00:00:00 --- log: started retro/06.07.08 02:25:25 --- join: neceve (n=claudiu@unaffiliated/neceve) joined #retro 03:07:02 --- join: Cheery (n=Henri@a81-197-20-242.elisa-laajakaista.fi) joined #retro 03:34:38 --- quit: Cheery (Read error: 104 (Connection reset by peer)) 03:34:38 --- join: arke_ (n=f2@bespin.org) joined #retro 03:57:33 --- quit: arke (Read error: 110 (Connection timed out)) 04:08:44 --- nick: arke_ -> arke 05:19:31 --- join: crc_work (n=cchilder@h-64-105-207-163.phlapafg.covad.net) joined #retro 05:21:55 I will be coming and going until around 1.30 or 2pm, when I get home 05:29:54 what time is on your side crc? 05:30:58 8.30 am 05:31:28 ah, ok. 05:33:30 hmm, a byte is even on x86 big endian, could that be right? 05:34:26 no, the bits are still stored in little endian order IIRC 05:34:52 hmm, ok. 05:40:35 hmm... 05:41:20 well I wrote a little program in C which puts 'fa' to stdout, I piped the output into a file and then I hexdumped it 05:41:50 result: 00fa, so I would say that it prints it in big endian 05:44:30 00FA is a word size, not a byte sized value 05:44:37 so that is little endian 05:44:43 when outputting another byte it gets filled into the high order byte so when this byte is 'cb' it's: cbfa 05:44:55 but the byte itself is bigendian 05:45:09 so not this one: af 05:46:36 well, I said I have problems in thinking in little endian 05:48:22 it's still stored as little endian in memory 05:49:17 that wasn't really the question 05:49:35 x86 is purely little endian 05:49:45 I meant is a byte on x86 fa or af 05:49:52 fa 05:50:06 but stored with the bits in the reverse order of what you would expect 05:50:45 yeah, the next byte is stored before the last byte. 05:51:09 so when you have: fa cb then it's stored like this: cb fa in memory 05:51:15 yes 05:51:36 at a byte level, the endian issues are irrelevant 05:53:06 that's why I love bytes, they don't make so much headaches. 05:53:47 yeah 05:55:13 oh, I love the idea so much of bytecode streams. 06:41:31 how much systems use different data types? I know that it occures that a 8bit systems talks to it's memory with 16bit, but how often does that happens? 06:43:00 in most systems 06:43:21 x86 lets you ignore alignment issues, but most RISC systems do not 06:44:29 well, I never saw a 32bit CPU which doesn't talk to it's memory in 32bit adresses 06:44:52 x86 will let you access bytes, or words directly 06:45:22 even across 32-bit dwords if you want 06:49:01 aha.. 06:49:18 so what's the answer? 06:49:54 the answer depends on your field. desktop/laptop or embedded? 06:50:17 desktop/laptop is mostly x86, so you can have many addressing modes 06:50:26 on desktops/laptops I think that's normal 06:50:46 eh? I didn't ask for adressing modes. 06:52:35 ? 06:53:03 8 bit, 16 bit, 32 bit, are all data widths that can be used on x86 systems 07:17:02 http://retroforth.net/paste/?id=13 07:31:15 this adds support for three prefixes: @ ! and ~ 07:31:31 ~ gives the xt, @ fetches from a variable, ! stores to a variable 07:51:10 neat! 07:51:28 I will be putting this into the repo in a few minutes 07:56:20 done :) 07:58:42 when I get home, I'll start redoing some of the modules (anything loaded after 'retro') to use this 08:00:46 crc_work, what do you think about simplification? 08:02:47 it's a good idea 08:02:56 I tend to do a lot of it myself ;) 08:03:21 though my normal stategy is to quickly develop something that works, then simplify as I refine it 08:05:50 yeah, it's important to have understandable pieces of software. programs ok, they are written for a special purpose and so they can be very complex, but libs or platform code(interpreters, compilers or libs) should be understandable. 08:06:14 I agree 08:12:21 :-) 08:27:12 there's a process of developing bad stuff, throwing it away, and replacing it with good. trouble is, many modern apps are stuck in the first stage. 08:27:19 e.g. X11 08:27:56 bloated=bad, of course ;) 08:29:59 especially for x11 a bad protocol, or did you look at it? 08:30:17 just the doc crc sent me to last night 08:30:57 also the code you showed me 08:31:31 was that my x11 features app? 08:31:39 I think so 08:31:47 probably not your latest 08:53:17 1 more hour :) 09:18:57 thinking on classes, I *could* write a class that responds to "messages" (think something vaguely like smalltalk) as well as the normal state-based model 09:19:11 I may do this when I redo the gtk bindings 09:22:30 that sounds like it might be a good idea, it might keep the complexity of the message passing to an understandable level 09:23:28 there's some stuff that just doesn't seem to belong on a stack. 09:23:53 it would also keep the word count down to an approachable level 09:26:36 right 09:32:16 --- part: crc_work left #retro 10:14:54 home at last :) 10:20:28 hiho... 10:33:38 two more patches in the repo, one fixes a bug in the prefixing (now prefixes can be used in definitions) 10:59:35 I have been working on the colorscheme for http://retroforth.net/discuss 10:59:48 any comments or suggestions before I proceed further? 11:02:38 looks fine to me 11:04:54 cool :) 11:05:26 can be used in definitions? 11:05:31 * crc likes this better than smf, it's smaller, and less prone to problems than the current smf 11:05:39 virl: the prefixes 11:05:44 e.g, ~words 11:05:48 rather than ['] words 11:08:00 smf? 11:08:06 simple machines forum 11:08:18 and that's what? 11:08:20 the second generation of yabbse forum 11:08:44 which was a php version of the original YaBB forum (the original was written in perl) 11:10:56 I really like the new forum as well 11:22:41 --- join: Cheery (n=Henri@a81-197-20-242.elisa-laajakaista.fi) joined #retro 11:24:07 I just updated to the 1.0 release, it may fix some of the problems you encountered when we were doing the initial tests 13:12:28 --- join: Icefall (n=Miranda@194.126.180.10) joined #retro 13:34:54 * docl has been playing with putting strings in cons cells 13:35:01 http://retroforth.net/paste/?id=15 14:27:13 I am about to install a new hard drive (160gb) and SUSE 10.1; will be back tonight if possible 14:27:54 --- quit: Cheery ("Leaving") 14:29:43 ok, good luck 14:29:50 thanks 14:29:51 ttyl 14:30:00 --- quit: crc ("hardware work") 14:36:45 --- quit: neceve (Remote closed the connection) 14:53:03 --- part: arke left #retro 15:05:27 --- quit: Icefall (";-)") 15:18:10 --- join: snoopy_1711 (i=snoopy_1@dslb-084-058-161-076.pools.arcor-ip.net) joined #retro 15:18:10 --- quit: Snoopy42 (Read error: 131 (Connection reset by peer)) 15:18:34 --- nick: snoopy_1711 -> Snoopy42 16:20:05 --- join: jas2o (n=jas2o@WNPP-p-203-54-32-69.prem.tmns.net.au) joined #retro 16:34:26 --- join: crc (n=crc@pool-70-110-215-240.phil.east.verizon.net) joined #retro 16:34:44 --- mode: ChanServ set +o crc 16:40:18 hello crc 16:41:38 hi 17:08:23 * crc is now running SUSE 10.1 17:08:35 it's not too bad so far 17:08:55 I always wanted to try SUSE 17:09:06 I have an install dvd, but that is no use to me 17:09:25 I also have a livecd (gnome), but my laptop doesn't have enough ram :( 18:51:49 --- part: jas2o left #retro 19:24:32 --- join: snoopy_1711 (i=snoopy_1@dslb-084-058-136-004.pools.arcor-ip.net) joined #retro 19:25:58 --- join: nighty (n=nighty@CPE00508be190fc-CM0012253ec1bc.cpe.net.cable.rogers.com) joined #retro 19:32:49 --- quit: Snoopy42 (Read error: 145 (Connection timed out)) 19:32:53 --- nick: snoopy_1711 -> Snoopy42 23:59:59 --- log: ended retro/06.07.08