30
<pmichaud> er, 19:30
<barney> Should be half past seven, same as here
<chromatic> We should start now anyway.
<pmichaud> or maybe that's 19h30
<Coke> now that I'm a minute late instead, chromatic, you go first.
<chromatic> I fixed a handful of bugs, mostly little polish things.
<chromatic> I did check in a tiny optimization that stayed in.
<chromatic> Thanksgiving holidays mostly killed my productivity, but December should be much better for me, so if you have tickets for me to look at, let me know.
<Coke> anything else?
<chromatic> That'si t.
<Coke> * cleanups in languages - killed basic/interperter, cleaned up *STATUS, fixed bug in Lisp, id'd issue in cola which mrjoltcola is looking at.
<Coke> * RT queue maintenance
<Coke> * Some updates to http://www.perlfoundation.org/parrot/index.cgi?useful_rt_queries in an effort to start using the faux [TAGS] we've been adding to tickets. Point people looking for work to there.
<Coke> </>
<Coke> barney?
<barney> Added stub for languages/eclectus.
<barney> Worked on making languages more selfcontained.
<barney> .eor
<Coke> kjs?
<kjs> worked on back-end of pirc/new. Goes fairly well, first PASM instructions are popping out
<kjs> still needs more work, and clean up after that. Once that is done, I should have a look at register allocation and/or PBC file format
<kjs> .end
<pmichaud> ECOKETIMEOUT ?
<chromatic> Yep, go ahead.
<pmichaud> * Worked with chromatic++ to get 'perl6' working in time for 0.5.0 release
<pmichaud> * PCT is now mostly developed
<pmichaud> ** all the basic node types work
<pmichaud> ** still missing: repeat_until, try, chain
<pmichaud> ** also I have a couple of code optimizations I want to try
<pmichaud> * NQP is mostly developed
<pmichaud> ** added namespace support
<pmichaud> ** added module/class keywords
<pmichaud> ** added 'make' keyword
<pmichaud> ** added load_bytecode function -- this may become a load_bytecode keyword
<pmichaud> ** notably, nqp is now able to run blue_rect.pl and exercise SDL
<pmichaud> * I ended up moving nqp into compilers/nqp/ .
<pmichaud> * Rewrote 'abc' in nqp instead of TGE+PIR
<pmichaud> ** Added lots of documentation to the transformation file (feedback welcome)
<pmichaud> * Added a couple of new RT tickets, resolved/answered a few others
<pmichaud> * PerlJam++ submitted a patch to update Perl6Regex syntax in PGE
<pmichaud> * Added a ROADMAP document to perl6
<pmichaud> * Started blogging about this stuff in my use.perl.org journal
<pmichaud> .end
<chromatic> Anyone else we missed?
* Coke returns.
<chromatic> Any questions?
<Coke> time for questions. I have 2.
<Coke> chromatic: please work on http://rt.perl.org/rt3/Ticket/Display.html?id=47816 ?
* pmichaud queues one question
<barney> There are a few TODOs in docs/ROADMAP.pod. Should they be filed in RT ?
<Coke> also, can we remove more dead branches. Should be a CAGE task to ID whether any pending development is left...
<Coke> barney: in general, yes. ROADMAP needs to be revamped, getting things to have tickets is a good start.
<Coke> pmichaud: ?
<barney> My suggestion is to move old branches to branches_old.
<Coke> I disagree.
<pmichaud> I'm attempting to do almost daily blog entries about the development of perl6 and the compiler tools.  Should I leave those only in use.perl.org, or should I repost them to the mailing lists?
<barney> I wouldn't want to remove other peoples stuff.(even if it's in history)
<Coke> pmichaud: just send a reminder to the list so folks can add you to their aggregator. (you should be added to planet parrot)
<Coke> barney: it's all our stuff. it's in history. they can always get it back if they need it.
<pmichaud> coke:  okay, thanks.  That leads me to question #2, which is totally non-parrot related:  what's a good aggregator, since I don't have one?  ;-)
<pmichaud> many of the branches have been there for years
<Coke> very old, stale. Chip had a good quote about this in the ROADMAP, actually.
<chromatic> pmichaud, I like Akregator on Linux.
<Coke> the more cruft there is to wade through, the less likely people can find the stuff that's useful and interesting.
<pmichaud> chromatic: okay, thanks.  I'll start there.
<kjs> I have one question for Pm
<Coke> "Eliminate dead documentation.  Let historians dig through our subversion
<Coke> history.  We look forward toward successful deployment.
<Coke> "
<Coke> having heard no more questions, go ahead. =-)
<kjs> A minor thing. pmichaud: you mentioned "load_bytecode" as a possible function
<kjs> or statement?
<pmichaud> currently it's implemented as a function.  I may turn it into a statement.
<kjs> anyway, isn't load_bytecode more or less doing what Perl's "use" is doing? 
<kjs> so, as NQP is perl-like, why not call it "use"?
<pmichaud> yes, but there are some notable differences.
<pmichaud> at some point I may actually write it as 'use', but if they have some substantial differences I'd prefer to leave it as load_bytecode for now
<kjs> ok. just a thought. Writing underscores in hlls looks low-level :-)
<pmichaud> what I'd really like to have is a Perl 6-like syntax to access Parrot opcodes
<chromatic> "eval"
<pmichaud> does eval have adverbs on it that could indicate the language being used
<pmichaud> ?
<kjs> like some c compilers allow assembler in a _asm { } block?
<Coke> pmichaud: something like PIR::inline { ... }
<Coke> (taking a hint from the similar perl5 modules...)
<Coke> Tcl has [inline PIR { ... raw pir ... } ]
<pmichaud> PIR::inline has some potential.
<pmichaud> I know $Larry was looking at some ways to do per-language inlining with heredoc syntax... but I don't know if anything happened with it
<Coke> alrightee. On a related note, if anyone can think of actual project-managery things that need doing, feel free to bug me about it in #parrot or via email.
<Coke> (UN related note, that is.)
<Coke> Would someone like to volunteer to send a link to the log post-sketch to the mailing list ?
<pmichaud> aha, eval *does* have a :lang attribute!
<Coke> (on a regular basis?)
<kjs> can't that be automated? :)
<pmichaud> I'll post it
<pmichaud> I'll do it as part of my various posts to the mailing lists and journal
<Coke> danke.
<pmichaud> we've lost our Perl 6 summarizer, so I may take up that task (as far as compiler progress goes) and can include the link there
<Coke> At this point, your journals will fit the bill nicely.
<Coke> Anyone blocking on anything other than time?
<pmichaud> I'm a bit concerned about lexical handling in Parrot
<pmichaud> but I think I'll post to the m/l
<chromatic> I'd like to optimize Lua some, but I still can't figure out how to run individual programs from languages/lua.
<pmichaud> in particular, :outer() and 'newclosure' confuse me
<Coke> pmichaud: ok.
<Coke> list is probably the best place for those, yes.
<Coke> chromatic: good question.
<barney> Another Q: Has the 'unified_testing' branch stalled?
<chromatic> Good question.  I patched Parrot::Test::Harness to use TAP::Parser where applicable, and it's nice.
<Coke> chromatic: ah, perhaps you're the one that broke all my tests...
<pmichaud> I was just thinking about that.  :-)  Since nqp is now in compilers/nqp/ and is built as part of the standard build process, it would be nice if nqp's tests could be part of the standard 'make test' suite
<Coke> pmichaud: sure, just move compilers/nqp/t to t/compilers/nqp
<pmichaud> coke:  I thought there were issues with having per-language tests 
<pmichaud> have those been fixed?
<Coke> oh, nqp is special.
<Coke> let's get unified testing working first, I guess. barney, do me a favor, since you asked here... ask on list for a status update. =-)
<barney> will do
<Coke> also, I heard some microgrants are still available for perl6.
<chromatic> I need to go.  See you all later.
<Coke> ~~
% chromatic has left chromatic!~chromatic@208.14.131.162
<Coke> Anyone else have anything?
<pmichaud> yes, I believe they are.  unified testing would be an _excellent_ microgrant
<Coke> Let us both do the proper encouragement via email once we see barney's post.
<Coke> Ok. Calling it. See everyone next week.
<pmichaud> see you next week!
% pmichaud has left #parrotsketch
% Coke has left #parrotsketch
% allison has joined #parrotsketch
% natacha29 has joined #parrotsketch
% kjs has left #parrotsketch
% natacha29 has left natacha29!~natacha29@ANantes-257-1-88-137.w90-25.abo.wanadoo.fr
% pmichaud has joined #parrotsketch
% Andy has joined #parrotsketch
<Andy> <burp>
<Andy> scuse me
<allison> My report for the logs (late from a time zone confusion):
<allison> - I've been working on the implementation of Events for the next milestone.
<allison> - Since events are the first system to use the newly spec'd concurrency scheduler, an initial implementation of that is the first step.
<allison> - Also filling in some practical details in the concurrency PDD as I work on the implementation.
<allison> EOR
% allison has left #parrotsketch
<Andy> I guess I must be late, too.
% Andy has left #parrotsketch
% rdice has joined #parrotsketch
% rdice has left rdice!~richard_d@CPE0014bfafbbd5-CM0011e6ecf48a.cpe.net.cable.rogers.com
% barney has left barney!~bernhard@p549A2DD4.dip0.t-ipconnect.de
% particle has joined #parrotsketch
% pmichaud has left #parrotsketch
% rdice has joined #parrotsketch
% rdice has left rdice!~richarddi@CPE001217e365c7-CM00159a01d44c.cpe.net.cable.rogers.com
% rdice has joined #parrotsketch
% rdice has left rdice!~richarddi@CPE001217e365c7-CM00159a01d44c.cpe.net.cable.rogers.com
% tewk has left tewk!~tewk@ekstrom.org