Parrot 6.5.0 "Black-winged Lovebird" | parrot.org/ | Log: irclog.perlgeek.de/parrot | #parrotsketch meeting Tuesday 19:30 UTC
Set by moderator on 17 June 2014.
00:59 FROGGS_ joined 02:00 FROGGS__ joined 03:10 Khisanth joined 04:46 khisanth_ joined 07:44 FROGGS joined 07:56 basiliscos joined 09:47 Psyche^ joined 10:04 jsut_ joined 13:42 rurban1 joined 13:46 Coke joined 13:51 Util joined 13:57 bluescreen joined 14:45 bighugedog joined
rurban trying to benchmark now < 1.8.0 with some patches and --m=32 15:31
16:10 Chirag joined
Chirag rurban: Please take a look at my latest commit... 16:11
rurban ok 16:18
you mean the lt vs le? 16:21
it's only implementing o, not p. and it's far more complicated than that, so I implemented it by my own :) way too hairy for a GSOC student 16:22
you also need to check the arity, the param_count at run-time and error if wrong and decide :optional based on that
17:02 davidfetter_ joined 17:11 FROGGS joined
Chirag oh.. I see.. had no idea about that :) 17:53
so what should be my next step? our fill_params looked good already... 17:55
rurban testing my branch would be good 18:14
with lots of other langs (rakudo, cardinal, ...) 18:15
I'm still buried in fixing <= 1.7.0 to run benchmarks with the jit in earlier versions 18:16
needed patch for older versions: -#define NUM_REGISTERS 32
+#define NUM_REGISTERS 40
and my ploticus image creation script broke with the new ploticus version 18:17
Chirag ok.. I will test your branch.. so is the task almost over or could we optimize by replacing other calls? 18:20
rurban #2 is almost over, LGTM (looks good to me). I'm not sure about :constant yet 18:34
But we got again only 2%
Chirag hmm.. but weren't you getting ~4% by coding for _self alone? 18:40
18:45 Hunger joined
rurban yes, my benchmarking is obviously flawed 20:08
21:05 cooper joined 23:11 particle joined