% particle_ has joined #parrotsketch
% particle_ is now known as particle
% leo has joined #parrotsketch
% allison has joined #parrotsketch
% particle has left particle!~particle@144.81.84.137
% particle has joined #parrotsketch
<chip> hello?
<leo> hi chip - hi all
<allison> hi
<particle> hi chip. waiting on a pm, i guess.
<allison> jesse mentioned he might not make the wed, meeting, but I think he expected to be at the mon. meeting
<obra> hey
<obra> sorry.
<obra> Got dragged by work
<obra> let's get cracking
<obra> Right then.
<obra> chip! how's things?
<obra> (or if chip's poppoed off for the moment, leo! how's it going?)
<leo> slowly
<chip> Yo dude!
<leo> - started hacking in the namespace pdd, e.g. this keyed syntax:
<chip> I've got some pdd21 (namespace) changes
<leo>   g = find_global ['Foo';'Bar'], 'baz' 
<leo> - not much else, distracted by $work, $family, and other $RL stuff
<leo> That was it
<obra> ok.
<obra> blocking on?
<leo> just my time
<obra> okie
<obra> chip: ok. sorry for the false start. go ahead
<chip> (sorry for the lag time, I'm really glad my cell phone serves as a workable modem but the lag time is measured in 5sec increments)
<chip> There are some issues to work on around the periphery of pdd21, like the idea of variable roots (analogs of the pwd concept in Unix), but basically it seems to be holding up under scrutiny
<obra> ok. what's up next?
% Coke has joined #parrotsketch
<chip> The PMC restructuring proposal is weighed and found wanting.  :-)  Too late in the arc to 1.0, and some of hte specifics are not really OK either (like eliminating the one-indirection vtable pointer)
<chip> I've got the other dev proposals to examine.  Oh, the $RL event for the day was a meatspace relocation
<chip> not blocking on anything but time.  thanks
<obra> ok. having fun?
<chip> one more thing to mention
% audreyt has joined #parrotsketch
<chip> the :non_volateile patch went in without review, so I asked it to be reverted.  
<chip> Can I ask a question out of order, against train connectivity issues that may hit ?
<obra> sure
<chip> leo: I thought it was standard that all register allocations are exclusive (or "non-volatile" as the patch said).  Now that the reg size is unlimited, anyway.  Not so?
<leo> no - temps are still reused
<chip> $P<n> is synonymous with "temps" right?
<leo> and if he wants a ptr to a temp, he needs such a flag
<leo> $Px or $Ix doesn't enforce a temp - it's more the usage of it
<chip> Does ".local" imply non-temp?
<leo> if $Px is used around a subcall, it's non-volatile (can't reused because of possible continuations)
<leo> .local gives it just a name and doesn't enforece (non-)volatile either
<chip> Hm.  Does the PIR compiler notice that a register is used as a lexical and automatically make the allocation non-temp?
<leo> just .lex forces a distinct reg, but that's ĂMC only
<leo> PMC only
<chip> OK then.  Sounds like the feature is good & necessary.  Will give it a blessed name and ask it to be recommitted
<chip> doen with Q
<leo> great - thx
<chip> exists $Parrot{fun} or die;  # SHOULD NEVER HAPPEN
<chip> ^D
<obra> ok
<obra> allison. 
<obra> how's things
<allison> pretty good, I worked on Punie conditionals this week
<allison> also on the operator precedence parser, long enough to try out Patrick's suggestion and send back the problems
<allison> I've also started a list of "hard learned lessons from large C projects" or "C Best Practices"
<allison> http://svn.lohutok.net/nam/trunk/parrot/docs/c_kwalitee.pod
<allison> Any additions are welcome
<obra> Ah. another ORA topic ;)
<audreyt> mm, I'm having a deja vu of certain chapters in Code Complete :)
<allison> actually, not intended for ORA (they'd never touch the topic), more intended for Parrot
<obra> Fair enough
<chip> hey, a rant repository.  :-)
<allison> audrey, not surprising, as many are simply common sense
<audreyt> *nod*
<allison> rant repository, wishlist, all are fair game
<allison> at this point its all greenlighting
<allison> not all are relevant to Parrot, but I'm including them anyway
<chip> are you taking mailed patches or asking for independent commits?
<allison> any and all contributions welcome, in any format
<obra> ok. whatcha blocking on?
<allison> time, of course
<allison> I'm trying to push a couple of things off my plate to free up more
* obra prepares to push more onto allison's plate
<obra> "ooh. free time to attack"
<obra>   
<allison> :) I'm clearing in anticipation :)
<obra> anything else new and exciting?
<allison> that's all for the moment
<obra> ok. particle: you're next. how's stuff?
<particle> nothing to report this week. tuits--. blocking on a kitchen remodel and other adventures in Real Life.
<obra> Fun! ok.
<obra> audreyt: how's your one year old?
<audreyt> doing fine; 6.2.11 expected in 48 hours for the anniversary
<allison> is a celebration planned?
<audreyt> triaging through 11018 subtests; down to <100 fails as we speak
<audreyt> allison: likely on #perl6, but probably not anywhere else
<audreyt> been reading papers on targetting register-based CMs
<audreyt> s/CM/VM/
<audreyt> a few weeks ago I asked leo if parrot has a good way to represent thunks
<audreyt> at that time there were none (had to use an explicit state variable in an outer closure, at a performance penalty)
<chip> kind of a general term, "thunk"
<audreyt> "delayed evaluation"
<audreyt> f() ?? g() !! h()
<audreyt> the g/h part 
<audreyt> but once they are evaluated, their value stays the same
<audreyt> (for the duration of that evaluation context, of course)
<audreyt> turns out SPJ and other lambdaheads discovered a clever way to mix lazy and eager evaluation in a single intermediate language, and compile them both to register-based VM without penalizing either side
<audreyt> so the crude simulation in the pugs->parrot backend can go away now
<audreyt> very happy about it
<audreyt> (details are on my journal under the "not not ml" paper)
<audreyt> the object model in PIL^N is solidifying very well
<audreyt> switching to object-based calculus (where binding and assignment are both just method calls) will occur right after 6.2.11
<audreyt> in time for perl5 and parrot targetting in the israel and vienna hackathons, respectively
<audreyt> end of report
<obra> blocking?
<obra> -ofun?
<audreyt> very much async.
<audreyt> -O9.
<obra> :)
<obra> Coke: you around?
<obra> and failing that: questions? other blockers? ideas?
<allison> birthday cake for Pugs!
<audreyt> a virtual one? :)
<obra> nah. just thousands of em
<particle> don't forget to email me a piece!
<allison> one in each city
<allison> I'll irc you a piece :>
<audreyt> /dcc particle piece-of-cake
<audreyt> /dcc send, even
<obra> okie. next week, then, everyone
<particle> mmm, cake. better than brains.
<audreyt> thanks obra
<obra> mm. braincake
<leo> thanks obra - afk a bit for dinner
<obra> thanks for letting me harrangue you, all
<audreyt> :) &
% audreyt has left #parrotsketch
% particle has left #parrotsketch
% Coke has left Coke!~wcoleda@host-216-153-142-58.alb.choiceone.net
% allison has left allison!~allison@ppp-71-139-1-208.dsl.snfc21.pacbell.net
% leo has left leo!~lt@feather.perl6.nl
% rafl has left rafl!~rafl@feather.perl6.nl
% chip has left chip!~chip@feather.perl6.nl
% rafl has joined #parrotsketch
% rafl has left rafl!~rafl@vs161067.vserver.de
% rafl has joined #parrotsketch
% rafl has left rafl!~rafl@vs161067.vserver.de
% rafl has joined #parrotsketch
% rafl has left rafl!~rafl@vs161067.vserver.de
% rafl has joined #parrotsketch
% rafl has left rafl!~rafl@vs161067.vserver.de
% rafl has joined #parrotsketch
% rafl has left rafl!~rafl@vs161067.vserver.de
% rafl has joined #parrotsketch