% particle has left particle!~particle@c-24-19-3-148.hsd1.wa.comcast.net
% particle has joined #parrotsketch
% pmichaud has joined #parrotsketch
% rdice has joined #parrotsketch
% contingencyplan has left contingencyplan!~contingen@cpe-76-186-27-146.tx.res.rr.com
% contingencyplan has joined #parrotsketch
% cognominal has left cognominal!~cognomina@82.67.232.89
% cognominal has joined #parrotsketch
% contingencyplan has left contingencyplan!~contingen@cpe-76-186-27-146.tx.res.rr.com
% contingencyplan has joined #parrotsketch
% contingencyplan has left contingencyplan!~contingen@cpe-76-186-27-146.tx.res.rr.com
% contingencyplan has joined #parrotsketch
% kj has joined #parrotsketch
% kj has left kj!~IceChat7@ip565fd420.direct-adsl.nl
% particl1 has joined #parrotsketch
% particle has left particle!~particle@c-24-19-3-148.hsd1.mn.comcast.net
% particl1 is now known as particle
% jonathan has joined #parrotsketch
<jonathan> My report early, since I gotta go now and can't be around for #ps
<jonathan> * Did various Rakudo hacking, including type annotations on variables and parameters, a couple of cases of the handles trait verb and refactoring the type hierarchy to have Any and have Int etc inherit from there, and so forth. Also some initial work on captures, and I did .= but it's got a bug...will fix very soon, if it's what I think it is.
<jonathan> * Hacked on PCT to make setting flat actually emit :flat
<jonathan> * Made various PMCs work better when subclassed as a high level object
<jonathan> * Helped a copule of people get Parrot and Rakudo built in their laptops. Watched at how fast their laptops built it compared to mine.
<jonathan> .EOR
% cotto_work has joined #parrotsketch
% kj has joined #parrotsketch
<kj> I'm reporting early:
<kj> little to report, except that all solutions to the exercises of the PCT tutorial are posted now.
<kj> .end
% kj has left #parrotsketch
% Coke has joined #parrotsketch
% Infinoid has joined #parrotsketch
% Senaka has joined #parrotsketch
% chromatic has joined #parrotsketch
<chromatic> Hello everyone.
<pmichaud> hola
<cotto_work> hi
<particle> hi-o
<particle> jonathan pasted his report earlier today
<particle> so did kj, just 15 mins ago
<chromatic> Okay, who else is here?
* Infinoid <--
% rgrjr has joined #parrotsketch
<chromatic> Let's go in alphabetical order then.
<chromatic> I fixed a couple of bugs.
<chromatic> Mostly I worked on the Test::Class PIR library.
<chromatic> Coke?
<chromatic> cotto_work?
<cotto_work> ok
<cotto_work> * no more info on the legal situation (boss is unexpectedly out of town for the week)
<cotto_work> * PHPArray more or less works with other PMC types
<cotto_work> * also pointed out a couple bugs to Infinoid
<cotto_work> eor
<chromatic> Infinoid?
<Infinoid> * Minor rakudo tweaks, implemented %hash.keys() and %hash.values().
<Infinoid> * Codingstd tweaks, fixed some warnings, applied some patches from RT.
<Infinoid> * Reorganized the ops2c code and its output.  Things are in a better order now, which helps C++.
<Infinoid> 1;
<chromatic> particle?
<particle> ~ met with allison this weekend. very productive planning discussion, no lines of code written
<particle> ~ busy period for gsoc, tpf got ~25 submissions, and ~16 look really good. 4 of those are parrot- or perl 6-related. we'll find out this week how many slots we get from google.
<particle> ~ please help me prep for next week's release by keeping trunk passing. great work so far this month.
<particle> .end
<chromatic> rgrjr?
<rgrjr> * updated some doc refs to find_global and store_global.
<rgrjr> * flushed the venerable queens.pasm example.
<rgrjr> * finished continuation barrier writeup (posted for comment).
<rgrjr> * started an update (a rewrite, almost) of docs/pmc/subs.pod.
<rgrjr> * started laying the groundwork for deprecating the stack ops.
<rgrjr> .eor
<chromatic> Tene?
<Tene> I'm fairly certain I didn't do anything parrot-related this week.
<chromatic> tewk?
<chromatic> pmichaud?
<pmichaud> Mostly worked on closing older tickets (think I got about six of them)
<pmichaud> more of those to come
<pmichaud> also worked on getting travel arrangements in place for yapc::na and oscon
<pmichaud> still reviewing code changes, which are happening quickly
<pmichaud> wrote a status report for the mofo/tpf grant, posted to use.perl journal
<pmichaud> I have two items for discussion today
<pmichaud> eor
<chromatic> Let's go to question time then.
<chromatic> pmichaud?
<pmichaud> the yapc::na parrot hackathon will be the weekend before yapc -- i.e., jun 14+15
<pmichaud> there's also been discussion of a "parrot workshop track" at yapc, which I think is a neat idea
<pmichaud> I've been thinking that perhaps the hackathon and/or workshop should have some "structured events" in place
<pmichaud> i.e., set times where we say "people who come will be engaged in doing xyz"
<chromatic> I think that's what kid51 was working on with the organizers.
<pmichaud> this could be things such as "building parrot, building rakudo, building a small language using pct, etc."
<pmichaud> almost tutorial-like
* Infinoid grabs a question-ticket
<pmichaud> opinions?  should we pursue it, is it unlikely to work, other?
<pmichaud> initially I'm thinking we could flesh out ideas for things on the parrot wiki, then post a schedule to the yapc::na wiki 
<Coke> should we pursue the structured events? I've asked if we can get some kind of feedback from the organizers so we better plan out the workshop.
% diakopter has joined #parrotsketch
<pmichaud> this wouldn't need to absorb all hackathon participants, and I'm not wanting to fill the available time -- there should still be plain random hacking time
<pmichaud> but for people who are interested in hacking but want something more than "show up and we'll find something for you to do", it might work better
<Coke> ... I thought we were talking about the workshop.
<pmichaud> I'm talking about both
<pmichaud> more structured things at the workshop (and working around the speaking schedule and other tracks), but not leaving the hackathon itself completely "loose"
<chromatic> Makes sense to me.
<Coke> +1
<pmichaud> anyway, I'll flesh it out a bit more on the mailing list and wiki and we'll see if it goes anywhere
<rgrjr> Hackathon before workshop is inconvenient, alas; makes it harder to recruit.
<pmichaud> rgrjr: there are two hackathons at yapc::na
<pmichaud> there's the parrot hackathon the weekend before, and a "general hackathon" on thu-fri after
<pmichaud> I think the parrot hackathon went before because many folks can't make it the thu+fri after
<Coke> we can have a post-con hack; just that's going to miss several of us.
<Coke> so, rgrjr, if you're going to be there post, get cracking. =-)
<pmichaud> well, the post-conference hackathon could be loose
<rgrjr> pmichaud:  good, then.
<pmichaud> and of course, nothing prevents folks from working on parrot at the post-hackathon
<pmichaud> okay, 2nd item
<pmichaud> obra asked me a week or so ago about the possibility of getting some sort of public pass/fail display for rakudo's tests
<pmichaud> i.e., similar to what we have for "make smoke"
<pmichaud> last week as part of reviewing rakudo's status I decided this is *really* important -- more so than I initially thought
<pmichaud> independently bernhard brought up the point that compilers/nqp isn't in the standard 'make test' sequence, and should be
<pmichaud> and we've had a long-standing issue with getting language tests to run from the main prompt
<pmichaud> so I think we really need a focused effort to fix parrot's testing system
<pmichaud> the many-little-harnesses-and-test-frameworks is starting to break down
<pmichaud> (or has broken down)
<pmichaud> anyone agree, disagree, or have suggestions for how to proceed?
<pmichaud> I'm willing to contribute tuits (because I think it's that important)
<Coke> it's been an open issue for some time. We can definitely focus on it in the short term.
<Infinoid> its just a question of where the tests go, right?
<particle> we need a working harness for the release
<pmichaud> Infinoid: no, it's more than that
<chromatic> It's also a question of how to *run* the tests.
<Infinoid> I mean, imcc and nqp are both built as part of the main "make"... yet imcc's tests are in "make test"
<particle> and current == working
<pmichaud> because different tests are written in different languages and require different runtimes
<chromatic> Many test scripts are not Perl programs.
<chromatic> Not all test scripts have shebang lines.
<pmichaud> particle:  we wouldn't break the existing harness until the new one is ready
<pmichaud> particle:  i.e., I'm thinking it would proceed in a branch
<chromatic> Something needs to know 1) which program to use to run the test and 2) what command line to use to invoke #1
<Infinoid> is someone working on a new harness?
<particle> pmichaud: i'm all for it. this itch needs scratching
<pmichaud> Infinoid: I think there have been several initial attempts, but I don't know the current status of them all
<particle> take a look at colin's branch and see if it's a good start
<Infinoid> I would like to contribute to the new harness project.
<particle> there's branches/smoke, and branches/unified_testing/
<particle> the latter is colin's work (and andya, ewilhelm, etc)
<pmichaud> I'm thinking that we rewrite t/harness anew, and perhaps make it TAP::Harness based
<Infinoid> cool, I'll check it out.
<particle> sounds like this can move to the list now
<chromatic> allison just apologized offline for not making it to the meeting.
<chromatic> Coke, did you have a report?
<pmichaud> okay, I'm not hearing any opposition or "watch out for XYZ" notes, so we can proceed to next item.
<pmichaud> thanks
<Coke> chromatic: yes.
<chromatic> Go ahead.
<Coke> fixup build deps, rip out deprecated stuff, proposed simplification to ops syntax."
<Coke> .
<chromatic> Thanks.
<particle> chromatic will write us a parser for .ops
<chromatic> When are you next within fork-stabbing range?
<chromatic> Infinoid, do you still have a question?
<Infinoid> yes
<Infinoid> diakopter and I are thinking about making a parrot evalbot, so you could tell it to evaluate a line of, e.g. tcl or lolcode.  would this be useful to have in #parrot?
<chromatic> +1 from me
<Infinoid> I CAN HAS EVALBOT?
<rgrjr> security issues?
<Infinoid> we'd have to sandbox it somehow
<particle> chroot jail etc
<Infinoid> probably impose some resource limits on time and memory size, too.
% Senaka has left Senaka!~senaka@124.43.227.54
<Infinoid> it'll run on feather
<particle> the security pdd is still a few months away, so we can't do it safely using pir atm
<particle> it'll run on a vm on top of feather
<Infinoid> all details aside, if its useful, we will see what we can come up with.
% Senaka has joined #parrotsketch
<rgrjr> sounds good.
<Infinoid> ok, thanks!
<Coke> +1
<rgrjr> Q:  Is anyone thinking about a PIR API for symbolic debugging?
<particle> i'm dreaming about one right now :)
<rgrjr> me too . . . but so far they are pretty vague.
% rdice_ has joined #parrotsketch
<chromatic> Are there any other questions?
<chromatic> Blockers?
<particle> no blocking incidents to report.
<Coke> could use some feedback on my recent ops stuff.
<Coke> (and I owe kid51 feedback on his recent stuff..)
<chromatic> Is anyone looking into 64-bit problems and C++ compiler problems?
<Infinoid> I am
<Infinoid> Senaka tried to build with g++, and found lots of issues
<Infinoid> we're knocking them out, slowly.
<Infinoid> (that's what prompted the ops2c stuff.)
<Infinoid> and I'm testing regularly on amd64, fixing warnings as I find them
<particle> still can't run 64bit parrot built with msvc
<particle> recent stacktrace here: http://nopaste.snit.ch/12650
% rdice has left rdice!~richard_d@CPE0014bfafbbd5-CM0011e6ecf48a.cpe.net.cable.rogers.com
<chromatic> Could MSVC stacktraces be any uglier?
<Infinoid> particle: I still don't fully understand that, but if the pointer size is determined by the data size on win64, can we fix that up by being more careful about the data types we toss around?
<particle> Infinoid: yes
<particle> lemme get you the output of a sizeof program for basic types
<Infinoid> thanks
<Infinoid> we also have some reports of dismal failure on pa-risc
<chromatic> Probably casting and alignment woes again.
<Infinoid> alignment
<chromatic> Are there compiler warnings for such?
<particle> pa-risc has some funky alignment rules
<Infinoid> compiler errors, in fact
<Infinoid> http://nopaste.snit.ch/12643
<chromatic> Those should be fixable then.
<Infinoid> oh, I'm sorry, those are warnings
<chromatic> Pointer size assumptions, yay.
<chromatic> Is there anything else to discuss this week, or shall we adjourn?
<chromatic> Okay, see you all next week.
<Infinoid> thanks, chromatic
% Infinoid has left #parrotsketch
% rgrjr has left #parrotsketch
% chromatic has left #parrotsketch
% Senaka has left #parrotsketch
% diakopter has left #parrotsketch
% rdice_ has left rdice_!~richard_d@CPE0014bfafbbd5-CM0011e6ecf48a.cpe.net.cable.rogers.com
% rdice has joined #parrotsketch
% jonathan has left #parrotsketch
% cognominal has left cognominal!~cognomina@82.67.232.89
% rdice has left rdice!~richarddi@CPE001217e365c7-CM00159a01d44c.cpe.net.cable.rogers.com
% contingencyplan has left contingencyplan!~contingen@cpe-76-186-27-146.tx.res.rr.com