|
17:44
ilbot2 joined
|
|||
| moritz | now we're double logged | 17:44 | |
| paco | I have tell my client to log everything .. | 17:49 | |
|
17:51
NotFound joined
18:17
jhorwitz joined,
allison joined
18:22
jonathan joined
18:30
chromatic joined
|
|||
| chromatic | Greetings. | 18:31 | |
|
18:31
wknight8111 joined
|
|||
| jonathan | ahoyte | 18:31 | |
| allison | hiya | 18:32 | |
| pmichaud | hello. | 18:33 | |
| jhorwitz | yo | ||
| particle | ciao | ||
| Whiteknight | hello | ||
| chromatic | No Coke, so let's start. | ||
| allison? | |||
| allison | - I spent the week tracking down and fixing bugs in the Concurrency branch. | 18:34 | |
| - Added a 'branchcc' op that acts like a bsr but is continuation-based, so it's safe to use with Parrot subroutines and exceptions. | |||
| EOR | |||
| chromatic | I fixed several bugs. | ||
| I applied several patches. | |||
| I made CONST_STRING work in all PMCs (even dynpmcs and language PMCs) and am working on a caching system for truly constant const strings. | 18:35 | ||
| I have some other small fixes to PMCs in the works. | |||
| jhorwitz? | |||
| jhorwitz | completed an implementation of ModPerl6::Registry. | ||
| working on slides for YAPC/OSCON and publishing live examples on smashing.org | |||
| parrot revs after r27801 break mod_perl6 with method resolution errors. out of my league, so putting together a test case so others can debug. | 18:36 | ||
| EOR | |||
| chromatic | jonathan? | ||
| jonathan | * Finished with the talks for a while; ones in France seemed to be well received, and also gave one at Vienna.pm last night that I think was OK. | ||
| * Handed mutables work over to pmichaud, who did a great job of merging it all into trunk | |||
| * Did some patches at the airport on the way back from France, which mostly broke things and got reverted. | |||
| * One of them was for a GC bug that was causing errors in Rakudo, especially in interactive mode, and would probably hit other languages using lexicals too. My patch wasn't quite right - seemed to cause infinite recursion in the marking of contexts - so the GC bug still needs fixing. Not had chance to look at it again yet. | |||
| * Did some other little bits and pieces that worked out OK, including smart matching on junctions and preparing mutables to support ro (read only). | |||
| * Now talks are over, it's back to the hacking; will have Rakudo day later on this week. Planning to get role composition working some more, and also do the IMCC hacking required to get us :instanceof(...) adverb for subs. Plus get some tests in place for things that work, but aren't tested yet, so we don't get regressions. | |||
| * Got a grant from DeepText in Russia, to do an extra day a week on Rakudo during July, and will spend these extra days on trying to get MMD working the Perl 6 way. | |||
| EOR | |||
| chromatic | particle? | 18:37 | |
| particle | ~ coordinating the purchase of parrot.org, paperwork is now under review by tpf | ||
| ~ mainly working on rakudo's test system in the past week, applied patches from bacek++, moritz++ | |||
| ~ had trouble building parrot for a bit, turns out i needed a fresh checkout. no idea why | |||
| ~ hoping to concentrate on presentations and related commits this week | 18:38 | ||
| .end | |||
| chromatic | pmichaud? | ||
| pmichaud | ** Rakudo spectest_regression: 55 files, 638 passing tests (+16, +344 from last #parrotsketch) | 18:39 | |
| Nearly everything I did this week was Rakudo focused. | |||
| == Rakudo stuff | |||
| : Did major refactors updates of List and other classes | |||
| : List context now works (e.g., my @a = 1; my @b = (@a, (1, 2), 3); ) | |||
| : 'hash' composer now works ( my %h = hash(a=>1, b=>2); ) | |||
| : Applied many patches from contributors (bacek++, dolmen++, moritz++) | |||
| : Merged jonathan++'s mutable branch with trunk (still need some work there) | |||
| : Cleaned up variable handling in actions.pm | |||
| : Started refactoring parameter declarations | |||
| : *Lots* of other little improvements and changes | |||
| == PCT stuff | |||
| : minor improvements to PAST nodes | |||
| ** queue one question | |||
| ** plans for this week: more rakudo, close more RT tickets from the Rakudo queue, add more useful features. May get to PGE refactoring. | |||
| EOR | |||
| chromatic | Tene, are you here? | 18:40 | |
| Tene | I think so. | ||
| looks like I did some work on the cardinal class hierarchy and aggregates. | |||
| Started work on a smalltalk grammar, but ran into trouble debugging it. | |||
| .eor | |||
| chromatic | Whiteknight? | ||
| Whiteknight | * created gsoc_pdd09 branch. | 18:41 | |
| * set up some basic data structures for the new GC | |||
| * Prototyped pool initializers and new object allocators | |||
| * Setup basics of dod run algorithm and tricolor mark system | |||
| one question to log | |||
| EOR | |||
| chromatic | Did I miss anyone? | ||
| Okay. pmichaud has the first question. | 18:42 | ||
| pmichaud | allison: How much slower (or faster) is branchcc compared to bsr/ret? | ||
| (I'm not at all advocating we keep bsr/ret, I'm just curious.) | |||
| allison | it creates a continuation an then jumps to a lable | 18:43 | |
| so, slightly slower | 18:44 | ||
| chromatic | Except that it doesn't do as much GCable work. | ||
| pmichaud | still faster than a full subroutine or method call, though? | ||
| allison | but, if we do ever remove bsr (not deprecated yet) that means we can entirely eliminate the stack | ||
| much faster | |||
| pmichaud | also nice because I don't have to pass a lot of registers around. At any rate, allison++ for that | ||
| allison | aside from capturing a continuation it's just a 'goto' | ||
| pmichaud | end of question for me. | 18:45 | |
| chromatic | Whiteknight? | ||
| Whiteknight | GC-related question (obviously) | ||
| why do we hard-code GC initializer functions into src/gc/memory.c:mem_setup_allocator and not use a function pointer in Arenas like we do for deinit? | |||
| allison | no good reason | 18:46 | |
| Whiteknight | ...and if that changed...? | ||
| chromatic | We could get rid of that #ifdef tree. | ||
| allison | if it changed, and all the tests still pass, no complaints | ||
| Whiteknight | exactly, that's what I want to do | ||
| okay. end of question. | |||
| chromatic | My question: NotFound and moritz seem like they're well on their way to getting commit bits. | 18:47 | |
| Thoughts? | |||
| pmichaud | +1 | ||
| moritz | chromatic: jonathan invited me to one | ||
| already sent CLA | |||
| chromatic | That's easy enough. | 18:48 | |
| allison | +1 _1 | ||
| er, +1 +1, that is | |||
| pmichaud | bacek may also be a good candidate for a commit bit in the very near future | 18:49 | |
| chromatic | I'm not yet sick of applying bacek's patches... but I agree there too. | ||
| particle | we're a happy family | ||
| i've applied quite a few of bacek's patches as well, they're all sane | 18:50 | ||
| chromatic | Alright, I'll mentor NotFound. | ||
| Are there any other questions? | |||
| pmichaud | if moritz needs a mentor, I volunteer either myself or particle :-P | ||
| chromatic | Is anyone blocking on anything other than time, or a supply of fruit-flavored jellybeans? | ||
| pmichaud | I'm definitely low on fruit-flavored jellybeans. And time. Other than those items, no blockers. | 18:51 | |
| BTW, allison++ for quick resolution on named argument handling last week | |||
| (thank you.) | |||
| particle | i volunteer myself or pmichaud as well | 18:52 | |
| chromatic | Alright, let's wrap up this week then. Good work everyone. | ||
| pmichaud | reminder: parrot hackathon sat-sun june 14-15 | ||
| allison | thanks all! | ||
|
18:53
chromatic left
18:55
NotFound left
18:57
jonathan left,
cotto_work left
19:15
jhorwitz_ joined
19:16
Auzon left
19:25
allison joined
19:34
moritz left,
Whiteknight left
19:40
smash joined
20:29
davidfetter joined
|
|||