00:00:00 --- log: started retro/10.05.02 00:29:40 --- quit: docl (Ping timeout: 245 seconds) 05:43:49 --- quit: virl (Remote host closed the connection) 06:45:35 --- join: erider (~chatzilla@pool-173-69-156-197.bltmmd.fios.verizon.net) joined #retro 06:45:47 --- quit: erider (Changing host) 06:45:47 --- join: erider (~chatzilla@unaffiliated/erider) joined #retro 06:55:20 hello all 08:21:58 --- join: docl (~luke@216-161-87-126.ptld.qwest.net) joined #retro 08:25:36 http://retroforth.com/paste/?id=3366 08:26:59 --- quit: docl (Read error: Connection reset by peer) 08:32:12 --- join: docl (~luke@216-161-87-126.ptld.qwest.net) joined #retro 08:57:57 --- quit: crc (Ping timeout: 276 seconds) 08:59:33 --- join: crc (~charlesch@184.77.185.20) joined #retro 09:43:30 good afternoon 09:45:46 whats up crc 09:45:51 how are you doing 09:45:55 doing well 09:46:05 how is the baby 09:46:06 working on retroforth 10.6 (now in a fossil repo) 09:46:09 baby is good 09:46:14 cool 09:48:30 I am implementing in wsh a program to to check files for signature malware 09:53:18 wsh? 10:30:36 windows script host = wsh 10:30:49 wscript/cscript 10:38:37 I am using the regex object that is has built-in 10:56:39 something simple I was working on for a friend 11:06:41 crc http://paste.pocoo.org/show/208934/ 11:07:35 reminds me of visual basic 11:13:03 yeah its a hybrid of javascript/VBscript I don't like it but my buddy cannot use anything else on his computer 11:13:35 why not? 11:14:11 he can not install new software 11:14:22 its a work computer 11:14:46 wscript/cscript come preinstalled 11:14:54 comes* 11:14:56 ok 11:20:37 crc plus I am learning some new kind of because I am porting python code to wscript 11:21:09 something new * 11:47:40 new things are good 12:24:31 --- quit: erider (Ping timeout: 265 seconds) 13:10:16 --- join: virl (~virl__@chello062178085149.1.12.vie.surfer.at) joined #retro 13:51:21 is the retroforth fossil repo supposed to be possible to clone? it is currently denying me access. 13:56:37 it should be clonable 13:57:49 --- join: belus (~belus@188.24.213.241) joined #retro 13:57:57 --- part: belus left #retro 13:58:21 Error: not authorized to clone 13:58:54 try again 13:59:43 --- join: erider (~chatzilla@pool-173-69-156-197.bltmmd.fios.verizon.net) joined #retro 13:59:51 wb erider 13:59:54 works :) 13:59:57 hi erider 14:00:12 ok, I had some of the permissions set wrong for anonymous users 14:06:57 As I was saying last night, I like prefixes. They make variables seem more like constants or numbers. 14:07:56 they were the last thing from 9.x that I missed 14:10:12 there are a couple shortcomings with the current implementation though. 14:10:29 what? 14:10:38 you can't do @@var to follow a pointer multiple steps 14:11:00 that wasn't possible originally 14:11:13 also, if you have a word that returns an address you can't do !foo to store to that address 14:13:37 the second point may be addressable soon 14:13:45 :) 14:17:03 ok, I think I get what it is doing. prefix is class blind so it is handling the word as if it were .data class. it just looks up the xt and stores it, rather than running .word on it first. 14:17:37 right 14:26:59 --- join: erider_ (~chatzilla@pool-173-69-156-197.bltmmd.fios.verizon.net) joined #retro 14:28:25 --- quit: erider (Ping timeout: 245 seconds) 14:28:26 --- nick: erider_ -> erider 14:28:48 --- quit: erider (Changing host) 14:28:48 --- join: erider (~chatzilla@unaffiliated/erider) joined #retro 14:29:37 docl: I can extend them after the rest of the core is loaded 14:31:39 : xt:class ( a-aa ) dup xt->d d->class @ with-class ; 14:31:40 here is __@ ] ( a-n ) xt:class &@ .word ; 14:31:40 here is __! ] ( na- ) xt:class &! .word ; 14:35:23 that works. 14:43:08 I'll push a commit shortly 14:45:22 brb 14:45:25 --- quit: crc (Quit: http://retroforth.org) 14:56:35 --- join: crc (~charlesch@184.77.185.20) joined #retro 14:56:49 ok, I've got this working; now leveraging it in the core 14:58:56 cool 15:00:53 done. both repos should have the patch now 15:03:38 nice! 15:04:09 your first case currently can't be done cleanly, sorry 15:08:22 darn 15:08:33 oh well 15:12:36 * crc wonders what the reaction of the ans people would be towards prefixes 15:18:51 good question 15:32:21 http://retroforth.blogspot.com/2010/05/prefixes.html 15:52:10 --- quit: erider (Ping timeout: 245 seconds) 17:28:03 --- join: erider (~chatzilla@pool-173-69-156-197.bltmmd.fios.verizon.net) joined #retro 17:29:40 test 17:30:14 you are here 17:30:26 cool 17:30:36 --- quit: erider (Changing host) 17:30:37 --- join: erider (~chatzilla@unaffiliated/erider) joined #retro 17:34:51 crc so how are you changing the core of retro 17:40:44 I merged the various files that built the core into one, then did a refactoring 17:41:33 treating it as a single piece allows me to do more interesting cleanups and organize things better 17:45:22 so it is smaller 17:54:27 yes 18:08:30 the reference implementation of the vm is now 610 lines 18:08:39 the rx-core source is 843 lines 18:13:39 so it is smaller with better performance? 18:14:08 I've not measured performance 18:14:36 code size is smaller, so with a decent vm implementation it should run faster 18:15:34 so you stripped out stuff not needed 18:16:02 yes 18:16:13 that is cool 18:40:57 * docl has a nestable version of prefixes almost working. 18:41:13 docl: cool 18:54:41 docl: are you still reading about dbase 19:16:51 http://retroforth.com/paste/?id=3367 19:30:37 a bug that I was tripping over was caused by the recent patch... when __@ tries to look at an anonymous xt (without an entry in the dictionary) xt->d returns 0 and d->class returns 1. adding a 0; makes it simply return the xt. 19:42:14 is this ready to commit? 19:42:49 I have a patch for it, seems to work ok 19:43:16 http://retroforth.com/paste/?id=3368 19:47:17 problem there: ;then, fori, and nexti aren't defined until later in core.retro 19:49:02 and doing: @@@nonexistantword leaves a value on the stack 20:20:30 --- quit: erider (Ping timeout: 245 seconds) 20:24:02 http://retroforth.com/paste/?id=3369 20:24:57 should work better 20:25:58 * docl actually compiled from the source file this time 20:25:58 I'm trying to adapt it to the metacompiled part 20:26:11 cool 20:39:21 * docl notices that queue> isn't actually used 20:53:50 bug: 20:53:54 ok .s <5> 1 2 3 4 5 20:53:54 ok @@@@@@@cr 20:53:54 ok .s <5> 1 2 3 4 0 21:02:29 hmm. in that case it is assuming that cr is a word that gives an address as a result which it doesn't. 21:02:44 ahh, that would be it 21:02:49 * crc is getting tired :) 21:02:53 :) 21:03:12 I think I found a bug with compile-time 21:04:43 variable bar 21:04:43 10 !bar 21:04:43 : foo @bar ; 21:04:46 see foo 21:13:08 ahh, it was the use of .word in handle 21:13:18 should have been with-class 21:24:01 --- quit: virl (Remote host closed the connection) 21:26:58 docl: try http://retroforth.com/paste/?id=3371 21:31:30 nice! 21:31:58 does it work ok? 21:36:09 multiple @ doesn't seem to work. :( 21:37:25 @@foo is not found 21:37:48 it's found on my setup 21:38:14 do you get any errors building? 21:39:07 (e.g., in the core; anything other than "RETRO", a blank line, "Last header at ....", "MAIN @ ....", "RETRO", and another blank line? 21:40:41 nope 21:41:47 it seems to work for me: http://retroforth.com/paste/?id=3372 21:48:15 I used fossil revert, then applied the patch using patch -i 21:49:00 the @@@foo command just gives me @@@foo ? for some reason. @foo works fine. 21:49:33 it's working for me 21:57:29 ahh, I was trying it in the repo for retroforth rather than rx-core 21:57:40 that would do it 21:59:26 I pushed the updated source+images to both repos 22:01:04 excellent 22:07:41 ok, with that I'm going to go to bed. have a good night 22:09:44 you too 23:59:59 --- log: ended retro/10.05.02