Index

Show enters and exits. Hide enters and exits.

01:16:15slavahi evan
01:16:19slavanice blog post about jit
06:12:47kstephenshttp://github.com/kstephens/rubinius/commit/c59a3e56590b1863b523f615f6ec860e01c95923
06:12:56kstephensMakes sprintf.rb about 29% faster.
06:13:20kstephenshttp://github.com/evanphx/rubinius/issues/issue/207
06:14:32kstephensI think if we could get rid of all the *args splats and rest args, it could be faster, but the spec are preventing me. :(
06:17:05kstephensThe mocks in modulo_spec.rb seem kinda weird in regard to #to_ary, etc.
07:50:44boyscoutPatch from postmodern for LLVM prebuilt checksums. Closes #202. - 7eb9cc9 - Brian Ford
07:51:16postmodernw00t
07:53:12brixenheh
07:56:00boyscoutCI: rubinius: 7eb9cc9 successful: 3037 files, 12072 examples, 36356 expectations, 0 failures, 0 errors
08:22:35evankstephens: you around?
08:27:32dbussinkevan: playing with his benchmarks / crashes?
08:27:41evanjust read the patch
08:27:50evanit would appear he made a bunch of stuff slower
08:27:51evannot faster.
08:28:14evanactually.. mm.
08:28:41evanat the very least
08:28:45evanall that .freeze stuff should go.
08:28:50slavasup evan
08:28:55evanslava!
08:32:49slavawhat's new?
08:33:33evanup in SF
08:33:35evanchillin'
08:33:37evanyou?
08:35:00slavachillin' :)
08:35:08slavajust got back from a trip around the north island
14:58:53kstephensevan: really its slower? How did I do that?
15:10:52kronos_vanoThe main problem in String#% is regexps. We should avoid them
16:20:59kstephensShould we just port the MRI sprintf.c?
16:22:23kronos_vanokstephens, We should just make String#% as fast as possible :)
16:23:24kstephenswhich might be using C? :) sprintf is a highly abused Ruby function.
16:23:38kstephensor used Ruby function.
16:24:12kronos_vanokstephens, as fast as possible in ruby
16:24:58kronos_vanoafter that we shoud take a look at benchmarks, and make a decision about moving it in c++ =)
16:26:48kstephensIt looks like some of the regexp could just be a check on the leading char, then parse the remaining chars, if needed.
16:27:51kronos_vanokstephens, we dont need regexp at all I think
16:27:55kstephensyea
16:33:26kronos_vanokstephens, in String#% I mean
18:31:05sbryantevan: would you mind if I put together a presentation for rubinius and give the presentation at the local ruby users group?
18:31:16evannot at all
18:31:19evango right ahead
18:31:40sbryantAwesome, I'll put it together and run it by you I suppose.
18:34:15evanif you'd like, sure.
18:36:05boyscoutFixed sexp for class/module. - 80a6276 - Brian Ford
18:42:45boyscoutCI: rubinius: 80a6276 successful: 3037 files, 12072 examples, 36356 expectations, 0 failures, 0 errors
19:19:37dbussinkevan: hmm, also just had a crash during gc when ri was generated, might be related to #204 i guess
19:19:43evanyeah
19:19:45evani've seen that
19:19:48evanhaven't tracked it yet.
19:20:11dbussinki have a apple crash report log, can try to reproduce it in gdb
19:21:10evansure
19:22:39dbussinkevan: #204 has been crashing reliably for me though
19:22:45evanok
19:22:47dbussinkok, i've caught it during ri in gdb
19:22:51evani'll look at that today
19:22:57dbussinkif you want to get anything out of it, let me know
19:52:39sbryantxb
19:52:41sbryanterr
22:21:55justin-georgehpricot 0.8.2 broken on master?
22:23:30rueBroken how?
22:23:44Defilerhpricot :(
22:24:19justin-georgewon't compile
22:25:05justin-georgehttps://gist.github.com/59860743dfeb6feb485e
22:25:27justin-georgethere's about 400 lines of more 'subscripted value' problems but whatev.
22:32:38rueDefiler: Did not make it to MWRC?
22:33:07rueThat looks like what you get with missing include files
22:33:50rueOhm, .rl file. Dunno, should probably try
22:59:20justin-georgerue: should probably try... ?
23:35:16ruejustin-george: Should probably try running myself, I meant
23:35:31justin-georgeoh, cool
23:36:04justin-georgeyeah it's not a big deal, just got my app running under rbx for the first time <3 (hpricot isn't vital)