00:00:00 --- log: started retro/11.02.07 05:45:27 --- join: aisa (~aisa@86.sub-69-96-208.myvzw.com) joined #retro 05:46:50 --- join: aisa_ (~aisa@150.sub-75-208-11.myvzw.com) joined #retro 05:49:45 --- quit: aisa (Ping timeout: 240 seconds) 05:49:50 --- nick: aisa_ -> aisa 06:08:04 --- join: aisa_ (~aisa@86.sub-69-96-208.myvzw.com) joined #retro 06:09:28 --- quit: aisa (Ping timeout: 276 seconds) 06:09:30 --- nick: aisa_ -> aisa 06:19:01 --- quit: aisa (Quit: aisa) 06:38:08 --- join: aisa (~aisa@173-10-243-253-Albuquerque.hfc.comcastbusiness.net) joined #retro 07:56:47 --- join: qbject (417dcfe2@gateway/web/freenode/ip.65.125.207.226) joined #retro 08:10:57 Morning, Folks. 08:15:34 hi qbject 08:17:50 Hey, cfa. 08:30:34 I've found some rather old posts when searching Retro Forth from a guy named Tom Novelli. Was his project related to this effort? 09:02:00 --- quit: aisa (Read error: Connection reset by peer) 09:02:03 --- join: aisa_ (~aisa@173-10-243-253-Albuquerque.hfc.comcastbusiness.net) joined #retro 09:08:56 I think it was, originally. 09:15:45 tom did the first four versions of retro, and helped with the sixth after I took over 09:17:52 Ah, neat. 09:19:26 crc: some good responses on c.l.f 09:20:20 yup 09:20:35 I've modified the vms to use the C style behavior 09:22:55 great 09:24:08 did you check out the 'division and modulus for computer scientists' paper? 09:24:26 not yet 09:25:50 ron pretty much sums up my position 09:26:13 (albeit a position that was founded in ignorance of the problem) 09:33:26 cfa: is this paper available online? 09:34:07 sure, we're talking about this thread --http://groups.google.com/group/comp.lang.forth/browse_thread/thread/efb075747bf91e34# 09:34:24 the paper is http://www.google.com/url?sa=D&q=http://legacy.cs.uu.nl/daan/download/papers/divmodnote-letter.pdf 09:34:41 hopefully pasting didn't mangle that too much 09:43:37 Those worked for me. Thanks. 09:45:21 np 09:45:33 heading home; 09:45:37 * cfa idles 09:56:09 --- join: Mat2 (5b4085ac@gateway/web/freenode/ip.91.64.133.172) joined #retro 09:56:24 hi folks 10:03:41 Hey Mat2 10:03:51 hi 10:06:18 hi Mat2 10:07:39 --- join: Mat2_ (5b4085ac@gateway/web/freenode/ip.91.64.133.172) joined #retro 10:07:52 hmm, i'm zombied 10:08:16 --- quit: Mat2 (Ping timeout: 245 seconds) 10:08:42 --- nick: Mat2_ -> Mat2 10:08:46 hi crc 10:09:20 i'm just reading about the /mod problem in comp.lang.forth 10:10:04 * cfa returns 10:10:05 hey all 10:10:20 hi cfa 10:11:41 hey Mat2 10:13:31 --- quit: aisa_ (Read error: Connection reset by peer) 10:13:43 --- join: aisa (~aisa@173-10-243-253-Albuquerque.hfc.comcastbusiness.net) joined #retro 10:16:15 Mat2: any thoughts re: /mod? 10:16:49 for me the euclidean method is more natural 10:18:25 interesting 10:19:09 is this due to prior experience, or just a general feeling? 10:19:39 because I've learned the modulo operation this way in school 10:22:14 it is also superiour to the floored and truncated method because of its regularity 10:25:02 maybe so 10:25:13 i came to the problem rather informally, i have to say (just check the scrollback) 10:25:27 i intuitively expected -5 2 / to yield -2, hence i had certain expectations of /mod 10:25:47 though i originally didn't twig that the "unexpected" output of /mod resulted from the flooring method for division 10:27:31 (i actually encountered the discrepancy between /MODs in my own forth [i compared the stack with gforth] and thought it a bug; i guess my nose was too close to the source) 10:28:22 ((sorry, that's a lot of "i"s) 10:29:12 hmm, there exist some traditions for programming languages. Pascal oriented tend to define a mod operator in the euclidean way for example 10:30:03 gforth is C based so thes may be the reason for the different handling 10:30:24 hese 10:30:29 ^these 10:30:38 (these damn keyboard) 10:32:22 or better depends on the mod implementation of gcc because C doesn't define a standard handling 10:34:22 perhaps for retro, two words are in order 10:34:53 s/mod and f/mod 10:35:04 yesterday I compiled my current vm sources with clang (llvm) and this compiler handle the % operator differently to gcc. That's a great pitfall which cost me some time of debugging 10:35:18 no kidding 10:35:24 no 10:35:28 surely that's standardised? 10:36:04 I think, no 10:36:44 mod can certainly be a headache eh? 10:37:56 to be save, it's better to use a self made function or macro for this operation (if you implement a forth vm in C) 10:39:52 by the way, the new clang version and llvm generates impressive good code for threaded interpreters 10:49:24 neat 11:17:05 ciao 11:17:09 --- quit: Mat2 (Quit: Page closed) 11:33:41 fwiw, I now have an android device (nookcolor), so a retro for android will be coming later this year 11:43:15 i can help out if you like -- i have an htc wildfire here 11:43:24 plus i'm quite familiar with the android sdk 11:46:18 sounds good. I've done a bit of java coding (years ago, before "Java 2" and such), but nothing for android 11:46:31 it's a nice platform 11:46:39 do you have eclipse installed? 11:46:44 not yet 11:47:38 adt is really good 11:48:13 --- quit: qbject (Ping timeout: 245 seconds) 11:48:15 plus the sdk comes bundled with a simulator which makes development very pleasant 12:06:19 interesting reply from andrew haley on the /mod thread 12:16:48 I've dabbled with the simulator (2.1, 3.0-preview1) 12:22:35 i had no surprises when i installed an apk tested in the simulator onto the wildfire 12:22:51 the only thing is we need to test under a variety of OS versions and screen sizes 12:46:06 it can't be much worse for testing than the javascript implementation was 14:23:23 --- quit: aisa (Quit: aisa) 23:59:59 --- log: ended retro/11.02.07