00:00:00 --- log: started retro/12.04.10
03:58:17 --- quit: crc (Remote host closed the connection)
07:28:25 --- join: crc (~crc@li125-93.members.linode.com) joined #retro
07:31:14 <CIA-98> retro: 03crc <crc@questor> * rcrc@questor-20120410143050-rsmu6adgr59qb4ny 10rx/examples/misc/easter.rx: example of calculating date of easter
08:03:10 --- join: __tthomas__ (~Owner@24.130.7.34) joined #retro
08:36:31 --- quit: jyfl987 (Quit: Lost terminal)
09:02:51 --- join: Kumul (~Kumul@adsl-72-50-66-190.prtc.net) joined #retro
12:34:16 --- join: Mat2 (5b4085c5@gateway/web/freenode/ip.91.64.133.197) joined #retro
12:34:24 <Mat2> hello
13:18:13 <crc> hello Mat2 
13:18:26 <Mat2> hi crc
13:19:18 <Mat2> what's new ?
13:24:57 <crc> not much
13:26:08 <crc> I've been working on a better testing framework for retro-11.4
13:30:43 <Mat2> i'm experimenting with instruction streaming
13:36:52 <__tthomas__> hey Mat2
13:37:05 <CIA-98> retro: 03crc <crc@questor> * rcrc@questor-20120410203641-iym99xifcsovejkj 10rx/test/base.rx: partial coverage of core language with new test framework
13:37:05 <Mat2> hi !
13:46:27 <__tthomas__> I have been playing with sharpdevelops ability to automatically convert C# to boo, python, ruby and visual basic, it works way better than I expected..
13:47:39 <__tthomas__> and installing monodevelop to test C# version with mono
13:49:01 <__tthomas__> I think I have sockets pretty well figured out, can use select to see if will block with c and python, so should be able to make them behave identically on at least 3 platforms..
13:50:14 <CIA-98> retro: 03crc <crc@questor> * rcrc@questor-20120410204951-wj01r506y425yo92 10rx/ (library/test.rx test/base.rx): improve output formatting of tests
13:50:39 <Mat2> great
13:52:49 <Mat2> i've found a nice waY for far better vm performance without native-code compilation just though instruction streaming of prior interpretated code-sequences which eleminating BTB stalls
13:52:59 <Mat2> completely
13:54:16 <Mat2> and speed up execution by 70% depending on the cpu architecture
13:54:26 <Mat2> up to 70%
14:03:27 <__tthomas__> very cool.. 
14:04:40 <__tthomas__> seems like dups should be very optimizable..
14:15:11 <Mat2> there exist an elegant method for resolving all stack operations by bringing stack based code into a sequence of monadic and dyadic functions
14:16:33 <Mat2> as there consume only one or two parameters all stack accesses can be mapped into accessing two registers, one accumulator and one operand register
14:16:54 <__tthomas__> I figured there had to be a way.. would be faster to just keep value on stack then duplicating it..
14:19:22 <Mat2> cache tos in a register and dup get identical to store tos onto stack
14:27:31 <Mat2> need some sleep, ciao
14:27:35 --- quit: Mat2 (Quit: Page closed)
17:14:54 --- quit: __tthomas__ (Quit: Leaving.)
20:49:20 --- quit: Kumul (Quit: gone)
23:05:28 --- join: jyfl987 (~jyf@unaffiliated/yunfan) joined #retro
23:06:35 <jyfl987> crc: what tools does the retroImage made by? i just found that in retro.py it dont have any code to made that
23:46:51 --- quit: ivan`` (Quit: ERC Version 5.3 (IRC client for Emacs))
23:49:12 --- join: ivan`` (~ivan@unaffiliated/ivan/x-000001) joined #retro
23:59:59 --- log: ended retro/12.04.10