weekly Perl 6 status meetings with phase transitions: Wed 17:30 UTC | IR clogs at irclog.perlgeek.de/phasers/today
Set by moderator on 30 September 2011.
03:23 cognominal joined 03:24 cognominal joined
sorear Prereport: 15:43
DID:
* set out a design for an internal debugger API (like DB:: stuff), to be done this month if tuits allow 15:44
* went back and forth with TimToady to figure out the best handling of cross-platform line endings 15:45
* asked the P5 unicode maintainer about how best to go about doing \\p{Whatever} in P6. Reply in a nutshell: "Use the XML UCD" 15:46
* helped diakopter to get 6model/lua into a mostly working configuration
* fixed a few random niecza bugs, including thunky xx 15:48
* updated some of niecza's internals documentation
WILL DO:
* not much, severe tuit drought expected
* priorities: finish and implement debug interface, regex improvements including general Unicode character classes and an imitation of :graphs, native types, 6model convergence/implement roles 15:49
EOR
may even miss #phasers 15:51
16:55 colomon joined
colomon Pre-report: 17:02
* Got TagLib# working with NIecza.
* Wrote scripts to look for missing tags, and fix tags which can be deduced from the MP3 file's path. 17:03
* github.com/colomon/TagTools
* Blogged on this.
Plans: 17:04
* p6 work likely to be scanty for the next week due to real life suckiness. 17:06
EOR 17:07
pmichaud my report: Medical issues dominated this past week, so still running short of p6 tuits. 17:13
EOR
colomon pmichaud++ 17:20
jnthn Preport... 17:30
In the last week...
* Went to Twin City Perl Workshop and gave a couple of talks, including a new one on optimization stuff
* Used various bits of travel time to do optimizer design work
* Did a couple more optimizer implementation bits, but nothing major
* Eliminated need for nqp::want usage in native ops
* More work on bigint; optimistic we can have it merged before the next release
* Helped diakopter++ a bit with nqplua
In the coming week...
* Will go to London Perl Workshop and give talks there
* More bigint stuff
* Perhaps more optimizer stuff
* Figure out where I want to go with NCI things
RFC
* I'm pondering moving PAST over to 6model, and dragging it into the NQP repo as part of that. I know we intend to eventually replace it with QAST but... - I really want the memory improvements, like, yesterday - It's blocking getting nqpclr and nqplua transitioned over to new nqp; they still use parrot-nqp So unless anybody has any good reasons why not, I plan to JFDI.
EOR
o/ 17:31
tadzik \\o
moritz \\o
my report:
did a bit of C hacking on the bigint stuff
tried to do more exceptions stuff; blocks on not being able to stub joint-namespaced names in the setting 17:32
need to produce a minimal example and hand to jnthn++ for fixing
EOR
tadzik did: splice(), attempt at META.info parser, failing due to regex bugs. Friendly pair of eyes welcome to look at gist.github.com/1344631 eor 17:33
17:33 japhb joined
jnthn tadzik: Where does it fail? 17:35
tadzik: Should some of those token be rule?
japhb Did: Faster Str.trim*() Working on: Str.Numeric() rewrite. EOR
jnthn tadzik: As it stands, that grammar won't allow whitespace *anywhere*.
moritz fwiw I've started a 'nom' branch of JSON::Tiny 17:38
japhb moritz++
moritz it currently fails in the actions because the ~ parser generator currently surpresses captures
ie '(' ~ ')' <pairlist> makes <pairlist> not capture 17:39
jnthn moritz: Did you file a ticket for that one?
moritz not yet, I think
tadzik jnthn: that's fine, I'm not testing it on anything contatining whitespace
jnthn: fails in #46, $<pair> is nil
moritz: oh, that's similar to my bug 17:42
moritz: in my case <foo> ** ',' supresses captures too
jnthn Suggest filing ticket(s) and assign to pmichaud++ 17:43
moritz anyway, at least my problem is workaroundable
18:22 diakopter joined
diakopter did I miss it? 18:24
[Coke] seems to have died down about 45m ago. 18:26