00:00:00 --- log: started forth/03.04.17 00:01:44 --- join: karingo (karingo@158.portland-01-02rs.or.dial-access.att.net) joined #forth 00:02:57 Heh, now this is an interesting spam management technique. 00:03:06 Suppose we have two computer users on a network, Alice and Bob. 00:03:08 (of course) 00:03:50 Alice sends a message to Bob. In the process, Alice's e-mail client does two things: first, it generates a random number, and appends it to the subject heading. Next, it stores that random number for later use. 00:04:45 When Bob wants to reply to Alice, that random number is preserved in the subject heading. When receiving mail, if Alice's client sees that the subject heading's number matches one of the random numbers it generated in the past, it lets the message through without additional red tape. 00:04:57 Otherwise, the mail is considered suspect, and more traditional white/black-list techniques are applied. 00:05:17 that's not the problem though 00:05:40 i mean, it's a technique.. 00:06:06 mixed in though with white/black-list it does something.. 00:06:18 but you still have to look at all your pending email to make sure something new isn't stuck in the queue. 00:06:30 which means you still sort through spam. 00:06:33 That's the point -- this is intended to work with white/black lists. 00:07:03 Only with this technique, messages that are known replies to your outgoing messages appear instantly. 00:07:10 Those should always have priority over new stuff anyway. 00:11:15 back 00:12:18 kc5tja: sounds close to.. well i forgot what the program is called.. 00:12:49 but basically when you send outgoing mail it appends some numbers to the email addr (assuming you have one that allows for extensions) 00:13:28 and that can be used for some period of time, or a number of messages, whichever... and if they don't use that then they have to go through a authorization thing 00:13:49 sorta like when you sign up to a mailing list you have to reply to the email making sure you is you 00:14:15 bayesian spam filter works good enough for me tho (actually really well) 00:14:37 * kc5tja nods 00:14:48 I haven't used a Bayesian filter yet, but I've heard some really, really good things about them. 00:14:58 Paul Graham has an article or two on them, IIRC. 00:15:01 yep 00:15:24 I like spamprobe.. there are some other better known ones but I found spamprobe to be best 00:17:01 I used to get about 15 spam per day.. now I get one per month in my inbox.. no false postives 00:18:13 Is this for Windows? 00:19:07 linux, I think it might run windows (with cygwin tho) 00:19:23 I know there was some discussion about it recently on the mailing list 00:19:55 I use procmail, which finds all the mailing lists and filters them, then I check it with spamprobe and it goes to Inbox or junk depending 00:21:36 Hmm...I'll have to research it when I get my own mail account running on a local box someday. 00:22:24 Yeah.. it isn't really applicable to hosted email.. compared to something like spamassassin, since you have to train it 00:32:27 hmmm ... I wonder how easy it would be to build a primitive convection-powered solar wind turbine. 00:32:35 Something like a scale model of a solar chimney. 00:36:00 howso convection-powered? 00:40:13 Sunlight (or some other form of external combustion) is used to heat air at the bottom of a loop of piping of some kind. 00:40:43 Hot air rises, being less dense and hotter because of its larger volume. 00:41:07 right 00:41:16 hmm 00:41:28 Because there is a positive mass flow rate, the hot air passing by the turbine will cause it to spin, since it becomes pressurized otherwise. Thus, volume is, at least temporarily, turned to a pressure ratio. 00:41:39 It's literally equivalent to a Brayton cycle engine, but without a compressor stage. 00:41:53 would need to capture a lot of sun 00:41:54 hmm 00:42:27 Not necessarily. 00:42:43 THe sun delivers 1kW of power per square meter. 00:42:56 Or, in another way, 37 watts per square foot. 00:43:11 Of which, 5% can be reasonably expected to be used for power. 00:44:04 So if I get 1.5W of shaft power out such a contraption with only a single square foot, flat-plate collector, I'm doing good. :) 00:44:20 (Heck, with only two such devices, I could power my 40m QRP rig!) 00:45:31 wow 00:47:34 Should be a lot easier to build than a comparably sized Stirling cycle engine. :D 00:47:40 At least with my resources. 00:48:02 --- quit: karingo () 00:48:27 hm I must sleep now... 00:48:47 Ditto; I have school... :/ 00:48:49 73 00:48:55 --- quit: kc5tja ("THX QSO ES 73 DE KC5TJA/6 CL ES QRT AR SK") 00:49:01 night 00:49:03 --- quit: bwb ("Later") 01:49:57 --- quit: natty (Read error: 60 (Operation timed out)) 02:12:41 --- join: Fractal (kdihgrkv@dont.try.configuring.openbsd.on.stronglsd.com) joined #forth 03:41:02 --- join: natty (~n1ywb@155.42.84.139) joined #forth 04:16:01 --- quit: fridge ("http://lice.codehack.com") 04:27:20 --- join: gilbertdeb (~gilbert@fl-nked-ubr2-c3a-29.dad.adelphia.net) joined #forth 04:49:24 --- quit: Fractal (Read error: 110 (Connection timed out)) 04:58:14 --- join: crc (12345678@AC88CB27.ipt.aol.com) joined #forth 05:12:55 --- quit: crc ("Leaving...") 05:30:06 --- join: mur (jukka@baana-62-165-189-22.phnet.fi) joined #forth 06:50:22 --- join: Herkamire (~jason@h0030657bb518.ne.client2.attbi.com) joined #forth 07:36:25 --- quit: natty (Read error: 60 (Operation timed out)) 07:36:56 --- join: natty_ (~n1ywb@155.42.84.139) joined #forth 07:38:37 wb natty_ 08:28:03 --- join: flyfly (~marekb@ip164.ktvprerov.cz) joined #forth 09:10:47 --- join: remote (~nicolas@Toronto-HSE-ppp3715612.sympatico.ca) joined #forth 09:12:56 --- part: remote left #forth 09:29:19 sweet, got the forth programmers handbook pdf 09:30:13 got starting forth? 09:30:30 zorking sweet 09:30:41 no but I checked out a copy of thinking forth a few months ago. good read. 09:30:54 I should buy starting forth actually 09:30:58 i will soon 09:32:54 --- quit: gilbertdeb (Remote closed the connection) 09:39:55 --- join: gilbertdeb (~gilbert@fl-nked-ubr2-c3a-29.dad.adelphia.net) joined #forth 09:40:05 Confucius say: don't kill xdm 09:40:35 * Herkamire says don't use xdm 09:40:44 what to use then? 09:40:54 startx 09:41:09 I run it to manage xdmcp 09:41:21 the indigo2 has no monitor as yet. 09:44:35 so what does it do then? 09:44:56 do you have a keyboard hooked to the indigo2? 09:45:54 nope. 09:46:09 I run everything from here... 09:46:24 i make it export the Xsession to this machine. 09:55:14 wierd 09:55:22 is that faster than just letting ssh do it? 09:55:55 I didn't bother with ssh since its on my lan. 09:56:24 ok, well I don't know much about X authentication 09:56:50 it isn't about the authentication so much as it is about the managing of the sessions. 09:56:55 I thought xdm was just for giving you a graphical login prompt 09:57:10 as it is now, I run X, and get a chooser menu from which I select either the solaris or the irix box. 09:57:40 the chooser then calls the graphical login prompt from whichever machine I chose. 10:01:48 ok 10:01:54 I just use ssh which handles everything 10:02:17 and I hardly ever display windows on a remote machine. 10:02:40 you just use the console stuff right? 10:02:54 I use xterm for almost everything 10:03:20 xterm and mozilla 10:08:16 I use screen as my UI :) 10:08:21 for irc and such 10:13:45 how do you scroll in screen? 10:45:28 Herkamire: god I wish i knew 10:45:35 I dont think it works that way.. 10:46:09 if you figure out how, let me know... but what I do is just show only one screen, and i scroll up like it was only one terminal and it usually will have the previous contents of whatever was written to whatever screen was active in the scrollback 10:46:26 there's another thing or two I can't figure out how to do in screen 10:53:25 ianni: are you opposed to xterm? 10:57:15 --- join: krish (KRISHNAKUM@61.1.220.221) joined #forth 11:19:06 I'm not familiar with it 11:19:18 er,, am i ? 11:19:21 I dont use xwindows 11:19:27 it's too slow .. I use Mac OS 11:19:34 my OpenGL terminal program is quicker 11:19:36 though 11:19:42 actually, terminal would be one justification for using X11 11:19:51 * ianni downloads the newest X11 for osx 11:20:18 ianni, do you use mac? 11:20:36 yes 11:20:46 hence the "I use Mac OS 11:20:46 " 11:20:48 :) 11:21:40 * mur whois 11:23:30 why, do you? 11:23:54 i thought that was in whois at real name place 11:23:57 but it was not 11:24:09 * mur woudl like to use mac 11:24:28 why? 11:24:33 they only have one mouse button! 11:24:47 gilbertdeb: not my mac at home! 11:24:53 I use a logitech - mouse wheel works too 11:24:59 it's a damn USB mouse slot.. :) 11:25:22 shhhh I'm trying to discourage him from using a mac. 11:25:24 ianni, what kind of mac? 11:25:31 --- quit: tgunr (Read error: 110 (Connection timed out)) 11:25:34 powermac g4 400mhz I got used 11:25:40 nice 11:25:44 yup 11:25:46 powerbook woudl be kewl 11:25:49 got a sweet deal i couldnt pass up 11:25:52 I am saving up for one, mur 11:25:55 --- join: tgunr (~davec@vsat-148-63-143-139.c189.t7.mrt.starband.net) joined #forth 11:26:00 whats the deal you got? 11:26:07 $500 11:26:08 which have you thought to get? 11:26:13 mur 15" or 17" 11:26:16 hmmm. 11:26:17 800mhz+ 11:26:30 15# has weird design :P 11:26:36 looks like 60 11:26:40 60? 11:26:41 's scifi computer :P 11:26:46 on top 11:26:47 well the 12 and 17 are new 11:26:48 but else it's nice 11:27:00 the 15 has not been updated yet 11:27:15 i want th e17' soooo bad 11:27:17 perhaps one coudl get powerbook used cheap 11:27:29 It has the fiberoptic backlit keyboard... droool 11:27:31 ianni, can you get one for me too :) 11:27:37 hehehe 11:27:39 fiberoptic? 11:27:42 ianni: have you gotten to the forth prompt on your mac yet? 11:27:44 if i win the lottery (even though I don't play) I will buy you one 11:27:56 gilbertdeb: sure gforth will run on OSX, I havent played with the firmware no 11:28:00 ive done it befoer ;) 11:28:20 how do you get to the prompt? 11:28:50 in console window? 11:29:16 gilbertdeb ... some keystrokes when you boot brings up a window 11:29:25 ianni, is it possible to tweak the gui colors anyhow? 11:29:30 er no maybe its whenever you want 11:29:32 * ianni looks it up 11:29:45 mur, yes... but the factory GUI looks nice 11:30:05 OSX is not very themeable... but it looks very niec without any special themes 11:30:17 you could run fullscreen X11 though, if you wanted I guess 11:30:31 ianni, i dnd't meant the gray and colourful but to change colours more than between the 2 choises 11:30:50 can you select colours like in windows for instance? 11:31:20 sure 11:31:58 for what? 11:32:14 mur: to spoil the desktop ? 11:32:30 noo.. to something else 11:32:43 like a bit different hue 11:32:47 for instance 11:32:56 it's cmd-opt-O-F at boot for the forth propmpt 11:33:29 Openboot Firmware 11:33:58 mur, you can change the selection colors and stuff 11:34:10 as far as changing window colors it's not as tweakable as windows, I don't think 11:34:13 but it's prettier than windows 11:34:25 and changing color from anyhting but light grey would make OSX look ugly, maybe.. 11:34:36 not sure extly what you're wondering 11:34:46 but it's definitely not very themeable. it's been done but it's very hack-ish 11:35:20 ianni: I think it is just as tweakable. 11:35:45 gilbertdeb: what haveyou tweaked? 11:35:56 the look of it all. 11:36:02 screenshot 11:36:04 colors and noises it makes. 11:36:05 :P 11:36:12 I don't have a mac anymore. 11:36:16 ah 11:36:44 what can't be tweaked? 11:37:13 window style 11:37:22 ah that. 11:37:22 ? not sure 11:37:29 can that be changed in windows? 11:37:31 yah, but you wouldnt really want to 11:37:34 i guess not 11:37:40 its just as hacky there too huh 11:37:52 but windows has like titlebar colors and stuff, and you can customize the colors of objects 11:38:00 that just kinda is diff. in Aqua 11:38:13 sure you can do that for the mac as well. 11:38:19 but it is usually part of a theme. 11:39:29 http://www.inpuj.net/~ian/desktop2.tiff 11:41:18 did it have to be a .tiff ? 11:41:48 and 2 MB long :P 11:43:32 --- quit: tgunr (Read error: 104 (Connection reset by peer)) 11:44:29 --- join: tgunr (~davec@vsat-148-63-143-139.c189.t7.mrt.starband.net) joined #forth 11:45:42 haha 11:45:46 i didnt know tiff didnt compress 11:45:48 --- quit: tgunr (Read error: 104 (Connection reset by peer)) 11:45:50 didnt notice :) 11:46:10 heard of jpg? 11:46:21 i think you can do TIFF JPEG 11:46:25 then it compresses 11:46:44 --- join: tgunr (~davec@vsat-148-63-143-139.c189.t7.mrt.starband.net) joined #forth 11:47:03 can you change the apple logo on left corner? 11:49:02 mur I think its possible. 11:49:18 it might be similar to changing the start button on the pc. 11:49:36 * mur is using linux on the pc :P 11:49:48 i dont see start button (sarcastic) 11:50:01 dragging picutre? 11:50:04 no. 11:50:13 right click and then some dialog? 11:50:19 editing explorer.exe with an icon editor 11:50:50 I see nothing wrong with that apple 11:51:01 he probably wants a peach or mango. 11:51:02 i suppose it'd be nice to change but it's not practical 11:51:05 gilbertdeb :) 11:51:33 maybe green what do you call these <>= shaped fruits? 11:51:46 paron rhmm 11:51:48 <> diamon shaped? 11:51:54 almonds? 11:51:58 first comes like circle 11:52:00 --- join: serg_penguin (serg_pengu@ts25-a145.Moscow.dial.rol.ru) joined #forth 11:52:01 then it goes on 11:52:21 .-''-.__i'll make 2 line examples 11:52:34 hi folks 11:52:39 '-__-'''' 11:52:45 hey serg_penguin! 11:52:49 zdrastivanya 11:53:01 privet 11:53:39 -> gilbertdeb did your BSD crashed ?? :(( 11:53:58 no I couldn't install > 4.8 on this machine. 11:54:31 debian is tolerable. 11:54:48 debian roxx ! RH suxx 11:54:49 I couldn't stand redhat or suse though. I recommend it to all bsd users. 11:54:54 it had MEGA holes 11:55:18 gilbertdeb: debian guru ? 11:55:22 a question is it tolerable to have dark background on dialogs in osx? 11:55:32 one was told about in RU printed magazine 4 stupid scriptkiddies 11:55:38 krish: debian user. 11:55:40 or will it become horrible and unusable 11:55:48 gilbertdeb: gotta problem 11:55:59 lets hear it krish. 11:56:17 there are a bunch of other debian users quietly lurking here :D 11:56:41 * mur hides 11:57:13 * serg_penguin uses stupid RedHat 11:57:25 wanna wipe it and setup BSD 4.8 11:57:29 * krish was using Mandrake 8 until now ... 11:59:01 where do u reside ? 11:59:26 who? 11:59:33 me ? 11:59:34 i'm now showing inet and IRC to a friend of mine, totally new to comp and net 11:59:41 who ? all of you 11:59:45 miami 11:59:51 i live in Salem, India 12:00:29 i only know Robert in Sweden :))) 12:00:34 4nah you dont want dark backgrounds on dialogs 12:00:53 you can write your own cocoa application fairly easily that would do that, 12:01:08 it would look ugly with dark colors 12:01:22 is cocoa a scripting language? 12:01:37 * krish thinks mur has plans for uglification if MacOS X desktop ... 12:03:07 gilbertdeb: cocoa is a OO framework 12:03:36 00 ? 12:03:50 obfuscated originalis 12:04:27 gilbertdeb: a fine framework 12:05:16 object oriented 12:05:46 gilbertdeb: tried objective-c ? 12:05:48 * mur is at finland 12:06:20 currently not at helsinki, will move in half year 12:06:44 krish, i'm arts student , i think i can manage .. 12:06:55 ObjC is nice 12:07:04 mur: manage what ? 12:07:26 colours 12:07:28 and graphics 12:07:35 and overall 12:07:45 oh.. ok ok 12:07:53 ;) 12:07:57 mur: no offence meant. 12:08:07 no offence taken 12:08:13 great. 12:08:33 krish never. 12:08:33 * mur is finnish, finns talking style is straightforward 12:09:45 name a famous finnish painter mur 12:10:00 gilbertdeb: never what ? 12:10:03 --- quit: tgunr ("Leaving") 12:10:27 gilbertdeb, why? i dont like finnish painters really :P 12:10:38 krish o-C 12:10:56 gilbertdeb: try it. 12:10:59 mur never heard of one. 12:11:01 gilbertdeb: simple and effective. 12:11:16 krish it is still c. 12:11:37 I resist C because i'm afraid lock-in might occur. 12:12:10 overwhelmingly popular macroasm language and unix/linux/bsd extension language of choice. 12:12:11 gilbertdeb: yes. but with all the good stuff. 12:12:21 krish: they good stuff comes from smalltalk. 12:12:28 and that, was a pain. :( 12:12:43 gilbertdeb: hmmm 12:12:55 gilbertdeb: read ESR's "How to become a hacker ?" 12:13:02 sure. 12:13:09 I've also read how to become a programmer in 10 years. 12:13:42 gilbertdeb: norvig ? 12:13:51 yeah. 12:13:58 gilbertdeb: cool fella 12:14:37 why did you ask about esr? 12:15:02 nothing particular. 12:15:13 gilbertdeb: just wanted to know if you read his article 12:16:26 I have a copy of brad cox's OOP an evolutionary approach. 12:16:36 I picked it up at a used bookstore a while back. 12:16:37 gilbertdeb: great. 12:16:58 gilbertdeb: the one that advocates software ICs ? 12:17:08 yes. 12:17:24 there is another copy of the book with a co-author. I don't have that one. 12:18:53 Novo[a-z]* 12:19:16 whats that? 12:19:30 does he not advocate software-ic in all his writings? 12:19:41 gilbertdeb: does he ? 12:20:20 I would'nt know. I only have this one book by him. 12:20:53 gilbertdeb: unlike C++ objc is very light on syntax 12:21:04 as is smalltalk. 12:21:29 krish: do you use it? 12:22:34 gilbertdeb: starting to ... 12:22:49 which languages are you playing with? 12:23:00 I suspect you are a collector/polyglot 12:25:34 objc is very easy to use and learn, and lots more flexible than C++ as well 12:25:51 and doesnt break C at all.. coexists in armoony 12:25:53 you harmony - 12:25:58 you can mix and match at will 12:26:14 do you use it much? 12:26:27 you too ianni. are you a polyglot? 12:26:38 yeah, I have two Objective C applications in production 12:26:46 what do they do? 12:26:56 GUI programs that interface with RDBMs 12:27:17 match input values to the database lexically 12:27:22 and using other logic 12:27:49 for integrating outside data with our database.. basically lots of table views and Java bridge for JDBC 12:28:24 Java bridge is amazing.. [[@"MyJdbcJavaClass new] callJavaMethod: @"a bridged string object as parameter"]; 12:28:34 what attracted you to forth? 12:28:39 [[@"MyJdbcJavaClass" new] callJavaMethod: @"a bridged string object as parameter"]; 12:28:45 polyglot = more than one language? 12:28:50 I speak a little french.. 12:28:55 a tiny amount of spanish 12:29:01 ianni: as in programming languages. 12:29:06 computer languages, of course, who only uses one language? 12:29:11 :-) 12:29:19 the chip itself. 12:29:20 well 12:29:33 gilbertdeb is back to hex ... 12:29:42 I like FORTH because ... there is nothing even close to it 12:29:51 it's one step ahead of everything 12:29:52 There is nothing simplier. 12:29:54 are you sure? 12:29:55 and it is very philosophical 12:30:03 reading the whitepapers on it hooked me. 12:30:05 gilbertdeb: forth is different, lisp is different ... 12:30:07 the essays and such 12:30:19 i was sold 12:30:37 --- quit: krish ("Client Exiting") 12:30:49 if we are to use the computer to do things... it only makes sense that the best program can be made with full understanding of the machine and the software which runs on it 12:31:52 ianni: thats what finally convinved me to take asm seriously. 12:32:14 but there are humongous languages like commonlisp... 12:32:54 * ianni learned a smidgen on scheme today 12:32:56 and they too offer fantastic advantages according to their supporters. 12:32:57 on=of 12:32:59 --- join: I440r (~mark4@1Cust71.tnt1.bloomington.in.da.uu.net) joined #forth 12:33:03 what did you learn? 12:33:04 *nod* 12:33:32 but a forth program can be so beautiful, very artistic 12:33:36 argh i FUCKING HATE windows 12:33:38 program/os 12:33:44 whatever 12:33:51 I440r: Me too. 12:34:00 gilbertdeb basics of the language only , just got my feet wet 12:34:01 I440r: did you fix the bsd partition problem? 12:34:03 it looks damn niec 12:34:15 ianni: are you very interested in it? 12:34:18 i love simple languages 12:34:18 ive had to reinstall 98 on this stupid fucking machine 4 times already because the fucking pile of shit keeps crashing 12:34:36 * gilbertdeb does too. 12:34:38 assembler is semplest. then comes forth 12:34:38 gilbertdeb: I'm more interested in forth. but it's not really *forth* so much as the power of controlling the machine 12:34:47 so yes, im very interested in scheme/lisp 12:34:56 or anything with different ideas 12:35:13 i love a simple language. thats how they should be 12:35:16 i have to keep UNINSTALLING my sound drivers in order to boot to anything other than safe mode 12:35:33 the boot keeps hanging and NOTHING except removing the soud drivers will fix it 12:35:34 I440r: are the drivers corrupt? 12:35:46 I can't imagine how much I440r hates windows... probably more than I do and I can't imagine hating windows worse 12:35:46 they come right off the sound cards cd 12:36:09 * ianni -class & 12:39:03 also, if i try do a windows update it crashes as soon as it finishes scanning for updates 12:39:17 every time. im trying again now with the sound drivers totally uninstalled 12:39:56 didnt crash this time. so the sound drivers are fucking with me 12:40:05 what about using w2k or wXP? 12:41:01 i just got the fuck rid of xp becaus it doesnt even have sound drivers for this card 12:41:29 besides its a like all ms operating systems, its a DOWNGRADE from the previous 12:41:38 dos downgraded to win 3.11 12:41:56 3.11 downgraded to 95 to 98 etx 12:41:59 etc even 12:43:25 --- quit: serg_penguin (Read error: 110 (Connection timed out)) 12:46:18 3.11 was nice. 12:46:35 considering how utterly tiny it was... 12:46:55 how the hell is it that so many people simply used it and never developed their own? 12:47:18 they did, it's called linux :) 12:47:26 heheh. 12:47:29 lol 12:48:01 don't you like linux? 12:48:09 I love linux 12:48:21 I have 3 linux boxes in my dorm room 12:48:44 I have 3 *nix boxes too. 12:48:51 irix, solaris, debian-linux. 12:49:15 do you run xdmcp ? 12:55:58 who? 12:56:05 natty. 12:56:10 or you. anyone. 12:56:20 nah 12:56:29 not me 12:56:34 I usually tunnel my x clients thru ssh 12:56:36 xdmcp is that program thing which lets you select the X machine you want to loginto. 12:57:04 since xdmcp is grossly insecure 12:57:18 i NEVER use any xdm progies 12:57:19 I don't need to worry about security on an internal network. 12:57:23 ever 12:57:28 gdm? kdm? 12:57:42 when i boot linux i dont want some lame ass bullshit thing starting x for me 12:57:46 i NEVER boot into x 12:57:46 thats lame 12:57:46 hahahaha 12:57:56 why not? 12:57:58 dpkg --purge *dm 12:58:00 don't you run remote X sessions? 12:58:07 yeah me neither, I don't go graphical boot, and I dont' run any display managers 12:58:13 I use the blackbox window manager 12:58:17 i dont need *dm to do that 12:58:24 i use windowmaker 12:58:32 and I use kde 12:58:34 *shrug* 12:58:44 ugh lol 12:58:46 I use 4dwm on the irix. 12:58:46 dude you can buy a 12 inch Iraqi Information Minister doll!!! 12:58:47 it is nice. 12:58:51 i hate kde and gnome 12:59:02 I hate gnome. kde is tolerable. 12:59:08 until I go back to twm ie. 12:59:15 kde and gnome are like windows if you ask me, look like shit and keep falling over : 12:59:20 :) 12:59:26 nonono 12:59:26 use windowmaker!!! 12:59:30 ahahaha. 12:59:32 heh 12:59:34 run my dockap :) 12:59:36 |/_ 12:59:37 like that? 12:59:39 ^ 12:59:46 ya 12:59:49 :D 12:59:50 LOL 13:01:36 kde was the first wm i tried basically and i hated it RIGHT OFF 13:01:52 it looked to me like it was trying to look like windows 95 13:02:09 I first tried icewm and still have it on my bsd box in ohio. 13:02:13 if the box is still alive. 13:02:23 gilbertdeb: your from ohio ? 13:02:32 but I got sick of icewm and switched permanently to twm on there. 13:02:39 relative to being in miami, yes. 13:02:43 why do all the people in ohio ALWAYS drive in the fast lane ? 13:02:45 but not native ohioan or anything. 13:02:53 grandma going 35 mph in the FAST lane 13:02:56 dammit 13:02:59 :/ 13:03:02 I440r: they have how you say? silly tendencies? 13:03:11 I dunno why do people from quebec always drive 100mph stradleing both lanes? 13:03:19 hahaha 13:03:19 bastards in ohio gave me a ticket for CONSPIRING to pull a u-turn :/ 13:03:37 I440r: ohio is almost as bad as mississipi. 13:03:38 take their half outa the middle ??? :/ 13:04:06 --- join: wossname (wossname@HSE-QuebecCity-ppp81121.qc.sympatico.ca) joined #forth 13:05:00 indiana aint so bad except the avarage height/weight is 5 ft nothing and 450 lb 13:08:16 I dunno why do people from quebec always drive 100mph stradleing both lanes? 13:08:21 wossname: do you have an answer ? 13:08:46 nope, i don't live in quebec 13:09:32 e'st ce que tu pas quebecois? 13:10:07 non~ 13:10:09 :( 13:12:32 c'est bien non? 13:34:53 --- join: tgunr (~davec@A17-205-45-104.apple.com) joined #forth 13:40:15 --- join: MrReach (~mrreach@209.181.43.190) joined #forth 13:40:20 Hi 13:40:21 hihi! 13:42:56 hihihi 13:50:17 --- join: Serg_Penguin (~z@ts25-b141.Moscow.dial.rol.ru) joined #forth 13:50:27 hi i440r, hi all ! 13:50:34 just 4 few minutes 13:50:38 greets, Serg_Penguin 13:50:56 whadda u code now ? 13:51:21 serg! 13:51:32 mr r !! 13:51:44 light/sound machine with Win32Forth 13:51:50 hihi, I440r 14:01:59 --- quit: Serg_Penguin () 14:16:00 I440r: whatcha workin on today? 14:22:39 installing windows 98 for the 5th time on this box since yesterday 14:22:50 arg! 14:22:57 anyone here know how the FUCK to get rid of the fucking gay bullshit "view as webpage" shit in 98 ? 14:23:14 i can do it ONCE. if i get rid of it it magically reinstates itself 14:23:25 turn view as web page OFF 14:23:43 go into folder options. set ALL directories with same properties as current 14:23:57 next time i look at ANY dir im seeing the fucking thing as a web page 14:24:02 its NOT A FUCKING WEB PAGE!!!!!!!!!!!!!!!!!!!!!!! 14:24:06 its a fucking directory 14:24:06 fuck 14:24:11 heh 14:24:23 check under desktop properties 14:24:33 i have 14:24:45 thats permanantly disabled 14:24:54 ahah 14:24:59 because if you have your desktop as a web page windows falls over every 2 minutes or less 14:25:29 yes, ActiveDesktop has never been a stable bit of software 14:25:54 it isnt software. its an abomination 14:25:59 like 98% of the rest of windows 14:26:00 * MrReach laughs 14:26:04 good night 14:26:14 :/ 14:26:21 ok, left click the desktop ->properties ... 14:26:25 * I440r calms down now heh 14:26:41 i do it that way 14:26:48 active desktop is disabled 14:27:02 yet my windows explorer keeps showing dirs as web pages 14:27:04 clicke the "web" tab, uncheck "view my active ..." ... then 14:27:09 I440r: u have 2 open only 1 windows expolrer window 14:27:28 it IS unchecked 14:27:31 I440r: disable that view as webpage thing globally 14:27:32 its never been checked 14:27:46 I440r: under the folder options menu 14:27:53 click on the "Folder Options" button 14:27:54 did that too 14:27:57 I440r: then close that window 14:27:58 yup 14:28:09 then i close explordr and when i come back in its viewing as a web page 14:28:11 I440r: & it should stay permanently disabled 14:28:14 k 14:28:23 let me chk the correct settings 14:28:32 check "custom, based on settings you choose" ... then "settings" 14:29:01 mrreach thats what i did to :/ 14:29:02 yup 14:29:18 --- quit: mur ("Congratulations! you have won the game. final score of IRC: 10590") 14:29:43 windows gives you 284529834658923 different intuative ways todo something and only ONE of them actually succeeds 14:29:53 and the "every folder what has web content" (translated from hungarian ;) 14:30:03 check "use windows classic desktop" "only for folders where I select web page" " 14:31:36 took me 2/3 tries to get it to perfarm as I wanted ... can't imagine why you're having such problems 14:31:57 did it :) 14:32:11 nothing is working. it might work if i reboot and try again 14:32:58 he 14:32:58 y 14:33:20 what is the ... 14:33:58 type of view selected? detailed? 14:36:36 detailed 14:36:49 and im NOT hiding extensions of knnown types 14:51:10 HAHAHA! 14:51:26 When librarians stage a protest ... do they still whisper??? 14:53:54 --- quit: flyfly (Read error: 110 (Connection timed out)) 14:55:59 lol 14:57:13 ok i pressed send but it might take a min or 2, mailcity (lycos) can be realy slow sometimes :/ 14:57:32 erm? what? 14:57:36 wrong window 14:57:47 i was mailing something to someone... that was a mischan :) 15:05:52 did you get your folders working properly? 15:06:03 im rebooting now, ill try again 15:06:43 --- join: kc5tja (~kc5tja@ip68-8-206-137.sd.sd.cox.net) joined #forth 15:08:24 * kc5tja sighs 15:08:27 So very typical. 15:08:28 --- join: ramnull (~nicad@12-241-145-39.client.attbi.com) joined #forth 15:08:40 What's up? 15:08:53 how's that? 15:08:54 I show up for my first potential customer for my company, and while I'm out, I get a telephone call for a job interview at Nokia, Inc. 15:09:06 I440r: I'm rewriting that framebuffer stuff. 15:09:41 I440r: Using lookup tables instead of structs. 15:09:45 Of course, I'll continue to keep up the company, just in case the interview with Nokia doesn't pan out. 15:10:01 * MrReach nods, "Tough descisions" 15:10:11 Yeah. 15:10:41 I'll even keep the company on the side, which Nokia should let me since it's for a user interface specialist position (how they got my name for THAT position, I'll never know!). 15:10:41 cool :) 15:11:07 heh 15:11:08 Just one more thing to put on ye olde resume. :D 15:11:16 probably hiring temporarily for focus groups 15:11:57 All jobs in the technical or IS fields are temporary. Unless you run your own company, there's always the threat of being laid off. 15:12:09 Alright, I'm curious. How is a Forth in general supposed to behave when deleting data from a lookup table? Does it reclaim the space? Change the cell/word to all zeros? Leave a "blank" spot in memory? 15:12:28 huh? 15:12:31 it does what ever you tell it to do 15:12:32 Of course I realize... 15:12:41 what word are you "deleting" with? 15:12:42 --- quit: tgunr (Read error: 104 (Connection reset by peer)) 15:12:49 That a lookup table is nothing more than a bunch of allocations. 15:13:20 MrReach: That question answers my question. 15:13:22 ramnull: That depends on the application. Most tables just leave the field as-is, and maintain a separate bit-map somewhere, indicating whether a cell is allocated or not. 15:13:23 using "allocate" ? 15:13:26 or allot 15:13:29 * ramnull grins 15:13:32 actually, a table should probably be a single allocation ... perhaps resized as it grows 15:14:15 i didnt implement resize/realloc whatever yet 15:14:18 So there are no fixed rules for allots and/or allocations. 15:14:18 A multi-level table (e.g., a trie; yes, pronounced "tree") is another solution. Above three levels, it gets unweildly though. 15:14:56 CPU page table TLB lookups are implemented this way. 15:15:04 I440r: I'm using allot at the moment. 15:15:05 Well, for Intel CPUs and Motorola CPUs at least. 15:15:29 Whoa! 15:15:36 I just got it. 15:15:47 Forth = No rules. 15:15:53 Heheh :D 15:15:57 ok, good, your question didn't make much sense, as worded 15:16:04 HAHA! 15:16:13 Think of Forth as the Ultimate Fighting Championship, but with software versus hardware instead of Judo versus kick-boxing. 15:16:24 * ianni grins 15:17:04 * ramnull nods 15:18:16 Hmmm....Well, I'm working on writing a SIMPLE interface for the Linux framebuffer. 15:18:40 which forth? 15:18:44 The C code and the examples over at linuxassembly.org use structs. Structs are a pain in the ass. 15:19:27 Right now I'm using IsForth. But I would like to make it usable for some other Forths eventually. 15:19:58 there's no reason that structs have to be difficult 15:20:15 MrReach: True. 15:20:20 * kc5tja laughs at the assertion that structs are a PITA. 15:20:33 they are :/ 15:20:34 heh 15:20:37 So many Forth programmers make the claim, but don't back it up with anything except opinion. 15:20:57 MrReach: But I'm use to working at very high level(compared to Forth) i.e. Ada95 15:21:14 I've done both table-based and struct-based programming, and I see *zero* difference in expressibility, run-time performance, or other attributes. 15:21:15 many times, allocated memory blocks are the best way to pass wournd chunks of data 15:21:22 I'm finding Lookup Tables to be so much simpler, personally. 15:21:35 and I don't see any way out of that fact 15:21:53 What are you storing in your tables? 15:22:15 But I'm not MrReach, I'm not I440r, I'm not kc5tja...so you're guys milage may vary. 15:22:23 Absolutely. 15:22:26 * ramnull grins 15:22:43 I'm not saying NOT to use tables; I'm just saying many people make the claim, as if it were hard proven fact, that tables are better than structs in Forth. 15:22:54 lookup tables are simply one of many software constructions that can simplify a problem 15:23:02 Yes. 15:23:02 heh 15:23:09 kc5tja: Right now I'm just using it to hold the initial values (term state, color tables, ect...) 15:23:15 However, tables aren't the grand cure-all data structure (sorry, a table is still a structure) 15:23:39 kc5tja: It is a structure. No argument here. 15:24:02 For example, I'm using struct-based implementations for my word parser and string buffer management, because tables simply cannot be used for these applications. 15:24:11 Not, at least, without wasting huge amounts of RAM in the process. 15:24:17 In a 16-bit environment, that'd be suicide. :) 15:24:45 yep ... allocs don't work so well on a 16bit machine 15:24:49 Are there any rules of thumb that I should go by when choosing between tables and structs? 15:24:50 why not? 15:24:52 MrReach 15:24:56 MrReach: Yes they do.. 15:25:10 hard to keep track of memory limits 15:25:15 MrReach: why is that 15:25:21 (asking to ask) 15:25:24 ramnull: Performance is my personal metric when deciding struct-versus-table. 15:25:31 the process works fine, but have to watch resources so very carefully 15:25:41 Tables are higher performance when accessing *ONE* record frequently. 15:26:22 When accessing numerous records, especially in a loop, structs are better; the locality of all the fields makes better use of CPU cache. 15:26:25 (in that case) 15:26:30 It just depends on what you're going to be using it for. 15:26:36 I see. So a network router might use a table, whereas a regular database app might use a struct. 15:26:36 In most cases, it really doesn't matter. 15:27:20 Or at least use structs for an index into the database. The columns themselves are probably best served using tables (not because of speed, but because of the ease with which you can add or remove columsn with a table) 15:27:41 * ramnull nods 15:27:46 I'd also like to note that SQL databases, or relational databases in general, are pretty universally slower than pointer-to-struct (e.g., "object") databases. 15:28:00 I should add, "on modern equipment." 15:28:04 Cache has a lot to do with it. 15:28:23 It just depends if you do predominantly row-major or column-major accesses. 15:28:40 Byte for byte, a table and a struct-based approach should use the same amount of RAM, assuming equal populations of data. 15:29:45 kc5tja: I've also been using the Kdb database which has wicked column-major accesses. But because it's written in K, it's supposedly "Object" based. I dont know what to make of it. 15:30:09 kc5tja: I'm thinking Forth should be able to do something similiar. 15:30:16 I'm not familiar enough with K to know how to respond, but I've seen some of the claims and some sample source code. It looks very interesting. 15:30:21 * kc5tja nods 15:31:14 Wonder if one could hack a Forth to run locally on a GPU. 15:31:19 Note to self: try not to talk much when you have a sore throat. It only makes it hurt more. :( 15:32:02 HAH! That'd be a blast. :) You could set up a pretty wicked AMP system (Asymmetric multiprocessing) that way. Exchange data and programs in source form via shared memory channels. :) 15:32:04 what is GPU? 15:32:18 This no rules concept has got my coming up with wild ideas. 15:32:28 s/my/me 15:32:28 Graphics Processing Unit -- those 256-bit to 1024-bit microprocessors on graphics cards responsible for real-time raytracing and what-not. 15:32:42 ok, gotcha 15:32:51 Well, hypothetically... 15:33:44 If you get a board with 5 pci slots, and plug a decent yet cheap graphics card into it, with a Forth/GPU system, you could have a Beowulf in a box. 15:34:05 Just imagine being able to employ the GPU, at will, as a vector coprocessor for scientific data manipulation and such. 15:34:12 --- quit: I440r (Excess Flood) 15:34:26 * kc5tja nods 15:34:34 --- join: I440r (~mark4@1Cust71.tnt1.bloomington.in.da.uu.net) joined #forth 15:34:38 And hey, why not hack a Forth on some sound card DSPs too? 15:34:42 practically all graphics cards are cheap 15:35:09 ah, all computer stuff is cheap nowadays 15:35:15 it sickens me 15:35:26 why? 15:35:37 oh, cheap rather than inexpensive 15:35:44 MrReach: Quantity versus quality. 15:35:52 sifbot: 2 4 + . 15:35:54 MrReach: 6 15:36:00 Anyone curious enough to try something like that? I'm gonna try it once I get the "skillz". Heh. 15:36:03 sifbot: .( where is the love) 15:36:05 Klaw: where is the love 15:36:08 * kc5tja still has his Amiga 500 locked away in storage. That thing fell down two flights of stairs, and still works perfectly. 15:36:58 back later ... shower time 15:37:03 Maybe I'll port FS/Forth to it someday. :) 15:37:16 i'm still at novice :-/ 15:37:30 i've finally found a place where i can apply forth to my work though. 15:37:40 use it as postgresql procedural language 15:38:05 Sweet. How so? Does postgresql have a mini-Forth interpretter built-in, or does it call out to any arbitrary script interpretter? 15:38:33 Well, I've been looking over the specs for the Ignite processor. Even if one cant get enough documentation from the card makers to do something like that, one could still build a co-processor card using an Ignite based PCI board. 15:38:45 --- join: tcn (~r@tc2-login24.megatrondata.com) joined #forth 15:39:04 hey 15:40:39 In fact, there was company selling G4 PCI boards a while back. 15:41:19 * kc5tja nods 15:41:32 Yeah, they specialized in coprocessor cards designed for heavy scientific computing. 15:41:51 I think the same company produced DEC Alpha cards too. Some of their cards had up to four Alphas on them. Ouch! :) 15:42:28 tcn!! 15:42:37 whussup! 15:42:53 I'm thinking something similiar could be achieved for a lower cost using stack based processors, like the RTX or the Ignite. 15:43:52 RTX, definitely. But I'm not sure of its floating point performance. Of course, for a Forth programmer, floating point probably doesn't mean much. 15:44:02 But for most scientific programmers, floating point is everything. 15:44:26 kc5 unless your an ans "forth" coder... then floating point is required!!! 15:44:26 * kc5tja has 7 65816 CPUs sitting in his possession. I've often entertained the thought of creating a 7-way 65816-based SMP system. :D 15:44:27 :) 15:44:34 I440r: No it isn't. 15:44:41 hi. i'm back from #tunes.. had enough of Water's bullshit :) 15:44:44 I440r: The floating point wordset is optional. 15:45:03 roger waters? 15:45:08 kc5 yes. the wordset is - but i see alot of people LOVE floating point in ans forth 15:45:26 tcn water has always been anally retentive in #tunes 15:45:28 Because it's one of the few things ANS actually gets right. :D 15:45:29 natty: it's his nick 15:45:52 natty_: Ahhh! Another ham who likes Pink Floyd! Small world, I tell ya... :) 15:46:14 Heh. 15:46:17 Though, I like Roger Water's solo stuff too, just not as well. :D He got wishy-washy in his old age. 15:46:18 It seems like fixed point would be a better solution in alot of cases that people default to floating point. However since I'm still on volume 1 of TAOCP, I cant say for sure. ;-> 15:46:22 who doesn't like Floyd? ;) 15:46:25 pink floyd rocks 15:46:30 tcn: No idea. 15:46:35 kc5tja: postgresql allows to easily plug in C functions.. i've got someone trying to plug a Forth into that 15:46:38 tcn: Some young bastards. 15:47:02 my thought is that since postgresql is multithreaded, it could in theory act as an HTTP server. 15:47:07 ramnull: Fixed point is cool stuff. However, it's actually *slower* on modern processors than floating point. 15:47:09 robert: but they sure don't like old Swedish music either, do they?! 15:47:27 but aside from that, pl/pgsql (the postgresql procedural language) is a kludge compared to forth 15:47:43 For example, the Intel Pentium series of CPUs are configured so that IMUL first converts the operands to floating point (1 cycle), then does a floating point multiply (1 cycle), then converts back to integer (1 cycle). 15:47:57 As a result, an integer multiply blows both the integer and floating point pipes for a loop. 15:48:06 The trouble with floatint point is, the error is unpredictable 15:48:22 tcn: How old? 15:48:38 tcn: For the cases where FP is used when fixed point could also be used, the error is usually quite predictable. 15:48:55 Here's what I definitely would LOVE to see more of though -- RATIONAL math. 15:49:06 robert: really old.. ancient.. :) 15:49:09 kc5tja: Couldn't that be overcome through creative use of ADDC, shifts, and MOD? 15:49:17 tcn: By the way.. I don't really get what you mean. I do like Pink Floyd, and I tried to answer your question... kind of seriously. :) 15:49:18 rational, yes 15:49:23 I haven't seen integer rational math since PC/GEOS. Pity, because that's a wonderful way to represent numerics. 15:49:40 ramnull: Please clarify? 15:49:42 robert: well, i wasn't being serious :) 15:50:28 Lisp has rational numbers.. that doesn't make Lisp good, but.. :) 15:50:40 kc5tja: Well, usually the ADDC and shift instructions only take one cycle to complete, whereas an IMUL takes several. 15:51:00 tcn: Yay! 15:51:07 ramnull: But to multiply two 32-bit numbers, you need up to 32 shifts and 32 adds (total: 64 cycles), versus the IMUL, which totals 3. 15:51:09 I usually use MUL .. 15:51:29 IMUL and MUL are architecturally the same. 15:51:30 kc5tja: Good point. 15:51:35 lemme time them.. 15:51:46 tcn: What processor do you have? 15:51:57 AMD Athlons might exhibit different timings than the Pentium series. 15:52:08 (in fact, I'm positive of it) 15:52:30 I have a Pentium MMX and a K6-II-3dnow here 15:52:31 I've been trying to get access to the AMD developer center, but they only let in the big dogs. 15:52:56 * ramnull is using a k6-3. 15:52:58 Yeah. That pissed me off. I wanted specs on the new Hammer series of processors, so that I could support the 64-bit operating mode. 15:53:09 fuck AMD 15:53:16 Fuck Intel. 15:53:25 fuck 'em all! 15:53:59 Intel is the one that royally fucked over (illegally at that) AMD back in the mid-80s, so it's only fair AMD rocks Intel's house. 15:54:10 fyi, this 150mhz mobile pentium is as fast as the 350mhz desktop K62. 15:54:30 So? 15:54:36 That says nothing about the Athlon. 15:54:38 ok, it cost more 15:54:54 Nor does it say anything about the CPUs. 15:55:03 What kind of cache is equipped? 15:55:24 My 455MHz K6-III wiped the floor with my roommate's 200MHz Pentium, which suggests your specs are highly suspect. 15:55:30 it's true 15:55:40 Sandpile has a good writeup on the CPUs. 15:55:41 k6-2 is a painful chip 15:55:49 I don't deny that. 15:55:50 p200 > k62 400 15:56:03 well, don't talk about pre-athlon amd chips :( 15:56:03 the k6-3 may be a lot better 15:56:22 before the athlon, there was no amd 15:56:24 k6-3 is what i got on this laptop 15:56:39 k6-3 works well enough for me. 15:56:53 * kc5tja nods 15:57:08 K6-III is a sweet processor. 15:57:11 I love it to bits. 15:58:15 i've never used a k63 :/ i have an amd 386, it's no better than an intel 386 15:58:27 I'd really like to build an Ignite co-processor card for my PC. Guess I'll have to yank out those elecronics books. 15:58:30 i440r, did you ever get an assembler for isforth? 15:58:30 That's because it's Intel's design. 15:58:36 not yet 15:58:37 y ? 15:58:41 nope :) 15:58:48 i started to code it 15:58:50 Intel couldn't keep up with demand for the 386 or 486 processors back then, so they outsourced fabrication of those chips. 15:58:56 but halted codeing it 15:59:00 for now 15:59:12 blehg 15:59:13 What happened later on, though, was as soon as Intel had their fabs up, they *lambasted* AMD for "copying" their chip designs. What a joke. 15:59:39 i thought amd did 386 illegally, or maybe 486... and they had a huge charge on all outsourced chips 15:59:43 Hmmm....I dont suppose there would be a market for something like that? 16:00:03 i read that they forced the chip companies to create their own specs and charged them on top of it :l 16:00:10 That's just it -- Intel cut AMD's contract without notification. 16:00:28 So AMD kept producing chips, but soon found themselves without payment. That's when the whole lawsuit thing took off. 16:00:37 heh 16:00:50 why if the chip companies are getting screwed don't they rally around sparc? 16:01:05 gilbertdeb: Because there's no money in the SPARC arena. 16:01:13 Everything is x86 for the desktop. 16:01:18 :( 16:01:19 and do you trust Sun? 16:01:21 gilbertdeb: Sparc costs an arm and a leg? 16:01:30 ramnull it is 'open'. 16:01:35 SPARC is an open standard actually. 16:01:45 Like MIPS. Only MIPS has the benefit of being public domain. 16:01:50 Personally I'd go with IBMs new CPU. 16:01:55 This is why there are so many MIPS clones on the market, and MIPS can't do a thing about it. 16:02:11 If they can get it out the door. 16:02:15 i440r: I need an assembler to time the MUL instruction.. gcc compiles IMUL's.. 16:02:22 tcn: NASM 16:02:27 NASM 16:02:29 help me write one 16:02:36 isforth woulda been convenient 16:03:00 kc5tja: I didn't realise MIPS was public domain. 16:03:24 MIPS architecture was design at Stanford University under Dod/DARPA contract. Government money at a educational facility == public domain. :) 16:03:35 s/design/designed/ 16:04:00 hmmm. 16:04:01 Anyone got any opinions on the IBM PowerPC chips that are suppose to be on the way? 16:04:04 i440r: I was writing a pretty good C assembler.. it should be translatable to forth.. 16:04:10 kc5tja: can anyone clone it for no $$$? 16:04:40 gilbertdeb: For the MIPS-I and MIPS-II architectures, I believe yes. 16:04:54 You might not be allowed to call it MIPS, but you can say it's "MIPS compatible" or some such. 16:05:08 cool. 16:05:14 I'll be back in a bit. 16:05:17 --- quit: ramnull ("This isn't Happy Hour!") 16:05:17 ramnull: I'm not familiar with the new PowerPC designs out there. 16:05:26 no wonder Nintendo used MIPS 16:05:40 tcn but they switched to PowerPC 16:05:42 or was that Sony 16:05:51 it makes me wonder if I chose the wrong chip to study... 16:05:52 Both. 16:06:06 tcn are the opcodes done in octal ? 16:06:21 Sony uses a custom superset of MIPS, while the Nintendo 64 used a real MIPS-brand MIPS (since it basically was designed by SGI). 16:06:23 hmm, i forget 16:07:00 :/ 16:07:01 The Nintendo Game Cube uses a PowerPC processor. I'm really not at all sure why they changed processor architecture. 16:07:18 i440r: yeah, octal opcodes (a few in hex, where it makes sense) 16:07:30 yes thats why I am wondering if I should have picked another processor to play with. 16:07:35 but why has SGI abandoned the MIPS? 16:07:56 so far, my r10k 195 has made me happy. 16:08:41 i440r: all the regs, opcodes & aliases are defined in a text file, 1 per line 16:09:15 gilbertdeb: Money. 16:09:45 gilbertdeb: They aren't making any money on MIPS-architecture SGIs anymore. If it sits on the desktop, x86 is the "cheapest" solution, because that's what sells the most. 16:09:50 x86 is fun, actually 16:09:52 if chips weren't that expensive to manufacture, I bet the current trend in software, to do stuff for the hell of it would happen too. 16:09:58 (in reality, even expensive PowerPCs are dollar-for-dollar cheaper than Intel CPUs, but...) 16:10:13 but kc5tja cheapest in what sense? my I2 rocks! 16:10:18 tcn dcc me the files :) 16:10:25 it was manufactured in '97... 16:10:27 erm - zip em and email 16:10:27 gilbertdeb: You're really not understanding business economics here... 16:10:33 its been developed since '95. 16:10:34 gilbertdeb: Cheaper for the company to produce. 16:10:34 sure 16:10:47 Your I2 rocks. No questions about that. 16:10:51 ah I thought you meant cheaper for a customer to buy. 16:10:59 But the company isn't making nearly the amount of money on selling I2s as x86s. 16:11:02 what's an I2? 16:11:04 No, never. 16:11:13 Cheaper to produce != cheaper to purchase. 16:11:17 tcn indigo2 16:11:25 an sgi machine I got off ebay. 16:11:29 it is purple ;) 16:11:33 oh 16:11:40 and it was only 99$s 16:11:49 I remember when the I2 first hit the market, too. 16:11:54 good deal :) 16:12:04 SGI positioned it to compete squarely with the Commodore-Amiga 3000. :) I was proud when that happened. 16:12:12 tcn for a machine that sold at around 40k when it was born, yep :D 16:12:19 rofl 16:12:44 kc5tja: the Amiga is yet another classic machine with die hard fans that makes no $$$ 16:12:57 gilbertdeb: They're not being made anymore, either. 16:13:10 gilbertdeb: But I still have my Amiga 500, and will continue to have it for the rest of eternity. 16:13:13 the new amiga ;D 16:13:31 I guess Amigas had great coprocessors because the 68000 wasn't THAT great 16:13:40 kc5tja: exactly... thats why I asked 'cheaper' in what sense earlier and made that noise about the greatness of my i2. 16:13:45 The 68000 was amazing for its time, what are you talking about? 16:13:45 * gilbertdeb pats the purple one. 16:14:09 the 68000 was really pricey in 1980, too 16:14:20 tcn: This is 1987. 16:14:39 (of course, why they were still using the 68000 when the 68010 was on the market is besides me, but I digress) 16:14:41 kc5tja: but it seems they are currently making machines from chips which are 32 bit, to _await_ the full flared release of intel's 64 bit when they've had 64 bit for eons! 16:14:59 gilbertdeb: Yup. 16:15:00 i've used an early Mac and an early Amiga, both 68000's.. the Amiga was great, the Mac? I'd rather have a IIe or a C64 :) 16:15:12 tcn was it the OS ? 16:15:16 tcn: Don't blame the hardware for the deficiencies of the operating system. 16:15:49 tcn: MacOS is world-reknowned for its sluggishness and stupid memory management policies. 16:15:55 no, the Amiga had excellent graphics & audio chipsets, like the C64 before it 16:16:02 kc5tja: sluggishness? 16:16:06 i thought that was windows? 16:16:08 :P 16:16:08 ianni: Yes, sluggishness. 16:16:12 the Mac was slower even with its little B&W display 16:16:17 old mac os's always seemed REALLY snappy 16:16:20 ianni: Remember, windows is still trying to be like MacOS 1. 16:16:28 * ianni never used pre-8 16:16:39 but then 16:16:46 i guess lots is snappy compared to explorer.exe 16:16:47 ianni: Then you've never used AmigaOS. :) 16:16:51 kc5tja :) 16:17:14 tcn: No, sorry, it wasn't. 16:17:33 tcn: The 68000 is demonstrably faster than a comparable 80286 or 80386 processor at the same clock speed. 16:17:45 kc5tja: it seems that anytime there is a terribly popular but relatively low quality product, there is a high quality product to match it lingering in obscurity. 16:18:21 dirty water always surfaces. 16:19:08 gilbert: heh.. i'm seeing that in the HVAC industry now :( 16:19:51 Hvac? 16:19:59 high voltage ac ? 16:20:00 high volume .. 16:20:32 some upstart cuts into our market with a gimmick, undercuts us, and we have to copy the gimmick.. now his product is failing after a couple years, but the damage is done.. 16:20:34 Heating, Ventilation, and Air Conditioning 16:20:39 * tcn nods to kc 16:20:53 (My dad was in that business) 16:21:16 and if our product fails similarly, the whole industry is hurt.. people'll go back to wood stoves 16:21:29 tcn: in 1886, the AFL was formed to prevent that sort of thing happening to craftsmen :) 16:21:39 and you thought the history of labor unions was irrelevant. 16:21:42 :D 16:21:48 gilbert: craftsmen? it's mass production now.. 16:22:03 it's a damn shame 16:22:05 but the business is specialised, with a certain set of standards. 16:22:23 the designers remain craftsmen as there aren't that many of them to go around. 16:22:33 in order to boot to windows 98 i have to boot to safe mode, uninstall my sound drivers, reboot and reinstall them 16:22:44 the deployers can be s/burgerking/hvacdrone/ 16:22:52 fucking pile of shit operating system is fucking pissing me off 16:22:55 I440r: Oh, that's normal for Windows. :D 16:23:17 "mouse movement detected, please reboot your machine to let the change take effect" 16:23:38 You rebooted your machine. Please reboot your machine for changes to take effect. 16:23:50 hehehe 16:23:55 heh.. this machine has win98 on it.. wouldn't boot when I enabled the external serial port!! 16:24:06 delete all files, [yes] [yes] 16:24:18 or was it because I had to disable the IR port to enable the serial port? who knows.. 16:24:43 win98 couldn't find my IR port so it just waited for it with a black screen. how nice :) 16:25:46 heh 16:27:08 tcn: I say the paperful office should make a comeback :D 16:27:22 computers have _increased_ the amount of paper/paperwork in an office. 16:27:24 there is no paperless office ;l 16:27:31 * kc5tja agrees with gilbertdeb 16:27:42 i've printed out a couple of electronic books, actually 16:27:45 Despite having a computer handy, I find I do nearly all of my arithmetic work on paper. 16:28:03 i would never have printed out these books without a computer?! 16:28:06 kc5tja: nothing as versatile as the paper-pencil duo 16:28:22 Nope. 16:28:42 and with the addition of a straight edge and a compass, no geometrical shape is beyond it. 16:28:56 the possibilities are endless! 16:28:57 Except the triangle. 16:29:01 the future, is bright. 16:29:03 gilbertdeb: construct the trisection of an angle 16:29:11 I440r: it is an impossibility. 16:29:16 :) 16:29:20 * kc5tja nods 16:29:21 :) 16:29:28 yeah, if I had a scanner or something I'd write everything on paper and scan it in :) 16:29:29 I studied geometry. 16:29:33 Yes on my own. I was bored. 16:29:38 really bored 16:29:45 Nothing wrong with studying geometry. 16:29:58 I studied electrical engineering ... on my own. 16:30:01 heh i had a professor of maths going over a method i came up with trying to disprove it - it took him 2 weeks to figure out what was wrong with it 16:30:09 kc5tja: not at all. I went on to algebra, trigonometry and then Calc1 in the same 3 mth session. 16:30:11 fun times. 16:30:24 he did it all with equasions. i could have told him why it was a flawed method heh 16:30:29 without going near any math :P 16:30:32 Oh, yeah, this is another major issue I have with computers nowadays. 16:30:40 SOlid state implies no moving parts. 16:30:46 So why the hell do I have so many frigging fans in my box?! 16:30:51 hahahah 16:30:56 hahahah 16:31:07 the i2 has 3 fans and makes noise. 16:31:10 I bring this up because one of my fan's bearings is starting to fail. I can't locate the fan that's making the noise. 16:31:19 There's a possibility that it could even be the harddrive. 16:31:23 stick your finger in each one :) 16:31:42 THough it makes the same noise with either the Linux harddrive or XP HD in it, so I'm pretty close to 99.9999% confident it's not the harddrive. 16:31:51 tcn: That requires I open the box. :) 16:32:04 fans tend to be junky 16:32:08 it's a fan 16:32:21 I totally fear the day when liquid cooling becomes a necessity. 16:32:23 And it will. 16:32:33 --- join: bwb (~bwb@ip68-4-121-108.oc.oc.cox.net) joined #forth 16:32:38 nope.. air cooling is the wave of the future.. passive air cooling.. 16:33:03 My friend has his Pentium over-clocked to 4GHz or so, and it's liquid cooled. It's pretty neat -- the whole setup looks amazingly like the cooling system of a car, right down to the radiator and polyethalene glycol. 16:33:16 the future is a computer utility~ 16:33:21 hmmm "death to smoochie" sounds like a GAY movie :/ 16:33:25 hahahahhaha.. he's nuts 16:33:28 what are we trying to do? cool down computers 80% of whose cycles remain idle. 16:33:30 it is. its got robin wiliams in it 16:33:32 its GAY 16:33:32 I440r: it was fun 16:33:37 now thats funny. 16:33:46 robin williams and edward norton and what's her name 16:33:48 i do believe 16:33:54 tcn: No kidding. He has the CPU and the GPU cooled, plus some of the motherboard chips, all on one fluid circuit. 16:33:55 and jon stewert 16:34:17 At least it's quiet though. 16:35:01 130+ channels of fuck all 16:35:14 but i get to watch them on my computer heh 16:35:16 you seen those new VIA boards? some are fanless.. they're not fast.. well, alot faster than this old box! 16:35:35 we got 4 satelite decoders in the house now, one of them is in my room connected to my tv card :) 16:35:46 i thought you were foreclosed 16:35:59 were in a rented house now 16:36:21 hehehe and the bare necessities are getting installed :) 16:36:33 heh.. watch what happens.. bank will sell the house to itself for $20000 and give you nothing back 16:36:37 I440r: can you stream daria? 16:36:42 actually, the only channel in here thats absolutely required is fox news 16:36:48 ? 16:36:52 ye gads. 16:36:54 on a dial up ? 16:36:59 not _Them_ 16:37:03 they'll BILL you for the difference 16:37:19 I440r: did'nt they throw in satellite broadband ? 16:38:20 i wish! 16:38:58 a dish to watch the news with. I stopped listening to what they had to say a while back. 16:39:16 if they'd guarantee daria and family guy, I might get a dish :D 16:40:40 i440r: i've got an expression parser that ought to work for that assembler.. 16:40:51 haven't merged it in yet 16:40:54 in forth ???? 16:40:55 :) 16:41:07 what do you think? 16:41:38 :/ 16:41:55 I'm back ... miss me terribly??? 16:41:56 oh, and ./m won't update the opcodes file.. 16:42:10 ./m ? 16:42:27 you have to run ./optable if you change the 'opcodes' file, to rebuild opcodes.h 16:43:01 what are you building? assembler or disassembler? 16:43:09 well you can guarantee i wont be doing anyting C with these files, ill take a look at them when i pluck up the courage 16:43:26 and try convert them to a sane language - but c can be very difficult to convert to forth 16:43:35 fucking dumbass local vars 16:43:36 :/ 16:43:37 heh 16:43:54 locals rule! 16:43:59 in fact ... 16:43:59 not in forth 16:44:02 --- quit: wossname ("mmmmmm phi") 16:44:15 I discovered this week that win32forth has LOCALALLOC 16:44:28 LOCALALLOC ( u -- adr ) 16:45:09 allocates u bytes of memory on the return stack and returns its address as adr ... it is autofreed when the word exits 16:45:36 (rather like C does it @:^) 16:46:17 it's *WAY* cool because the adr can be passed to any constituant word, and it's guaranteed to be there 16:46:46 I'm toying with a new language, like C with Forth influences 16:47:33 what will you end up with? 16:48:04 it'll have C syntax, minus some obscure stuff.. it'll have an interpreter-compiler option.. 1 stack, though.. 16:48:14 awwww c -snytax? 16:48:20 is'nt that obsolete? 16:48:42 heh, didn't lisp declare pseudoalgebra obsolete? ;) 16:48:59 s/bra/&ic notation/ 16:49:03 heh 16:49:31 it'll also allow assembly and *real* algebraic notation 16:49:44 no thats nearly impossible. 16:49:48 with implicit multiplication, fancy symbols, etc.. 16:50:00 real algebra is 2 dimensional. 16:50:20 ever seen Maple or Mathematica? 16:50:28 mmm yes. 16:50:53 this would need a special editor, but it would just parse an expression and pass it on to the main compiler.. 16:51:15 could be C, Lisp, Calculus.. 16:51:27 maybe even Forth 16:51:37 and which problem would it solve? 16:51:56 verbosity :) 16:52:02 oh really? 16:52:13 j/k/a+/apl 16:52:40 and nail too. 16:52:48 always gets left out that nial. 16:52:55 i dunno.. aren't those languages a little TOO dense? 16:53:13 they are not verbose at all :D 16:53:29 K may be fast compared to Java and SQL.. 16:53:45 and small! 16:53:59 small is good 16:54:29 K reminds me of Forth so much I don't even wanna go there :) 16:54:47 thats good. make a freek! 16:54:50 whats wrong wtih forth that you hate it so much tcn ? 16:54:54 tcn: what don't you like about forth? 16:55:11 free k 16:55:20 heheh. that came out wrong did'nt it? 16:55:26 herk: read my paper 16:55:32 tcn: I did 16:55:35 tcn url? 16:55:49 tcn: the one about why you left retro for small C or something 16:56:04 http://bespin.org/~tom/writings 16:56:41 I should rewrite it, but it basically says what I want to say 16:57:51 i440r: Have you tried using shared libraries in isforth? 16:58:13 tcn: any good resources to learn about small c? 16:58:29 google :) 16:58:54 tcn: you like the turbo pascal IDE better than any smalltalk's? 16:58:57 tcn no. its not something im in any big rush to add either 16:59:09 look for SCC6502, SMC88DOS, Small C/Plus, in particular, and "Let's Build A Compiler" by Crenshaw 16:59:10 tho... i prolly will eventually 16:59:39 I've never used Smalltalk 17:00:14 although I've seen enough of Smalltalk to doubt I'd like it 17:00:26 tcn: k, thanks 17:00:53 tcn: you don't do squeak at all? 17:01:05 the tunes people haven't got you by the throat yet! 17:01:11 heh 17:01:46 i440r: Reason I ask is, a C-like interpreter would get along with the C libs better than Forth.. and it would make it easier to use existing Linux stuff while improving it.. 17:01:56 tcn did you write any control structures in retro? 17:02:03 i440r: ie, using SVGAlib for graphics 17:02:23 herk: yeah 17:02:29 tcn: loops? 17:03:25 gilbert: i tried squeak once.. maybe I didn't get the point :) 17:03:25 tcn well i was going to give isforth a frame buffer extension.... :) 17:03:47 someone is actually working on that right now but he is a beginner. im giving him a chance to write it tho 17:04:10 ah Tcn now I don't feel too dumb :D 17:04:23 i440r I want create a simple window-manager type program. 17:04:35 :) 17:04:50 it's a mickey-mouse language 17:04:51 tcn: have you read this: http://cliki.tunes.org/Forth%20is%20NOT%20intrinsically%20slow 17:04:56 tcn literally. 17:04:57 heheheh yeah 17:05:25 herk: yeah, well, a few times, but not this week :) 17:05:31 tcn is a smallC compiler as relatively easy to write as a forth? 17:05:48 tcn: did you write loops in retro? 17:06:10 forth is *MUCH* easier to write 17:06:15 what were you thinking? 17:06:57 oh, I misread 17:07:10 herk: to be honest, i don't remember.. look :) 17:07:52 why do you ask? 17:07:57 tcn: I can't help thinking that you never really got the hang of forth. 17:08:06 calling libs in forth is not incredibly complex 17:08:18 complaining about stack twiddling is a surfire sign of a newbie 17:08:23 mrreach: yeah, like in winforth 17:08:35 the syntax is marginally more involved than C ... but it's fairly easy and relatively efficient 17:08:47 and it seems like you just wrote an OS in asm that has a forth interpreter. 17:09:02 C translates directly to Forth, Forth can never win the race. ... tcn I don't understand that. 17:09:19 if one becomes like another, then they are the same! 17:09:31 is that what was meant? or am Imissing yet another point? 17:09:31 yeah, i didn't do much forth programming, decided I didn't like it, so I never did "get the hang of it" you might say.. 17:09:45 forth *CAN* win the race ... some modern compilers output code as tight as any C compiler 17:09:56 tcn: ok, that makes sense. 17:10:02 tcn give forth a chance :) 17:10:07 mrreach yea - iforth 17:10:11 tcn: for some reason I started off with the impression that you were a hardcore forth guy for a while 17:10:13 i still DO use forth for little things 17:10:14 note... NOT isforth :)\ 17:10:43 ie, interactive forth 17:10:59 actually, I use TCL for little things 17:12:23 i dont think theres ANYTHING worth using except assembler and forth 17:12:42 well, I had loops in Retro 2 17:12:50 I440r: how far along is your assembler? disassembler? 17:13:08 tcn: Squeak is one of those "pointful" languag eenvironments. Don't feel bad if you "don't get the point" -- that's natural. 17:13:11 FOR-NEXT, BEGIN-UNTIL, BEGIN-WHILE-REPEAT, DO-LOOP.. 17:14:06 kc: maybe it's a proof-of-concept, which would explain why it's in its own little world and doesn't get along with Unix too well.. 17:14:22 You know, in the years that I've programmed in Forth, I've used DO/LOOP and its cousin FOR/NEXT less than 10 times total. 17:14:35 kc5tja: what do you use instead? 17:14:44 noloop? 17:14:47 tcn: Smalltalk is supposed to be THE operating system, not a task under a hostile OS. 17:14:50 begin-while-repeat 17:14:54 gilbertdeb: I use BEGIN/WHILE/REPEAT. 17:15:04 For me, it just makes more sense. 17:15:30 kc: heh.. if it can't cope with another OS, it's too goddamn fussy :) 17:15:45 tcn: agreed 17:15:54 Well, then C is pretty fussy then. 17:16:31 I'm thinking of a BEGIN ... WHILE/UNTIL (condition) ... LOOP in my new language 17:16:41 C has one hell of a time interfacing to an OS where Pascal (I won't even get into Oberon!) is the host language. 17:16:56 ouch :) 17:17:16 Most people don't realize that C is not exempt from the Language/OS unification issues. 17:17:32 kc5tja: when I was playing with squeak, the only way to run it was by itself, full screen with no wm running. 17:17:52 I'd say that the Pascal OS was poorly designed 17:17:58 gilbertdeb: I've never had any problems running Squeak in a window under Windows or Linux 17:18:03 tcn: No. 17:18:06 tcn: Not even close. 17:18:15 kc5tja: it made more sense since itself had a window manager 17:18:17 Pascal requires specific run-time requirements. 17:18:18 even Forth gets along fine with a C-based OS 17:18:21 C doesn't meet those requirements. 17:18:22 it felt a tad redundant to have a wm within a wm. 17:18:31 Oberon is worse: it has garbage collection! 17:19:04 kc5tja: is gc bad? 17:19:08 gilbertdeb: No 17:19:15 GC? sounds like Tunes 17:19:18 But try to get C to talk to a language that depends on garbage collection. 17:19:39 (and I'm not talking about reference counting; I'm talking REAL GC here) 17:20:18 kc5tja: whats hard about that? 17:20:22 C really is a just glorified assembler :) 17:20:34 ianni: In C, how do you tell the difference between a pointer and an integer at the binary level? 17:20:59 ianni: How do you create a structure to be used by an Oberon program? 17:21:09 garbage collection causes the taget data to be moved aorund 17:21:16 i dont know what Oberon is.. nor the answer to the previosu question 17:21:30 a C pointer might then be invalidated 17:21:31 ianni: or the reverse: if the Oberon program creates a single reference to an object, and passes it to C for processing, how does the environment know NOT to deallocate the object (remember, once passed to C, it no longer has a reference!). 17:21:47 any language that requires runtime time ID is gonna be slow.. and if you write the OS in that language, well.. it's gonna suck. 17:22:19 ianni: (a) You know that oberon is garbage collected, (b) you know that Oberon is a language, and (c) you know that it's hard to interface with C (otherwise, you never would have asked me what was so hard about it). 17:22:37 tcn: That just goes to show you how much you know about Oberon. Oberon is **NOT** slow. 17:22:47 It's easily at least as fast as C. 17:22:52 unless maybe you build special hardware to handle GC/types/objects efficiently.. there was a now-defunct Scottish company that did that.. 17:23:16 or the Lisp Machines.. 17:23:18 kc5tja: where do you suppose C gets its speed? from implementation? 17:23:28 object oriented hardware ????? 17:23:28 ugh 17:23:32 everyone uses C as a speed comparison. 17:23:43 C IS NOT INHERENTLY FAST! 17:23:51 gilbertdeb: Two things: one, CPU vendors cater to the C community, and two, the C compilers have sufficiently high-level information about a program to create nearly ideal instruction sequencing. 17:23:52 then where does the speed come from? 17:24:05 tcn: It is when the hardware vendors caters to it. 17:24:27 kc5tja: like which ones? 17:24:29 x86 ? 17:24:43 gilbertdeb: Intel -- they renovated the 16-bit architecture to support C when they created the 32-bit architecture. 17:24:57 hopefully they are the only ones? 17:24:59 gilbertdeb: The 68000 was designed *specifically* to run C and Pascal languages equally well (since both were equally popular back then) 17:25:06 gilbertdeb: *ALL* RISC architectures today, especially MIPS. 17:25:13 oh no. 17:25:18 C (or any typed language) makes it easier to write a fast compiler.. but you can still write a slow one. 17:25:39 tcn: is it easy/fast to write a smallC compiler? 17:25:43 tcn: C'mon. Look at the specs on code produced by modern C compilers. They're very often as fast as hand-written assembly, if not faster. 17:26:02 They have a whole battery of optimizers that do so many passes over the code it's not funny. 17:26:04 AHEM, C/Unix was developed on the PDP-8 & 11, of which the 8086 is a direct descendant. 17:26:11 Bullshit. 17:26:24 The 68000 has an instruction set that is 100% recognizable to any PDP-11 programmer. 17:26:30 The 8086 is a descendent of the 4004. 17:26:44 Right down to the 8 general purpose registers. 17:26:51 look at the PDP-11 registers and opcodes, then an 8086. Same thing. Both little endian, too. 17:26:57 Hmm... 17:27:14 MOV (R0)+,-(R1) looks astonishingly like move.l (a0)+,-(a2) to me. 17:27:15 I guess the PDP was very influential 17:27:17 Nothing at *ALL* like x86. 17:27:48 I strongly suggest that you actually spend time with these processors before making such claims. 17:27:50 gilbert: It's harder than writing a Forth compiler. 17:28:08 kc5tja: given that all the popular chips are designed to make c go faster, it makes no sense then to write in anything but c no? 17:28:20 no 17:28:32 gilbertdeb: Not necessarily true. 17:28:40 I am all ears. 17:28:45 all eyes. 17:28:48 interpreters and interactivity are not to be discarded lightly 17:28:53 gilbertdeb: Though other languages can exhibit slower run-times, remember that most programs are heavily I/O bound. 17:29:02 gilbertdeb: Therefore, they spend most of their time *waiting* for the user or some I/O device. 17:29:03 --- join: ramnull (~nicad@12-241-145-39.client.attbi.com) joined #forth 17:29:14 If this is true, then one can get away with a little inefficiency in exchange for quicker time to market. 17:29:20 THis is what made Visual BASIC so popular. 17:29:22 kc5tja: they are generally underutilised. 17:29:33 which makes me wonder, how are they able to sell gigahertz computers? 17:29:44 gamers.. 17:29:47 --- nick: MrReach -> _MrAway_ 17:29:49 ah them. 17:29:56 and newbies 17:29:58 tcn: aside them? 17:29:59 Yup. We wouldn't have GHz machines if it weren't for both Windows and the PC gaming industry. 17:30:01 oh those too. 17:30:36 Or Microsoft in general. It was quite laughable when Microsoft claimed that a 68020 was too slow for MS Word, and the Mac user should upgrade to the 68030. 17:30:37 heh. I bought a .15 ghz machine last year, about the same time my little brother got a 1.5 ghz :) 17:30:39 hmmm. I wonder, why didn't sgi push their machine as a superior gaming machine giving its graphics superiority? 17:30:55 they did.. N64 :) 17:30:56 good job tcn. 17:30:57 what did you do? 17:31:06 i mean get.. 17:31:17 a used laptop 17:31:21 cool! 17:31:25 tcn: And if they'd only released the CD-ROM for that platform instead of using cartridges, it WOULD dominate the gaming industry. 17:31:27 yeah, it's nice 17:31:35 But that's OK - at least PS2 uses a MIPS at heart too. :) 17:31:42 tcn was it expensive? 17:32:15 --- join: thin (~thin@acc-1-02.cariboo.bc.ca) joined #forth 17:32:20 I could have gotten a new PC (without monitor) for the same price 17:32:29 hi kc5tja i440r bwb _mraway_ tcn etc :P 17:32:33 re thin 17:32:36 hi 17:32:39 for a 15 ghz machine? 17:32:45 oops mhz 17:32:49 150 17:32:49 150 17:32:55 ;) 17:33:11 Hey, when I build my 65816-based system, I'm going to run it at a whopping 10MHz. I have zero need for anything faster. 17:33:14 Has anyone looked at the specs for the Ignite processor? 17:33:21 my 150 is quicker than his 1500 with XP, kazaa, viruses, etc :) 17:33:35 What OS do you run on the 150? Linux? 17:33:42 (or some related technology like BSD) 17:33:44 bare bones linux! 17:33:47 * kc5tja nods 17:33:50 No X11 then. :) 17:34:01 still goddamn fat, but it works. exactly, no X11, no glibc2 17:34:02 kc5tja: so given that most chips await user input and grossly underutilize the chip, lisp remains untouchable! 17:34:38 gilbertdeb: I assume you mean the software awaits the user? 17:34:45 gilbertdeb: and not the chips? The chips couldn't care less. 17:34:47 a Lisp *compiler* might be ok 17:34:49 kc5tja: :) 17:34:59 tcn: not even a compiler, an interpreter. 17:35:27 Hey, if Python can run Zope on commercial servers without compilation, so can Lisp interpret its code (which is in a form much more amenable to interpretation!) in live software. 17:35:28 well, if it's a dead language why learn it? 17:35:34 Because it's not dead. 17:35:37 Any more than Forth is dead. 17:35:39 things haven't changed that much and from the look of 'things', getting or making a reallly fast computer if there is no need for heavy number crunching or graphics is rather stupid. 17:35:50 kc: that explains how they sell 4ghz machines 17:35:55 I <3 forth 17:35:57 Yahoo!'s store is written in Lisp. 17:36:04 tcn: Bull!!!! 17:36:04 kc5tja: yes, that's true 17:36:07 C'mon!!! 17:36:09 Use your head! 17:36:11 kc5tja: no longer. they converted it to c++. 17:36:17 kc5tja: and it pisses me off to no end 17:36:22 * kc5tja is getting fed up with the fscking lies going around here. 17:36:34 Paul Grahm could almost convince me, if anyone could :) 17:36:36 natty_: ?? 17:36:59 Not so much because it's lisp based mind you, RTML was pretty easy to learn 17:37:03 tcn: I've *seen* Lisp code compiled, and it really is quite fast. Substantially faster than equivalent ML or Haskell code. 17:37:04 that "Taste for Makers" article is damn good, of course, it'll probably lead you to Forth instead.. 17:37:15 hold -- phone 17:37:21 RTML? real time markup language? 17:37:22 My biggest complaint is that I can't upload and download the site as a set of files 17:37:33 roomie got it. 17:37:41 ReTard ML 17:37:49 oh. 17:37:51 i don't even remember what it stands for 17:38:02 Real-time ML 17:38:03 it's yahoo's store language 17:38:07 Ohh... 17:38:07 that. 17:38:10 Yeah, I have no idea. 17:38:32 never dealt with Yahoo! store except to actually purchase stuff. 17:38:35 it's based on ML, as in SML and OCAML? they expect store owners to learn that??? 17:38:48 tcn: I learned ML real quick. 17:38:53 I think he meant markup language. 17:39:01 tcn: Anyone who's been through high school algebra can use ML. 17:39:13 gah I hate ML so very much 17:39:25 This is processor looks like something I could work with. 17:39:26 tcn: No it's based on LISP 17:39:39 based on lisp? 17:39:41 RTML == Yahoo Store's LISP based markup language 17:39:42 I didn't think so! 17:39:43 I find it amusing people program in Forth (arguably a functional programming language) but dislike ML or Haskell. 17:39:47 ML == teh suk 17:39:48 yeah, i guess it's kinda like BASIC.. i've even used ML a bit.. but then I've seen some fugly ML programs 17:40:14 Well, I agree -- the programming styles for ML haven't developed yet. 17:40:32 Forth and Lisp have the advantage of not being pure functional languages, and so, can make imperative optimizations where appropriate. 17:40:36 kc5tja: s/programming styles/programmers outside academia/ ;) 17:40:39 Ocaml seems to be doing well. 17:40:42 Going between ML and C is almost impossible since ML uses the characters so differently 17:40:55 natty_: Same could be said for Forth too. 17:40:57 at least for me 17:41:11 Nah because forth's syntax is mostly completely removed from C's 17:41:15 Although I'm kinda partial to J, K, and the Algol/Wirth line of languages. 17:41:16 Especially when you see things like : foo ( a b -- c ) 17:41:16 :) 17:41:19 i'm used to different operators 17:41:25 It looks like a function definition with parameters a b and c. :) 17:41:38 haha 17:41:41 Oberon is so far my favorite systems programming language. 17:41:56 kc5tja: yeah you mentioned that. 17:41:57 Well, infix programming language. 17:42:01 aren't the compilers expensive? 17:42:03 them Functional languages seem like an academic thing.. 17:42:05 I need to try oberon again 17:42:09 Forth is my favorite postfix language, and Lisp is my favorite prefix language. 17:42:13 tcn: right from birth. 17:42:23 tcn: Everything we do is rooted in academia. 17:42:25 Any of you guys use Ada? 17:42:28 tcn: then again, the algebraic style was meant for academia as well. 17:42:34 which leaves COBOL for the rest of us :O 17:42:37 ramnull: Too verbose for my liking. And too clumsy. 17:42:42 I've used VHDL and Pascal which are both similar to Ada 17:42:49 VHDL more so 17:42:50 gilbertdeb: AppleScript... :D 17:42:52 but we all learned algebra as teenagers 17:42:55 You'll find bits of all those languages in Ada. 17:43:00 VHDL is a proper subset of ADA, IIRC. 17:43:05 algebra was fun to learn. 17:43:19 kc5tja: Every seen the Spark subset? 17:43:34 kc5tja: everything we do is rooted in academia-military 17:43:43 Yup. Algebra is the root of writing efficient Forth programs too -- factoring, optimization by removing terms, etc. 17:43:50 kc5tja: You really gotta do something deliberate to fuck up a Spark program. 17:43:52 ramnull: Nope. Never heard of it. 17:43:58 actually I started on algebra & trig at age 11-12.. i knew it was powerful stuff :) 17:44:07 kc5tja: www.sparkada.com 17:44:19 ramnull: Ditto for Oberon. It's type system is provable -- it has only three (well known and documented) loopholes, all three of which require the programmer to do something VERY deliberate to exercise. 17:45:05 Hmmm...if Chuck Moore and Nicholas Wirth collaborated on a programming language, what would it look like? 17:45:16 They'd both kill each other. 17:45:27 hehe :) 17:45:41 it would look like TUNES!!! (thin air) 17:45:56 The nice thing about Oberon's type system is that it's strong enough to virtually do away with hardware based memory management units (for protection purposes at least; not so for virtualizing memory). 17:46:05 TUNES has Forth people and Lisp people.. ack.. :) 17:46:22 Forthers and Lispers get along very well with each other. Quite unlike Cers and Lispers. :) 17:46:45 I'm all for the speed and flexibility of Forth. But for goddamns sake I hate it when programs crash, hang, and especially when the mangle data. 17:46:52 oh right.. prefix/postfix for the algebraicly challenged :) 17:47:19 tcn: There's strong evidence that, before algebra classes in school, the overwhelming majority of humans think/process data in postfix. 17:47:43 Note that you don't select BOLD, then highlight text in a wordprocessor. You always highlight FIRST, then select "bold" SECOND. 17:47:46 kc5tja: only reason I came to #forth was because someone in #lisp said it was worth it :D 17:47:51 kc: agreed, but there are advantages to thinking about math problems in infix 17:47:56 gilbertdeb: Did we let you down? :) 17:48:03 hehehe. 17:48:05 no. 17:48:09 When I get to doing serious programming in Forth, you're likely to see ten lines of test code for every line of code. 17:48:12 tcn: There are also advantages to thinking of them in prefix too (e.g., sigma notation). 17:48:16 it's the hottest channel in here :) 17:48:34 damn straight 17:48:42 fierce debates :) 17:48:47 I downloaded Oberon, and played with it. Never got into it much. 17:48:55 god, i've been in here 3 hours 17:48:57 ramnull: I write software in C commercially. Guess what? Despite C's use of types, I still end up writing five to ten unit test functions for each function I write in C. 17:49:05 tcn: good for you, go for the record! 17:49:17 The result: next to no use of debuggers whiel writing code, provably correct code, and very low bug-counts. 17:49:18 kc5tja: wow a perfectionist :) 17:49:27 fortunately I only pay for 5 hours a day 17:49:28 No, I just practice test-driven development. 17:49:30 the record might be 36+ hours straight? ;) 17:49:41 tcn: ahhhh, an automatic safety valve, sneaky 17:49:45 kc5tja: how is your stuff provably correct? 17:49:47 kc5tja: The Spark guys claim that you can statically predict each and every branch your code is going take. 17:49:55 kc5tja: thats a good call. I also tend to test my stuff pretty well, but I usually do it by hand rather than writing functions 17:50:04 Of course I've never written anything incredibly complex 17:50:11 ramnull: only if you know your input in advance 17:50:26 gilbertdeb: Every build, I run the unit tests to ensure my previous changes didn't break the software. The unit tests are written in a form similar to formal methods. They permit reasoning about the code. 17:50:26 Usually I think through my algorithms to the point that I've already taken every scenario into account anyway 17:50:40 ramnull: Bull. I refuse to believe that until I see *one* independent and professional paper on the subject. 17:50:48 I spend about 10 times as long thinking about my code than I do actually writing it 17:50:55 tcn: Only if you write your program to only accept certain input, and send the rest to dev/null or somewhere else. 17:51:05 natty_: which is why pencil and paper remain rulers. 17:51:14 natty_: Thinking is good, but you can't think of everything. I tend to write what I need when I need it, even if it proves to be inefficient later on. 17:51:15 I am promoting PP solutions :) 17:51:24 kc5tja: They got papers up at that website explaining it. 17:51:30 ramnull: From them. 17:51:38 I want to see it from Berkeley, or from UCSD, or from MIT. 17:51:42 kc5tja: Nope. 17:51:43 ramnull: useless for the real world.. 17:52:14 --- quit: I440r (Read error: 104 (Connection reset by peer)) 17:52:24 Well, it's what the military uses for thier fighter jets, subs, etc... Britain is using it to control thier railway system. 17:52:32 I dunno, I would argue that you MUST think of everything, otherwise even if you write unit tests you cannot be certain that your code is correct 17:52:35 i mean, how can you predict what input an antiballistic missile is going to see if the north koreans fire something at us in 5 years? 17:52:43 but regardless, I must get back to my USB driver 17:52:49 see ya natty 17:52:56 ramnull: That doesn't prove anything except that it works for their applications. Ada is a good language, but not even it is exempt from basic rules of boolean algebra. 17:53:06 ramnull: That being, you can never prove a negative. 17:53:40 Don Knuth has $4096 up for anyone who can find a bug in the latest TeX sources. 17:53:49 But he's not saying that it is certified bug free. 17:53:50 heh.. the military is using Access for a lot of less-critical things :) 17:53:54 There's a big difference. :) 17:54:13 tcn: And WIndows NT for their destroyers. That (*@()*&#% frightens me. 17:54:54 i doubt they run stuff like weapons on NT 17:54:58 kc5tja: whats Tex written in? 17:55:01 gilbertdeb: C 17:55:02 and why is he so confident? 17:55:04 --- quit: thin ("gtg") 17:55:05 I know some of the stuff in the field runs on solaris 17:55:10 gilbertdeb: Because it's damn reliable software. 17:55:24 we don't run lasers on windows, but the layout program is dos or windows.. 17:55:24 because of the language? 17:55:29 or because of the way he wrote it? 17:55:30 Of course, he's had 30 years to work the bugs out. 17:55:30 for th earmy 17:55:32 gilbertdeb: He used literate programming methods with it, formalized the heck out of it, and then proceeded to test the crap out of it. 17:55:39 gilbertdeb: PURELY because of the way he wrote it. 17:55:58 ramnull: yeah, and he still hasn't written the freaking books!!! 17:56:15 Spark could be considered a highly anal way of doing literate programming, with built in compiler support. 17:56:20 ianni: I'd say the army is the largest consumer of software ever. 17:56:22 tcn: I can't say whether they run their weapons on NT, but I can tell you this: when the first NT-voyage crashed, their weapons systems were *entirely* off-line. 17:56:32 The WHOLE SHIP was dead in the water, and needed to be towed and dry-docked for repairs. 17:56:37 heh.. Knuth may find that parsing isn't all that complicated if you do it the Small-C way 17:56:50 tcn: whats the small c way? 17:57:04 kc5tja: DOWN with ms and their marketing. 17:57:14 Now we have Bell Labs copying Ada with thier Cyclone compiler. 17:57:19 MS is a marketing company for really unstable programs. 17:57:19 kc: you can bet they gave Bill Gates a piece of their mind!! 17:57:44 tcn: and they still continue to use it? 17:57:52 tcn: I doubt it. Gates probably would just shrug them off as collateral damage anyway. 17:57:54 I think mr gates knows who the fools are. 17:57:58 gilbertdeb: Yes. 17:57:59 Now if only people would stick to the Forth KISS principles, software would be alot less hectic. 17:58:01 --- join: TheBlueWizard (TheBlueWiz@pc16dn1d.ppp.fcc.net) joined #forth 17:58:02 MS is a conglomerate of software groups.. some good, some terrible. 17:58:05 hiya all 17:58:11 --- join: tgunr (~davec@A17-205-45-104.apple.com) joined #forth 17:58:15 tcn: The WIndows group is horrid. 17:58:24 tcn: which are the good ones? 17:58:27 hi tbw 17:58:43 the marketing department? 17:58:44 hiya gilbertdeb 17:58:57 I wonder if PTSC makes evaluation kits for thier processor. 17:59:17 Access, *some* of the various OS groups, marketing of course.. hehe 17:59:29 gilbertdeb: Microsoft Visual Development Studio group is fantastic. 17:59:46 V.Studio, they suck, I liked Access 97 better 17:59:51 uhh 17:59:58 VS has *zero* to do with Access. 18:00:01 Any of you guys seen the MS critiques at radsoft.net? 18:00:02 VS rocks. 18:00:18 Radsoft is cruel, and they write for the platform. 18:00:21 It's solid, has never, ever crashed on me (quite unlike Access), ... 18:00:28 Access 2000 uses the VS IDE. it's kinda awkward.. although Access 2k itself is improved.. 18:00:40 it seems I can't view clog's log at all 18:00:57 ramnull: Nope. Never heard of them. 18:01:24 They wrote a Windows Explorer replacement thats less than 10k with all the features. 18:01:47 kc5tja: www.radsoft.net They know what thier doing. 18:02:18 btw anyone ever used the access/msword email merge thing succesfully? 18:02:30 hmm.. suppose the DoD is really using Windows a lot.. they could demand to review the source, get things fixed, and pass that along to the everyday users.. 18:02:40 tcn: Nope. 18:03:02 tcn: At least not at the time they got the first destroyer out. 18:03:10 one would think that given its suite characterisitics, the whole damned thing would be pretty easy to do with outlook, with a access-report/outlook combo. 18:03:17 tcn: Maybe with the new "shared source" thing that Microsoft introduced as a result of the anti-trust lawsuit. 18:03:25 computer security HAS been identified as a national security concern.. 18:03:37 kc5tja: anti-trust blah blah. 18:03:38 the "Shared Source" thing only lets (selected) people take a peek at the code, but can't contribute or fix or whatever 18:03:57 in 5 years, they'll tell the world they invented share source as a great solution. watch and see. 18:04:14 TheBlueWizard: Did I say they needed to contribute or fix code? No. All they need is to audit the code. Shared source lets them do that. If they were so inclined. 18:04:15 * TheBlueWizard gleefully pee on "Shared Source" crap 18:04:26 * kc5tja vomits on it. 18:04:40 We'll give MS some stiff competition :) keep 'em honest 18:04:45 tcn: doesn't mean anything. Before long, we'll be the ONLY government on the planet using Windows extensively. 18:05:20 i worry more about banks using it 18:05:23 auditing...yeah...but it is only really good when one have a complete set of code....auditing just a small piece of code isn't enough at all 18:05:30 kc5tja: and when that time comes, I'll go to another continent and sell *nix as _the_ solution to be had. 18:05:58 tcn: they need to be kept honest urgently. 18:06:06 That's the real reason our government did NOTHING to Microsoft; they knew that if they did, the whole f***ing government would collapse; their computers' operating systems would be thoroughly unsupported, if not by necessity, then by retributitive actions on the part of Microsoft. 18:06:15 * TheBlueWizard notes the irony....he is working for fed gov't, and yes, his place is rock solid Win shop :-P 18:06:35 kc5tja: they've bought the govt. yep. there is little hope left. 18:06:45 gilbertdeb: Little, bull. NO hope left. 18:06:56 bbl. Gotta code. 18:06:58 --- quit: ramnull ("This isn't Happy Hour!") 18:07:07 The most we can do is repeat history, with home-built, small computers, made by hand, with self-written OSes on them. 18:07:12 the govt is the single largest, next to the army, consumer of software . 18:07:13 kc5tja: I believe that with the change in administration (Clinton -> Bush), the gov't decided to basically drop the antitrust stuff 18:07:16 well when I'm president, i'll cut back on computers.. :) i'm a semi-luddite 18:07:17 THis is how Apple got its start, after all. 18:07:19 kc5tja: YES! 18:07:40 kc5tja: that method is seeing a revival and it makes me warm and fuzzy inside. 18:07:57 tcn: you might be interested in my PP solutions. 18:08:02 and i'll lay off guys who don't do anything, like tbw ;) 18:08:03 And the nice thing about Forth is that "object code" is meaningless to it. Everything is distributed in source, or at least compressed source form. 18:08:15 Which makes a large class of portability problems utterly obsolete. 18:08:21 tcn: then find me a cool job ;) 18:08:39 it'll be at least 10 years till i'm prez ;) 18:08:48 tcn wanna hear a PP solutions pitch now? 18:08:54 I gotta get started see? 18:09:02 you got a web page? 18:09:09 I can tell you now. 18:09:13 gilbertdeb: Check out http://www.6502.org 18:09:13 PP == pencil/paper 18:09:18 hahahahah 18:09:27 :O don't laugh!!! 18:09:30 gilbertdeb: Lots of homebrew 6502 and 65816-based homebrew projects on it, plus data sheets and all. 18:09:33 it is my new paperful office soltion. 18:09:36 i'm not. i use PP all the time! 18:09:43 that way, we'd pay attention to managing paper properly. 18:09:50 i even use it for programming! 18:10:06 tcn: in that case, you might be interested in some tools. 18:10:12 I have a compass, and a straightedge :D 18:10:24 they are stream lined essential tools. 18:10:33 we have the gc system implemented as well :D 18:10:39 eraser with brush at the other end. 18:10:49 but those are useless without a 1.5GHz mechanical pencil overclocked and fluid cooled at 4GHz. 18:11:03 hahaha 18:11:11 And running Windows CE Drawing Edition 18:11:12 :) 18:11:25 you can get drafting equipment dirt cheap from engineering firms now 18:11:42 if it hasn't been melted down for scrap 18:12:17 Heh 18:12:25 * kc5tja needs a good quality set of triangles and a T-square 18:12:35 i'd take that over CAD any day.. well, not the diazo machines or Leroy lettering.. :) 18:12:52 kc5tja: are firm of qualified, professional sales folks will sell you a t-square :) 18:13:00 Audel's handbook tells you how to build a drafting table & equipement :) 18:13:13 our line solutions department is constantly researching new technologies for drawing straight lines. 18:13:36 Well, I can make all the stuff myself too. 18:13:39 we have perfected antialiasing! 18:13:42 I do a lot of stuff with papier mache now-a-days. 18:13:57 tcn: exactly. 18:14:10 our interface is extremely intuitive and can be used by a 2 year old. 18:14:18 I bet Xerox would love us 18:14:45 and Int'l Paper.. they're hurting.. 18:14:46 AND you can do 90 % of any programming task using PP technology :D 18:14:46 (and then sold for $2Million under the name Van Gogh) 18:14:46 if a 2 year old has a steady enough hand.... 18:15:18 tcn: they don't have a paperful office spiel. 18:15:23 they'd make a killing otherwise. 18:15:27 yeah 18:15:45 of course they'd actually sell less paper.. 18:15:49 Here's my question: does Intl Paper use their own products in their office? :D 18:16:08 do farmers eat their own animals? 18:16:26 Some do, and some don't. 18:16:27 does lockheed martin test its own missiles onitself? 18:16:36 it oughta! 18:16:44 ouch 18:16:48 Lockheed Martin is a cool place, man. Don't knock them! 18:16:53 * tcn nods 18:17:07 They've got the aerospike engine and all -- we need that for our next generation shuttle replacement. 18:17:16 Heh, whenever that'll get the funding again...if... 18:17:22 kc5tja: they make WMD :( and that saddens me a lot. 18:17:32 kc5tja: then again, my i2 came through ebay from them. 18:17:33 And WMD is?? 18:17:40 ya know, LM could use some better software support.. 18:17:48 weapons of mass destruction? those things we claim saddam has and have no proof of? 18:18:12 gilbertdeb: Actually we did find some proof. But I wouldn't have gone to war over it. It wasn't a lot. 18:18:14 it really saddens me that thats what they do for their bread and butter. 18:18:26 They do a lot more than WMD though. 18:18:31 ya gotta die sometime 18:18:57 kc5tja: mostly for the army and nasa. 18:19:06 cancer, heart attack, alzheimers, aids, bomb.. what's the difference? 18:19:16 Yes, but not all that they do is WMD. 18:19:16 kc5tja: from a research point of view, they do cool engineering stuff. 18:19:45 from a world-peace point of view, they should not be under govt control. 18:19:45 There are actually a LOT of other companies out there that makes WMD (one just behind my dojo!), that aren't in the business of aeronautics in particular. 18:19:51 pinpoint destruction is the name of the game 18:19:55 gilbertdeb: No, they SHOULD!! be under government control! 18:20:06 I do *NOT* want those kinds of weapons available to any joe schmoe on the street corner! 18:20:20 you mean pesticides? heh 18:20:32 pesticides are nerve agents.. 18:20:56 kc5tja: the gov't is not very different from a child with heavy heavable stones. 18:20:58 I digress. 18:21:02 Response: chlorine bleach, and dish soap. 'Nuff said. :) 18:21:07 --- join: thin (~thin@sc232-12.cariboo.bc.ca) joined #forth 18:21:16 is it me or is tunes.org down? 18:21:50 * kc5tja checks 18:21:58 their pinging but http's down 18:22:15 "The connection was refused when trying to contact http://www.tunes.org" 18:22:21 that was what I was asking earlier re: tunes.org....I have the same problem 18:22:26 they got rooted a couple weeks ago 18:22:31 TheBlueWizard: Sorry, I didn't see that 18:22:38 ok 18:22:45 gtg again :P 18:22:46 --- quit: thin (Client Quit) 18:22:50 and they did some major upgrading 18:22:55 hmm 18:23:35 hhmm, www2.tunes.org is also down 18:23:37 Right now, my biggest problem with making my own computer is how to generate video, and what to support. 18:23:57 oh.. ww2 used to be a different host but it's not now 18:23:58 I would like it to be at least 640x480 at VGA frequencies, but that does suck up a lot of bus bandwidth. 18:24:15 kc: why do you think I haven't bothered? ;) 18:24:37 Maybe I can go with an Atari ST-like graphics engine, where it's monochrome at 640x480, 4-color at 640x240, and 16-color at 320x240... 18:24:51 tcn: I won't let such small details stand in my way. I *will* find a viable solution... :D 18:24:52 kc5tja: I love monochrome displays. 18:25:01 amber/black or green/black 18:25:03 kc5tja: for 6502 @ 1 MHz, that can be a problem....I think it is doable for 65816 @ 20 MHz 18:25:05 iick. 18:25:14 TheBlueWizard: I'm using 65816 @ 10Mhz. 18:25:15 MHz even 18:25:28 20MHz is beyond the official specs of the chips. Not all 65816s run reliably at 20MHz. 18:25:36 hey, do a google for "e-paper" 18:25:37 hmm...then again I am no HW guy 18:25:57 tcn: Besides, I can always use a bus wider than 8-bits. 18:26:07 hmm...I wasn't aware of the unreliability of 20 MHz...hmm! 18:26:12 gilbertdeb: My display would be bitplaned if I do create it. Bitplanes rock. 18:26:31 Amiga style, eh? 18:26:33 TheBlueWizard: CMD's products works because they test each CPU they get first. :D 18:26:39 Yes. 18:26:45 ah to both :) 18:26:46 i've got a microcontroller book that discusses CRT controllers a little.. 18:26:54 it can be done with TTL chips 18:27:03 I *MAY* even go so far as to support a hold and modify mode too, but don't tell anyone. Amiga, Inc. still holds the rights to that technology. :) 18:27:31 aren't they defunct? 18:27:35 tcn: One solution is to use a dedicated processor, with various address lines controlling the various CRT lines. 18:27:41 --- join: thin (~thin@sc232-12.cariboo.bc.ca) joined #forth 18:27:42 http://www.amiga.com 18:28:17 bah, i'm supposed to be meeting some guy for a game of pool but the temporal-spatial flux is getting in the way 18:28:21 there's been some amazing developments in display technology.. e-paper, even direct-brain.. 18:28:50 tcn: Yeah, like that's going to appear in my home-brew project. 18:28:50 kc5tja: you might come up with an alternative version og HAM....I myself would divvy up 6 bit into 3 signed numbers (-1, 0, +1, ??)...downside: requires ALU units :( 18:29:00 what i don't get is why there isn't any EEG recognition going on 18:29:08 that's way better than voice recog 18:29:15 TheBlueWizard: ??? 18:29:15 and just as easy 18:29:23 people think that EEG is more complex *scoff* 18:29:26 thin: I can't see how. I can't control my EEG waveforms. 18:29:32 yes you can 18:30:01 if you see your EEG on screen, you can think specific thoughts and control the way it looks 18:30:02 tcn: No, Amiga, Inc. is still around...just that Commodore died, then an incredibly convoluted story involving Amiga followed, then ended up as Amiga, Inc. 18:30:29 you can train the computer to recognize it when you are thinknig "A" "B" etc 18:30:41 there's many different brainwaves 18:30:41 I doubt it'd be useful for much. 18:30:46 I'm highly skeptical of its practicality. 18:31:18 what if somethings surprises you, disrupting your thought processes and wreaking havoc on the computer.. 18:31:23 they've shown you can play arcade games with EEG 18:31:35 fly planes too 18:31:36 tcn: does hitting the keyboard randomly wreak havoc ? 18:31:39 thin: what do you do with your hands? 18:31:43 kc5tja: e.g. modify RGB color up and down...ex: RGB=0x8090A0...a (+1, 0, -1) would create 0x81909F 18:31:53 gilbertdeb: i'm sure you could think of something 18:31:54 I can think of virtualsomething. 18:32:10 hmmm control pixels with the mind, hands free. 18:32:13 hmmm. 18:32:19 aren't there 4-bit ALU chips available? 18:32:23 thin: Playing arcade games requires the equivalent of manipulating a joystick: up, down, left, right, and fire (maybe two). That's a lot simpler than emulating a 64-key keyboard, with precision pointer manipulation equivalent to a mouse. 18:32:26 i'm not talking about that complex 18:32:41 (talking to gilbertdeb) 18:33:01 tcn: Yes, but not fast enough to run at video frequencies. 18:33:03 tcn: I used to see 4-bit ALU chips, but that was a long time ago....I think they're still around, but not so hot.... 18:33:10 tcn: Even so, a CPLD can be programmed to do the logic. 18:33:16 kc5tja: a pointer is much more complex, and i wouldn't worry about that, but i think that a 64-key keyboard is achievable 18:33:19 or more 18:33:20 they have a mouse for paralyzed people.. you put a dot on your forehead (like a Hindu) and a camera tracks it.. 18:33:34 it's just a matter of thinking a specific thought and telling the computer what it is 18:33:44 blink to click :) 18:33:50 * kc5tja tried that EEG stuff in a science demo one day. It didn't work. 18:33:52 have they sold this to hawkinds yet? 18:33:55 hawkings.. 18:33:59 he needs such a thing . 18:34:01 definitely 18:34:04 tcn: Makes click-and-drag a pain though... 18:34:13 Hawking (singular0 18:34:17 gilbertdeb: he has trouble moving his head 18:34:24 poor fella. 18:35:06 kc5tja: precisely. Hawking singular. 18:35:06 there is none other like him. 18:35:06 Nope, I prefer good old fashioned, tried and true, known good, provably correct technology -- keyboards and mice. And with a humane interface, the need for a mouse is reduced to manipulating pure graphics. 18:35:35 no pun intended 18:35:36 yep 18:35:37 Hawking, Feynmann, and Einstein are my heros. 18:35:37 s/heros/heroes/ 18:35:37 feynman rules! 18:35:37 EEG keyboard shouldn't be any different than a regular keyboard 18:35:49 I convinved (almost) a friend to name their baby feynman 18:35:54 everytime you hit a key you are emitting a specific brainwave 18:36:01 just train the EEG comp to recognize that 18:36:07 and throw away the physical keyboard 18:36:16 The thing is, my brainwaves didn't change on the screen the last I tried. 18:36:23 kc5tja: you might like my idea for a computer footrest. 18:36:33 * kc5tja WANTS foot-pedals for his computer. 18:36:39 instead of having a space bar, and a return key, you have a footrest do it for you. 18:36:41 I can use foot-pedals for all sorts of stuff, including LEAPing. 18:36:45 so your Feet do stuff. 18:36:57 bah, use a body-glove! 18:37:01 :P 18:37:07 thin: too $$$ 18:37:10 thin, the point is to be cheap here. 18:37:13 I'm not made of money. 18:37:18 yeah 18:37:27 kc5tja: would a footpedal be that hard to make? 18:37:31 gilbertdeb: Nope. 18:37:34 all it needs are 2 or 4 keys. 18:37:43 I'd like to see space, enter, and escape. 18:37:43 :D) 18:37:48 I was going to hack one up to plug into the parallel port of my laptop when I ported FS/Forth to it, to serve as leap keys. 18:37:54 vi editing sessions at the speed of light. 18:38:01 Well, I may be talking nonsense about EEG, I don't know much about it and I haven't gotten to play with one yet. 18:38:04 what would leap keys do? 18:38:14 gilbertdeb: Are you familiar with VI at all? 18:39:01 kc5tja: I use it all the time. 18:39:09 gilbertdeb: THen you're familiar with / and ? then. 18:39:12 hell, I use ed especially. 18:39:16 kc5tja: sure. 18:39:41 gilbertdeb: Instead of having to hit ESC / to enter search "mode," you just press and hold a key called LEAP. There are two LEAP keys -- one for forward, and one for backward. 18:39:52 ah i see. 18:40:01 As you type (while still holding down the LEAP key), the environment does an incremental search for the text you've already entered. 18:40:12 When you release the LEAP key, it goes back to insert "mode." 18:40:25 To search again, you just tap both LEAP keys simultaneously. 18:40:43 Depending on which LEAP goes down first, it'll search in a particular direction. 18:40:59 I could do that in SVGAlib :) 18:41:16 I am working on some editor/viewers.. 18:41:27 what's tunes.org ip? 18:41:29 kc5tja: my foot pedal will do 'space', 'enter', 'esc' or 'esc shift:' 18:41:36 tunes is still down 18:41:43 maybe they were slashdotted :) 18:41:50 kc5tja: space and enter are the most typed keys in any editing session. 18:41:58 Mine would have four pedals: , and COMMAND, with the fourth being user programmable. 18:42:01 hmmm and parenthesis for lispers. 18:42:04 and the slashdotters were disappointed so they DoS'ed it :) 18:42:59 I am surprised /, has not implemented bittorrent caching for the sites url'ed on their pages. 18:43:06 tunes.org's IP is 206.63.100.249 18:43:07 surely the web page authors wouldn't mind! 18:43:29 i think they were afraid of that, actually 18:43:43 tcn: use nslookup...it's your friend :) 18:44:02 what'd you want the ip for anyways? 18:44:07 DNS is working 18:44:27 heh.. i wonder if I can log in 18:44:43 but with bittorrent caching, their webservers would not be overwhelmed. 18:44:51 unless they simply want the bragging rights. 18:44:52 yeah 18:45:15 besides, it would only be for a few days... 2/3 days at the most. 18:45:31 doh, nevermind 18:45:47 well, gotta go...bye all 18:46:12 --- part: TheBlueWizard left #forth 18:46:37 hmm.. only their web server is offline.. there are like 10 people logged in 18:47:08 heh 18:47:25 i can't ssh to my account, damn firewall :( 18:47:32 Now see, if they'd been using ATM networking, and a webserver written in Forth, they wouldn't have any of these problems, now would they? ;D 18:47:39 tcn: do you know anything about tunnelling with putty & a proxy server? 18:47:55 nope 18:48:06 i tried everything, tried tons of different proxy servers 18:48:21 kc5tja: ATM? 18:48:29 at the moment? automatic transaction machine? 18:48:45 kc5tja: async transfer m*? 18:49:25 "Actors on the latest Star Wars film watch instant replays of their battles with CG characters. ILM CTO Cliff Plumer attributes this amazing leap to the increase in processing power and a migration from using Silicon Graphics RISC-Unix workstations to Intel-based Dell systems running Linux." 18:49:49 nooooo 18:49:50 hm 18:49:55 i just think it's ridiculous that they improved by going from SGI unix to intel linux 18:50:03 prolly a whole rack of dells :) 18:50:09 what factors were improved? 18:50:14 how did they improve it? 18:50:19 my first reaction is that it's bullshit and that microsoft paid him to say that 18:50:30 why would MS pay them to say that? 18:50:37 Sgi is now doing intel/ms win! 18:50:49 SGI couldn't keep up with Intel 18:51:04 i mean, nobody can ;) 18:51:06 and hopefully intel can't keep up with AMD 18:51:09 surely I know that, they don't sell to the same audience. 18:51:26 thin: Asynchronous Transfer Mode 18:51:31 and profits are the order of the day. 18:52:15 gilbertdeb: Probably clock rate; MIPS doesn't have multi-GHz processors to my knowledge. 18:52:20 What say we do a little planning, you know, for world domination etc.? 18:52:37 kc5tja: but we both know that a ghz cpu idles longer than a slower cpu. 18:53:00 tcn: We need a mascot first, no? Preferably something small and cuddley -- sort of like the Penguin for Linux, and the devil for BSD. 18:53:09 if the other bottle necks were nonexistent, a ghz cpu would make a lot of sense. 18:53:10 gilbertdeb: Idles in what way? 18:53:22 kc5tja: wasted cpu cycles. 18:53:27 gilbertdeb: For raw number crunching, Intel CPUs are the best performers -- their FPUs are technically better than AMD's. 18:53:31 we need gradual OS improvements, open hardware, better language/development environments.. 18:53:36 gilbertdeb: No, they don't. Not wen running in cache. 18:53:52 Which they'll be doing when running ray-tracing and whatnot. 18:54:05 set aside speed for the moment.. we were talking 65816's a minute ago. 18:54:08 are you saying you are using at least 80% of your cpu now ? 18:54:14 MIPS ? ( i know the other mips (million instructions per second) but what's the architecture mips stand for?) 18:54:23 gilbertdeb: No. But I don't have a 2.2GHz machine either. 18:54:24 kc5tja: sure, for graphhics and such... and number crunching faster machines are better. 18:54:32 gilbertdeb: Well, that's what they use them for!! 18:54:56 thin: Microprocessor without Interlocked Pipeline Stages (MIPS). 18:55:07 you can get open-hardware Intel machines, anyways, if you want something fast 18:55:12 tcn: Yes, we were. 65816's are nice and easy to use in a circuit. 18:55:12 kc5tja: i like otters. i think they rule. they get to swim around and they have fun all day long. they're so cute! 18:55:26 thin: I like dolphins, but then you knew this already... :) 18:55:32 kc5tja: i'd pick an otter for mascot :P 18:55:33 heh 18:55:34 yeah 18:55:42 i was gonna stick with penguins for now 18:55:50 tcn: for retro? 18:55:52 starving penguins :) 18:55:58 Oof, no way we are going to get Linux running on a 65816. :D 18:56:05 we're talking about a mascot for an OS 18:56:08 maybe Lunix :) 18:56:10 kc5tja: even written in ASM? 18:56:13 Not without a *LOT* of work, and even then, it'd still not be worth it. Go with an ARM instead. 18:56:21 gilbertdeb: I grew up with assembly language. 18:56:32 I mean linux in asm? 18:56:59 gilbertdeb: Not calling any system APIs, no, but i have done one function in assembly before (number crunching function). 18:57:00 * thin used to say linux as "lin ucks" then someone convinced me it was "lih nux" (it was written that way in some linux book) 18:57:19 lienoose 18:57:22 lienooxe 18:57:25 The latter is more correct. 18:57:31 like Linus as in Snoopy.. heh 18:57:37 he was Finnish 18:57:37 because it comes from the name linus 18:57:45 Linus pronounces it "Lee Noox". 18:58:05 got us all :) 18:58:18 He pronounces it the Swedish way, with a finnish accent! 18:58:23 so the spanish linux hackers had it right all along 18:58:23 Hehe :D 18:58:25 harharhar Robert. 18:58:33 I pronounce it however I like and if anybody is lame enough to try and correct me I tell them it's my vermont accent :) 18:58:34 And Italians -- they pronounce it the same way. 18:58:58 I have heard it pronouned leeneex and that sounds really off to me. 18:59:03 i'm nominally italian 18:59:26 yes. novello is isn't it? 18:59:30 * kc5tja is half Italian, half German by blood. 18:59:50 italian is a language in which all words end in a vowel no? 18:59:51 novelli actually 19:00:01 * thin is boring: half english, quarter irish, quarter french 19:00:03 1/4 German, 3/4 Swedish. :) 19:00:25 * gilbertdeb is from outterspace and isn't telling which planet. 19:00:25 heh.. in Endicott, NY there were Novello's, Novelli's, DeNovellis's 19:00:37 gilbertdeb: Neptune. 19:00:37 * thin doesn't respect english, except cute english girls ;) 19:00:46 I don't want those NASA knuckle draggers all over my planet. 19:00:53 thin: I concur. 19:01:32 gilbertdeb: I'd say that 90% of the words in Italian end in a vowel, yes. 19:01:37 Those that don't are usually imported from other languages. 19:02:02 dave barry once said that irish people crave potatoes more than others because their blood remembers the potato famine :) 19:02:32 bah. their cooks know no other ingredient! 19:03:42 i guess i really am half irish :) 19:05:17 tcn: are half your dishes potato-based? 19:06:10 well, if we keep some contact between Isforth, Retro, hardware hackers, etc.. it could lead to something pretty good.. 19:08:12 get that assembler done, Small C, trim down linux, switch over to a new kernel.. 19:09:25 i don't mention Isforth because it's pretty well done 19:09:54 or we could just wait 50 years and use tunes :) 19:10:37 i got involved in tunes again to see if i was still interested.. i'm not 19:10:41 tcn: you really think so? 19:10:51 re isforth? 19:10:58 isforth kicks ass! that's coming from a C lover ;) 19:11:14 on what level? 19:11:39 i mean, it's a nice lean x86 forth 19:11:56 needs more tools, that's all 19:12:38 hehe. I was reading the SF book and doing as I was told, and I typed '>r 20 r>' in isforth 19:12:43 and it died. 19:13:18 we learned a lot working on isforth that'll help with other *simple* compiler/interpreters in linux/bsd 19:13:36 we? 19:13:52 i want to get rid of GCC, X, bloated libs, all that crap.. so eventually we're not tied to it 19:13:53 I thought i440r was flying solo on isforth. 19:14:08 i helped him on it last year 19:14:59 i revived it for him, actually :) got it debugged and actually *running* and that got him interested again just as I was moving on to other stuff.. 19:15:18 worked out great 19:16:09 oh.. you can't do >r and r> in the interpreter! 19:16:26 yeah well there was a point where i440r was -> <- away from deleting all the source code 19:16:35 heh.. when? 19:16:49 btw, i might time out any minute now.. 19:17:07 he deregistered the channel 19:17:10 and i registered it 19:17:27 1 year and 11 weeks 6 days ago 19:17:37 he was going to delete isforth around then 19:17:46 but he went to #debian and asked some questions 19:17:51 that's when I moved in :) 19:18:01 because #asm and #assembler weren't helpful 19:18:03 hm, i've got 20 minutes left 19:18:19 he was concerned about loading isforth into memory 19:18:21 or something like that 19:18:21 heh.. those channels are full of l33t lamers :) 19:18:34 and thought it would be impossible 19:19:32 i was using FreeBSD at the time, too, and I had to get Isforth to work with that before I could get started on the debugging 19:19:33 tcn: 20 minutes? 19:19:38 you thought loading it in memory would be impossible? 19:19:56 i bought 5 hrs/day from the ISP 19:20:10 i didn't 19:20:17 Sweet. The industry at large is converging on ATM for its backbone infrastructure. About damned time. 19:20:19 you can buy hours from an ISP? 19:20:21 herkamire: i can't remember what the actual issue was 19:20:27 herkamire: something about ELF i think 19:21:04 kc5tja: what processor? 19:21:12 processor(s) 19:21:16 I had help from Fare's Assembler-HOWTO, linuxassembly.org, asmutils, etc.. oh yeah, and this TeensyELF thing :) 19:21:19 thin: Huh? 19:21:22 ATM is for processors? or networking? 19:21:36 networking I think. 19:21:41 aync transf mode 19:21:43 thin: ATM is a networking protocol. 19:21:44 backbones 19:21:48 i'm a bit out of it right now heh 19:21:55 I just wish it would supplant Ethernet too. 19:22:00 yeah 19:22:04 i'm so out of it i'm having trouble working on the websites 19:22:04 kc5tja: why? 19:22:04 Not that I dislike Ethernet, but ATM is a superior technology. 19:22:08 --- mode: ChanServ set +o thin 19:22:21 gilbertdeb: Well, network overhead is fixed and predictable: 10% *always*, no matter what you're transmitting. 19:22:25 --- topic: set to 'channel website http://forth.bespin.org | native x86 linux forth http://isforth.clss.net' by thin 19:22:31 better response time, right? and simpler becaues it's point-to-point 19:22:38 --- mode: thin set -o thin 19:22:41 Data is packaged into 53-octet long cells, which makes multiplexing the medium MUCH easier and faster. 19:22:48 tathi (on ppc) just made an elf with gas. the code allocated some memory (with rwx perms) with mmap, then copied the parts of the code he wanted into it. After flushing the caches, this works great. 19:23:05 Quality of service is substantially better than Ethernet. Real-time video and audio intermixed with data becomes a reality. 19:23:10 Even with 100Mbps links. 19:23:23 kc5tja: will it do these things fast? 19:23:30 And as tcn points out, the network structure is simpler: only 3 layers instead of 5 for TCP/IP. 19:23:35 gilbertdeb: Damn fast. 19:23:36 and are they cheap/expensive technologies? 19:23:47 They're expensive because Ethernet's been around for two decades prior. 19:23:49 hmmm. 3 layers huh? 19:23:52 latency is the problem with ethernet 19:23:57 But there is nothing intrinsically expensive about it. 19:24:03 even modems have less latency on a good day 19:24:03 I understand. 19:24:14 just as a sparc is expensive vs a pentium 19:24:19 Exactly. 19:24:48 but why is it not being used more often? 19:25:03 shouldn't this be a law? 19:25:24 that whenever there is a popular technology outthere, there is a superior technology lingering in obscurity. 19:25:36 volume discount :) 19:25:49 thats the other word I was looking for. 19:26:01 gilbertdeb: ATM is used extensively today. 19:26:04 Just not on the LAN. 19:26:10 Well, desktop LAN at least. 19:26:16 ATM is used on campus LANs a lot. 19:26:21 compatiblity issues too.. like you change one, you gotta change 'em all, then it's expensive.. 19:26:21 really? 19:26:24 Yes. 19:26:30 how do they behave? do they use ip's as well? 19:26:35 Every time you pick up your phone and dial a number, you're using ATM networking. :) 19:26:45 hmmm. 19:26:51 not my old campus.. at least, the engineering lab had a coax ethernet link.. 19:26:54 Also, the vast majority of DSL modems use ATM as its networking protocol to the telco. 19:27:03 tcn: At the point of the computers. 19:27:13 But the inter-office or inter-building networks were likely to be ATM. 19:27:18 They were at Clarkson University, at least. 19:27:18 * tcn nods 19:27:35 oh.. a lot of my classmates transferred to clarkson 19:27:42 tcn: why? 19:27:50 from Broome.. you know any of 'em? 19:28:43 No, this was years ago. 19:28:54 I actually left Clarkson, with the potential for a class-action lawsuit against them. 19:29:02 oh.. i thought you were like 28 19:29:06 They royally fucked me over, and I'll never forgive them for it. 19:29:06 I am 19:29:19 I was 17 when I went to Clarkson. 19:29:25 oh i see 19:29:50 i didn't go to Broome till I was 22.. not much taste for schools 19:30:32 colleges tend to fuck people over 19:31:06 tcn: few curriculla are designed for the independent learner. 19:31:21 tell me about it 19:31:41 $20-40,000 a year and they'll deny you a diploma because you forgot some damn bureaucratic bullshit (especially state schools) or you missed a prereq.. 19:31:42 hehe, before going to college I imagined that all I'd have to do is go to a prof, pick an interesting topic, work on it, show him what happened, get guided, go back etc.. 19:31:43 i don't know why i'm going for an engineering degree 19:32:08 are you artistic at all? change your major :) 19:32:22 gilbertdeb: Well, so far, at Mira Costa college, I've had nothing but an exceptional education for extremely low prices. I know I sound like a commercial, but I've been through college twice in my life, and both times I dropped out because I couldn't afford it. Here, I'm getting worked *HARD*, and I'm paying only $600/semester. 19:32:29 go here: http://tril.tunes.org/bm/ and scroll down to "Alternative Education" 19:32:30 I never imagined spending so much time simply sitting through text-book reading lecture sessions :( 19:32:49 there's some colleges that are about independent learning 19:32:52 kc5tja: are they just making you do busy work? 19:32:56 in the usa anyways 19:33:07 I hope they aren't. 19:33:08 even at Broome *community college* we didn't read through textbooks 19:33:23 tcn: what did you do? 19:33:23 in fact we hardly ever opened 'em in some classes ;) 19:33:28 ah. 19:33:34 you mastered the art of passing exams :D 19:33:37 had some very good teachers, that's what ;) 19:33:49 that can help as well. 19:33:53 they'd tell us everything, draw it on the board, etc.. 19:33:58 but whats a good teacher like? 19:34:11 is a good teacher a spoon feeder or a rtfm commander? 19:34:14 i want a fully open university where everyone can get their degree thru them thru online correspondence. and I want the degree to be recognized :P 19:34:21 they're responsive.. when the students aren't getting it they change their tack 19:34:33 a book can't do that 19:34:46 gilbertdeb: a good teacher doesn't waste your time 19:34:54 and books just keep getting thicker, flashier, fluffier.. 19:34:55 tcn: I have boiled that down to poor comprehension skills. 19:35:11 tcn: there's always hundreds of textbooks on any given subject 19:35:11 tcn: notice how many people in college do next to nothing about literature? 19:35:17 at least one book out of all those is good 19:35:21 and how much it is they hate literature and writing courses? 19:35:27 a good book is better than a good teacher imho 19:35:50 heh.. i used to go to the library and read through quarter-inch-thick 100-year old calculus books.. more concise :) 19:35:51 gilbertdeb: I love writing; I hate literature. Go figure. 19:36:02 same 19:36:07 kc5tja: but you have to love reading... 19:36:08 and kid's calc books even ;) 19:36:25 kc5tja: do you read sci-fi books? :) 19:36:25 and the act of writing is one of the most powerful thought machines ever. 19:36:45 thought enhance, not machine. 19:36:50 my calc teacher taught kinda like the kid's books.. kinda whimsical, ya know :) 19:36:59 well, my times up for the day 19:37:08 bye 19:37:09 kc5tja: comprehension and writing go hand in hand imo. 19:37:09 --- quit: tcn ("ircII EPIC4pre2 cLIeNUX. Can you say that?") 19:37:38 gilbertdeb: I can comprehend just fine. 19:37:45 gilbertdeb: But that doesn't mean I love reading literature. 19:37:46 and you can write just fine :) 19:37:53 I find literature to be boring, a waste of my time. 19:38:27 I care not what happens to Hamlet. Though I do enjoy Macbeth. Romeo and Juliet is so cliche. 19:38:27 kc5tja: as do a lot of college going people and others. 19:38:34 but those people are not good writers usually! 19:38:48 re: hating literature. 19:39:01 My idea of good literature: J.R.R. Tolkein. Now that's an author! 19:39:13 heh, he's become cliche to me :P 19:39:26 thin: hollywood'll do that. 19:39:28 but he isn't. He *invented* the fantasy genre. 19:39:36 i really don't like how everyone spouts about how all fantasy books derived from him 19:39:37 no he didn't. 19:39:38 yeah 19:39:44 i don't like hearing that :P 19:39:52 gilbertdeb: As we know it today, he certainly did. 19:40:04 i remember hearing about some other author that wrote about elves, goblins, etc before j.r.r tolkien did 19:40:06 D&D simply wouldn't exist without him. 19:40:07 fantasy goes back to the oral traditions of various religions!!! 19:40:13 heheh 19:40:15 re: greek gods and their antics. 19:40:17 gilbertdeb: Of course. 19:40:24 and those were written down. 19:40:38 But that's more mythology than fantasy. 19:40:49 now we call it mythology. 19:40:57 They did back then too. 19:41:12 to me they are all the same. 19:41:14 "Myths and Legends" was a common idiom back then. 19:41:17 * kc5tja shrugs 19:42:52 kc5tja: from my experience, reading/writing are paramount to comprehension. 19:43:17 I'm saying that for those who simply don't get it, and who simply want to be told what will be on the exam, they tend not to appreciate these things. 19:43:35 why are you beating the dead dog? 19:43:36 * kc5tja nods 19:43:39 or horse 19:43:46 HORSE!!! It's a dead HORSE!!! 19:43:54 See, if you read literature, you'd KNOW this... :D 19:44:07 no, i like to say it dead dog 19:44:09 thin: which dead dorse? 19:44:11 everyone else says dead horse 19:44:14 i say dead dog 19:44:29 its either a dorse or a hog 19:44:36 much more fun to beat the dead dog >:D 19:44:52 didnt expect to see this conversation in #forth... 19:44:53 :D 19:45:15 ianni: people have secret fantasies which leak out every now and then. 19:45:55 i just leak the collective unconcious >:D 19:46:06 ianni: eg. you are secretly in love with forth, and you end up dreaming about forth code. 19:46:10 haha 19:46:24 no, it's no secret i am in love with forth 19:46:31 and i didnt see any code in the dream 19:46:39 just empty cells ... it looked so amazing 19:46:40 what did you see? 19:46:43 hehehe 19:46:49 draw a picture. 19:46:54 like as if i could pluck on one and it would appear 19:47:00 i could hardly do it justice... 19:47:24 but its simple i could maybe 19:47:25 sounds like ZIP (from The Humane Interface book) or something more advanced 19:47:56 it was just a horizontal row of cells. very small. centered in the screen.. maybe 10% of the screen in the middle was the height/width of the cell 19:48:18 and as the row of cells got towards the left and right edges of the screen, they expanded depending on the x-pixel coordinate 19:48:19 and you could zoom into the cells? 19:48:26 hm 19:48:33 yeah, you probably could, there would be not other explanation for such a ui 19:48:38 not=no 19:49:28 and it was just scrolling through them, moving them along from left to right 19:49:51 bending the edges of the display jsut for effect.. and a little red highlight on the top and bottom the cells.. it was pretty 19:50:09 just remember that split second.. im not smart enough to dream code yet :) 19:51:00 but yeah, cant wait for the jef raskin book to arrive 19:54:26 Wow. 19:54:55 After all this time, the revenue per bit of telecommunications services is amazingly low for IP; only 5% the total market share. 19:55:08 ATM comes in next at over 40%, and the rest pretty much being frame relay. 19:55:25 so atm is way more popular than you thought? 19:56:04 Not necessarily popular, but it does indicate that more ATM channels are being sold than IP channels 19:56:07 (pure IP that is) 19:56:28 Which vindicates my position back when I was at Armored Internet: Frame Relay and ATM are where it's at, and we just layer IP over that. 19:56:38 I mean, it makes sense; this is what IP was *designed* for after all. 19:56:50 to be embedded? 20:00:25 To be layered, yes. 20:02:01 is it possible to extract the difference of the contents of two files w/o writing a script? 20:02:23 given two sets (text files). 20:04:00 Depends on your host OS. 20:04:07 :D 20:04:07 Unix has "diff" as standard equipment. 20:04:10 assuming linux. 20:04:14 it was something as simple as this 20:04:15 http://www.inpuj.net/~ian/future.png 20:04:15 I just want the differences. 20:04:28 there is nothing more to show, just that it was much more aesthetically pleasing :P 20:04:44 where are the preddy colors? 20:05:01 they looked pretty in the dream 20:05:05 it was all faded and stuff 20:05:11 very smooth 20:05:15 i drew this by hand 20:05:16 hehe 20:05:52 sdiff rgb.txt /etc/X11/rgb.txt | grep '<' | sed 's/ different 20:05:55 that did it :D 20:06:23 I f_cking love textutils 20:06:42 gotta love sed, grep, and awk. 20:07:10 I love sed/grep by default because I think ed is the best of the best. 20:07:17 and by extension, I love vi and vim even more. 20:07:55 that was ridiculously easy. 20:07:59 * kc5tja nods 20:08:01 I love VIM 20:09:55 kc5tja: from a UI pov what do you think of it? 20:10:00 actually, i dont know why i dont love vim 20:10:02 do you feel it is strongly lacking? 20:10:08 ianni: what do you use? 20:10:11 i like the simplicity to some extent 20:10:12 emacs 20:10:18 * ianni ducks 20:10:19 hrmm. 20:10:31 if lisp weren't such a cool thing ... 20:10:42 i just have found no reason not to use emacs 20:10:57 gilbertdeb: The problem with VI is its modality. It's hard for people to think in terms of modes when they have so many keys on the keyboard. 20:11:00 i never wasa big vi head, so i was easier to adapt to a more human like system 20:11:05 kc5tja: exactly! 20:11:09 incidentally, emacs was begotten by teco, which rumor has it that it has a forthlike language. 20:11:13 the fact that vi is stateful limits you 20:11:15 gilbertdeb: On a computer like the Commodore 64 or Apple II, it makes sense, as there aren't nearly as many keys available. 20:11:43 ianni: It doesn't in anyway limit... 20:11:46 try making a repetitive edit, like parsing text like you would with sed by hand with a keyboard.. it's nearly impossible in bi... 20:11:47 vi 20:12:04 ianni: you write what you would in sed in the ':' prompt. 20:12:05 Actually, not true. 20:12:11 unless you start thinking in regexps.. at which point i'd like to have the option of either modeless editing or regeps 20:12:12 i know that 20:12:15 VI has plenty of features for making ranged modifications. 20:12:21 the fact is i *want* to edit it by hand 20:12:25 the sed command are the same as the vi ':' commands which are the same as the ed commands. 20:12:25 like i would in any text editor 20:12:39 arrow goes up, cursor goes up, 20:12:42 And the syntax isn't that different from SED. For example, "%s/foo/bar/gi" is used in VI's command mode to change all occurances, regardless of case, of foo to bar. 20:12:55 kc5tja no im quite aware 20:13:01 ed begat grep/sed/vi ianni. 20:13:02 i've use vi enough to know that 20:13:14 * gilbertdeb is promoting ed heavily :D 20:13:19 :%s/foo/bar/g 20:13:20 we noticed.. :D 20:13:27 you use ed?? 20:13:31 all the time. 20:13:35 haha nice 20:13:37 its either that or vi/vim. 20:13:39 i should 20:13:55 small, lovely, easily mastered... powerful. 20:14:01 i just havent found a reason not to use emacs.. 20:14:08 :D 20:14:12 im a lazy person.. and i have big powerful machines :) 20:14:29 I don't use emacs because it's slow on my machine, and rather large. 20:14:30 teco was written in forth by john draper aka captain crunch (i think??) 20:14:37 Otherwise, I think it's a pretty nice editor. 20:14:43 that guy does forth? 20:14:50 captain crunch wrote _a_ editor in forth 20:14:54 And since getting used to VIM, I'm actually significantly faster in VIM than in Emacs now. 20:14:56 not anymore 20:14:59 i talked to him on icq 20:15:02 kc5tja, thats the only argument i have against it.. and the only reason why it's not the only editor ill use 20:15:08 rotten has bad things to say about him 20:15:09 heh nice 20:15:11 thin: thats impossible! 20:15:17 gilbertdeb: nope 20:15:26 his website lists is icq number 20:15:31 url? 20:15:49 teco as in tape editing/corrector ? 20:15:53 http://www.rotten.com/library/bio/hackers/captain-crunch/ 20:16:34 http://www.webcrunchers.com/crunch/home.html 20:16:51 my link is HIS own website 20:16:53 what are some vi-like text editors? 20:17:15 theres got to be some forth editor 20:17:35 --- join: natty (~n1ywb@155.42.84.139) joined #forth 20:17:45 ianni: there's lots of forth editors 20:17:49 in forth :P 20:18:08 --- quit: natty_ (Read error: 54 (Connection reset by peer)) 20:18:11 in the forth implementations 20:18:19 im sure - heh this guys site has some awesome stuff :) 20:18:48 So he thought he would try and use his blue box to place a call. Just then, a police car pulls up and a policeman gets out and asks Woz for his ID. As the cop was checking things out, he noticed the box sitting in the pay phone and asked "What's that?". Woz said "It's my Synthesizer for an electronic project". The cop was pressing the buttons, playing with it, then handed it hack and said "A guy named Moog beat you to it" and left. 20:18:54 ahahahahahahahahah.. 20:18:55 That is funny. 20:19:25 ianni: I once wrote a VI-like block editor for Forth, called VIBE. 20:19:50 It's woefully underfeatured compared to a real editor though. :) 20:20:20 hehe 20:20:24 id just like to see some in action 20:20:47 kc5tja: as long as it can do what EDIT.COM can do, I won't consider it underfeatured 20:20:49 I'll try and find the sources again. I used to have them on my box here, but now they're missing for some bizarre reason. :( 20:20:54 kc5tja: did it have ed powers? 20:21:03 gilbertdeb: No, but they were relatively easy to add. 20:21:07 thin what is edit.com ? 20:21:11 is that from vms / 20:21:12 ? 20:21:16 gilbertdeb: DOS's editor. 20:21:24 bah that ugly thing! 20:21:25 It came after EDLIN.COM 20:21:30 yeah, see.. I think vi should have a modeless mode. 20:21:32 i didn't like edlin 20:21:37 then i think it would be twice as useful 20:21:42 ianni: lol. modeless mode 20:21:47 that's a mode! 20:21:49 right there :P 20:21:54 ianni: it is derived from ed, which was 'modey' . 20:21:56 that would only make vi kick more ass then :) 20:22:01 heheh 20:22:07 so to understand why vi is also modey, look at its papa 20:22:11 to be able to TURN OFF modes then - depends on how you implement 20:22:37 basically just so i can edit edit.com style - as bad as i feel for using that example haha 20:23:10 ianni: just as CM is not going to make forth use infix for any reason, no one is going to take the beloved beep mode away! 20:23:24 you either think postfix or use something else :D 20:23:46 ha 20:24:28 beep mode? 20:24:43 yes. 20:24:53 there are only 2 modes you know? 20:26:15 in vi? i guess maybe that is true. how can you navigate within the screen while in edit mode then? 20:26:54 hjkls 20:27:10 oops no s. 20:27:28 you don't go back and forth in edit mode :) 20:27:44 reason: ed, vi's papa is a line oriented text editor. 20:27:48 get it? one line at a time. 20:28:05 to see a previous line, or the next line, you give its address see? 20:28:18 I don't even use hjkl in VI for movement while engaged in heavy writing. 20:28:23 i should probably learn a little about ed, seeing as it's the parent of a lot of the stuff i use 20:28:27 ALL my navigation is often done with / and ? because it's just plain faster. 20:28:39 kc5tja: have you used ctrl * in vim? 20:28:41 i do see... and i think thats limiting 20:28:44 the tips on the vim page are fantastic. 20:28:49 kc5tja: wow, you're the ideal LEAP candidate 20:29:02 is ? reverse / ? 20:29:07 ianni yes. 20:29:13 ianni: Yes. It searches backwards by default instead of forwards. 20:29:16 nifty huh? 20:29:20 sure.. 20:29:26 gilbertdeb: Nope. 20:29:39 is too. 20:29:55 why not? 20:30:01 * bwb just got his SAT scores :) 20:30:01 gilbertdeb: I'm replying to your question. 20:30:08 * bwb :) :) 20:30:11 ah that. 20:30:17 bwb: Congrats, I hope... :D 20:30:20 bwb: good job. 20:30:28 anything above 500 is good . 20:30:53 hehe yes it is above :) 20:31:07 kc5tja: sorry, it is star. 20:31:15 it lets you search for the word under the cursor. 20:33:09 * ianni gains sudden interest in colorForth 20:33:19 uh oh. 20:33:24 I'll say nothing. 20:33:36 hahah 20:42:55 its neat to look at. god knows i couldnt use this 20:43:16 * kc5tja could. :) 20:43:48 it'd be fun to try though.. you know your OS is good if it runs on a floppy... 20:44:04 Yup. 20:44:13 But ColorForth isn't the only Forth that will run from a floppy. 20:44:17 FS/Forth soon will too. 20:44:33 I need to implement a 32-bit assembler for it first, then port over the target compiler. 20:44:45 that idea there is particularly alluring as i just build a x86 box out of spare parts but SCSI card or drives arent working.. could get a flopy drive going and use it :) 20:44:48 Then I need to hack FS/Forth itself so that it doesn't rely on DOS or BIOS calls. 20:44:59 * kc5tja nods 20:45:09 YEEHAA! 20:45:11 be using that asus mobo and 750ish mhz cpu for that tiny os .hahahha 20:45:15 weeee 20:45:16 Good propegation on 40m tonight. 20:45:26 40m? 20:45:34 You got it -- I have full intentions of running FS/Forth on my 800MHz Athlon box. 20:45:40 yeah 20:45:42 40 meter amateur radio band (shortwave). 20:45:42 thats like 20:45:54 a freaking... 10 thz cpu machine or something 20:45:55 relatively 20:45:56 My goal is to utterly and totally replace Linux as my OS of choice. 20:46:05 that's an awesome goal 20:46:26 First it was going to be via Dolphin (an AmigaOS-inspired OS I was writing), but now I think Dolphin will actually be written in Forth. 20:46:40 will you have web browser? 20:46:47 I'll try. 20:46:50 hehehe 20:46:55 well it cant be TOO impossible 20:46:55 kc5tja: ur doing this still in spite of tcn's paper? 20:47:02 I've never written a web browser before, so I can't guarantee any kind of success. :) 20:47:12 gilbertdeb: Hell yes. 20:47:36 yeah, i thought of tcn's paper when you said that 20:47:49 not that i haven theard that idea before, but it is no simple task 20:47:52 --- join: jdamisch (jdamisch@207.191.240.207) joined #forth 20:47:58 hi 20:47:59 if only it were here 20:48:06 kc5tja: how far along are you? 20:48:10 hello, jdamisch 20:48:12 hi jdamisch 20:48:27 Not far enough. The 16-bit DOS hosted version is almost to the point where it can interpret stuff from the keyboard. 20:48:33 Then I need to get it to interpret stuff from storage somehow. 20:48:47 what are we talking about? 20:48:52 And from there, I want to use FS/Forth to write the 32-bit assembler. 20:48:57 kc5tjaforth jdamisch 20:48:59 jdamisch :D 20:49:09 :^) 20:49:19 jdamisch: it is the favorite activity of forth programmers to write their own forths ! 20:49:20 * thin remembers wanting to talk about something with you but forgets 20:49:33 you might already know that though.. 20:49:47 i still haven't written my own Forth 20:49:54 don't listen to gilbertdeb he's all pumped up with the romantic notion of being a "forthwright" 20:49:59 and you're a forth programmer? 20:50:01 or "forthsmith" or whatever other nonsense 20:50:03 thin: not for me. 20:50:05 gilbertdeb: Well, writing your own Forth is a rite of passage. Besides, no existing Forth today can do the things I want it to do, in the way I want them to be done. 20:50:07 for the 'others'. 20:50:16 4thwright, 4thsmith, 4ther :D 20:50:36 yeah 20:50:46 thats my goal, in this arena of computing 20:51:02 gotta write my own 4th 20:51:13 the mantra of forth is this apt haiku: You, Forth, the machine. Simplicity :D 20:51:43 yeah - anything that is developed is almost infinitely more useful since you understand all aspects of it 20:51:55 or at least since it is all part of the system 20:52:10 or it seems like it could be 20:52:50 kc5tja: Do you think forth is a good choice for more high-level tasks? 20:53:19 bwb: the answer is forth is. whats the question? 20:53:19 :D 20:53:45 heh 20:54:16 yeah, my obstacle is me 20:54:23 Well I mean from a more practicle vs. aethetic pov 20:54:46 lack of libraries. 20:55:00 forth is a multipurpose language like C is, so just use it for whatever you want to 20:55:26 it can be worked into something very high level 20:55:33 jdamisch: do you use/like forth? 20:55:38 have you seen the old Washing Machine example? 20:55:41 forth > * 20:55:43 i have. 20:55:49 i love forth 20:56:01 oh :D 20:56:08 i like that there are programmers that love forth, and can't program in it 20:56:11 that's fairly unique.. 20:56:20 * ianni takes his hand off the *ban forever* button 20:56:24 forth * > 20:56:25 though i guess you might see a little of that with java 20:56:26 that's better :) 20:56:29 heh heh 20:57:11 --- quit: tgunr () 20:57:14 jdamisch: Well... disregarding speed, how about python? (I don't like that example... but in general something that abstracts memory a bit more) 20:57:18 Klaw: forth and the jvm have some interesting similarities actually, as Speuler was pointing out recently 20:57:36 i haven't tried python 20:57:38 the bytecode i guess... 20:57:52 python has some happy developers 20:57:58 the language is a differnet subject 20:58:03 --- nick: _MrAway_ -> MrReach 20:58:33 fun things are done in python all the time. 20:59:21 i'd like to see some sort of synthesis or middle ground of python and perl 20:59:32 I like python the way she is. 20:59:36 for middle ground, see ruby. 20:59:39 maybe if i learned perl, i'd think differently 20:59:40 or tcl. 20:59:50 yeah, but python and perl seem so much more efficient 21:00:13 ... than? 21:00:20 yeah, tcl, i dunno.. i had enough of that 21:00:28 I am trying to find something (or write up specs to make something) where the core language is simple/flexible (like forth) but also abstracts a lot of stuff.. and it would only have one data type (in the language, implementation will prolly use optimization of course), but one that is extremely flexible (list or an associate-array perhaps) 21:00:30 when scripting for irc 21:00:51 like a list in lisp ? 21:01:31 isnt that what everything is in that language 21:01:37 thats the impression i got 21:01:52 bwb: Perhaps not from an aesthetic point of view, but there isn't a thing that Forth cannot do that C can. 21:02:11 And once you really get to know Forth, you develop a unique sense of aesthetics, I think. 21:02:11 kc5tja: eg rewrite Xlibforth ? 21:02:16 I can't imagine that. 21:02:55 kc5tja my sense of aesthetics have changed dramatically, and i still do not really know forth.. 21:03:12 ianni: you gotta write a forth :D 21:03:51 gilbertdeb: I'm unfamiliar with xlibForth. 21:04:01 kc5tja: thats because it does not exist. 21:04:05 However, FS/Forth for Linux will have a raw xlib interface. 21:04:13 i dont even know what some of the most important words are 21:04:22 i just have gotten a feel for the dictionary and the stacks and words 21:04:23 ianni: swap? 21:04:32 no, like DOES> and immediate 21:04:37 stuff like that.. 21:04:44 ianni: Those are pretty advanced words, actually. :) 21:04:45 ianni: Sorta like a list in lisp.. Tho I don't like the parens much :/ 21:04:49 right... 21:04:50 (even with a good editor) 21:04:50 CREATE and DOES> are very powerful. 21:04:55 why i dont understand them :) 21:05:03 the stack operations make sense to me 21:05:27 i just need to do more and then i will understand them 21:05:28 create and does> are powerful 21:05:36 ianni: you don't like TCL? why? it's my favorite 21:05:36 They're used to create new "defining" words. 21:05:44 --- quit: Herkamire ("bedtime") 21:05:55 kc5tja: main thing I don't like about forth is lack of memory management.. great of embedded/low-level.. but... 21:05:58 CREATE actually creates the word in the dictionary, and DOES> tells the environment what that word does when it's executed. 21:06:10 ah i see 21:06:12 Ahh, but memory management is ironically one of its strongest points. 21:06:15 bwb, you can use the heap if you like 21:06:38 I have been doing some research on memory management techniques, and Forth's use of the dictionary does fall under what's called a region-based memory management scheme. 21:06:43 The dictionary is one region. 21:06:44 ianni: if you want to start from the utter beginning, you could search for the b5500 instruction set. 21:06:46 well i dunno 21:06:47 allocate, free, resize 21:06:57 CREATE and DOES> are best explained with illustrations of what a word looks like memory-wise 21:06:57 i didnt ever get good at TCL.. 21:07:01 not my style 21:07:13 i could get it now if i wantd 21:07:15 If you can generalize the concept into supporting multiple regions (but all within the same address space), you can get general purpose dynamic memory allocation, but with the convenience of garbage collection IF you program it right. 21:07:34 yeah, i can imagine they have direct relationship with the hardware 21:08:43 Therefore, every time you execute a MARKER word, or FORGET something, or ANEW, you free all the memory allocated by ALLOT in the previous run. And don't forget that you can also ALLOT a negative amount too. 21:08:51 kc5tja: hmm... got any urls? 21:09:09 bwb: It took me forever to find the last set of documents. :( Hold... 21:09:16 i dont know MARKET or ANEW either... im very new 21:09:20 MARKER :P 21:09:33 bwb: forth urls? 21:09:34 i need to learn these more advanced words... 21:09:52 well but that erases stuff you might not want free'd too.. but I guess that falls under "programmed right" 21:09:57 gilbertdeb: mm 21:10:05 n/m 21:10:16 what is MARKER and ANEW? 21:10:24 --- quit: bwb (sterling.freenode.net irc.freenode.net) 21:10:24 --- quit: sifbot (sterling.freenode.net irc.freenode.net) 21:10:36 I've never been netsplit. 21:10:47 --- join: bwb (~bwb@ip68-4-121-108.oc.oc.cox.net) joined #forth 21:10:50 what is netsplit? 21:10:55 http://www.daimi.au.dk/~beta/Papers/Train/train.html -- here's the "Train" algorithm which uses techniques similar to the region-based memory management I was talking about, but it's not quite the same. I'm still looking. 21:11:02 what just happened to bwb and sifbot. 21:11:27 netsplit 21:11:28 oh heh 21:11:28 got disconnected 21:11:28 not netsplit 21:11:33 oh. 21:11:35 bwb: Yes. And don't forget that in a Forth environment that supports multiple regions, you can have finer grain of control over what does and doesn't get removed. 21:11:36 server disconnected 21:11:44 well, thats netsplit :P 21:11:51 just i guess it disco's you 21:12:14 unless it just disconnected you with same quit messages it usese for netsplits.. 21:12:15 ianni: I thought a netsplit is when the irc network fragments? 21:12:30 yeah, it's when a server on the network 'drops off' 21:12:52 bwb: you and sifbot left simultaneously. 21:13:03 it happens a lot on servers run by linux / unix techies, do to inherent instability 21:13:12 and there had been globalnotice about the possibility so I put 2 and 3 together. 21:13:30 Klaw: what causes the instabiilty? 21:13:34 eh, it's just what happens when a server leaves a multi-server network ... 21:13:40 i thought 21:13:42 gilbertdeb: I took it to be the server crashing or the machine shutting down or something like that.. 21:13:44 what would you rather have it run on? 21:13:55 or, DoS 21:13:57 etc 21:13:59 loss of connectivity 21:14:06 it closed the socket this time 21:14:15 http://citeseer.nj.nec.com/tofte97regionbased.html -- here's the seminal paper. 21:14:23 regardless it split, its not on the network anymore 21:14:46 gilbertdeb: I'm being facecious 21:15:08 haha 21:15:09 however there are a lot of netsplits on freenode 21:15:10 Ooo, he has a paper on the theory of stack allocation in polymorphically typed languages... 21:15:27 voodo kc5tja, voodoo. 21:15:32 Interestingly enough, the stack-based/region-based memory management architecture was vindicated for real-time ML implementations, not for Forth. 21:16:11 Notice the authors of these papers: tofte is primarily responsible for the creation of ML. ;0 21:16:12 :) even 21:16:45 thanks 21:17:02 interesting.. 21:17:57 gilbertdeb: you have been netsplit 21:17:59 And it works pretty much similarly to how CHuck uses his Forth too, when one thinks about it. 21:18:02 when? 21:18:07 I am still here. 21:18:25 no, when a netsplit happens, there's TWO "here"s 21:18:42 the people on the otherside of the netsplit are still on the channel 21:18:43 When he wants to do something, he loads in a program, causing the dictionary to be flushed in the process. The program runs, allocating any necessary memory using ALLOT, spits out the data somehow, then reloads another program (causing the dictionary to be reset [e.g., memory freed] in the process). 21:18:49 and they see YOU get netsplitted 21:19:04 but I don't see it happen then? 21:19:11 right.. think about writing the irc server... 21:19:21 the channel is made up of any noumber of nodes ... 21:19:41 if you get split, you're still on irc, just not connected to any of the others, you're off the hub 21:19:43 gilbertdeb: when bwb & sifbot got netsplitted, YOU also got netsplitted 21:19:44 hmmm. I wonder if postscript is theoretically as easy to write as forth supposedly is. 21:19:52 :O 21:20:01 gilbertdeb: postscript is easy to write 21:20:05 i've coded in it 21:20:08 sorry implement. 21:20:11 the language... 21:20:20 i want to implement a postscript-like language in forth 21:20:26 it'll be highly useful 21:20:37 for graphics and printing.. 21:20:51 thin: we were not still in the channel, the whole host went down 21:21:03 still no postscript app by default on windows 21:21:23 bwb: ok, that doesn't happen too often 21:21:41 thin: yeah 21:21:45 klaw: ??? 21:21:55 to open postscript files 21:21:58 i've noticed that 21:22:03 klaw: notepad :P 21:22:09 hah 21:22:11 ghostview displays it properly tho 21:22:21 not as good quality but oh well 21:22:45 * thin isn't sure why they would have such a hard time displaying it nicely 21:23:07 i'm sure there are other apps like adobe apps like illustrator 21:23:09 that open ps files 21:24:53 i wonder if i can do some forth cgi stuff with the bespin account? 21:25:09 without having to to open a port or mess with apache? 21:25:15 it'd be nice if acrobat reader did 21:25:26 Sure you can. IIRC, there are several (Forth-related) sites that use Forth as the CGI interface. 21:25:40 yes, install ghostscript for windows is a PITA 21:25:41 All you need is access to environment variables and stdin and stdout. 21:25:43 doesn't speulers site have that? 21:25:49 ie forth.cgi ? 21:26:05 * kc5tja nods 21:26:11 yes 21:26:23 ok, I'll get the URL for MARKER 21:26:26 * kc5tja was at one time thinking of writing a Wiki using forth.cgi as a basis for my work. 21:26:33 forth.org speuler's site and paysen's site have cgi stuff 21:26:37 MARKER is an ANS standard word. 21:26:50 When you say MARKER FOO, it creates a new dictionary entry called FOO. This word is a marker only. 21:26:57 oh, hell that one is local to my network 21:27:19 When you execute FOO, however, it resets the dictionary to the state it was in just before MARKER FOO (e.g., it unloads all the code and data that comes after it, including FOO itself) 21:27:40 --- log: started forth/03.04.17 21:27:40 --- join: clog (nef@bespin.org) joined #forth 21:27:40 --- topic: 'channel website http://forth.bespin.org | native x86 linux forth http://isforth.clss.net' 21:27:40 --- topic: set by thin on [Thu Apr 17 19:22:24 2003] 21:27:40 --- names: list (clog onetom skylan Robert TreyB gilbertdeb MrReach kc5tja thin natty bwb @ChanServ) 21:27:43 --- join: ianni (ian@inpuj.net) joined #forth 21:27:48 thats a netsplit 21:27:53 and I'm stil lhere :D 21:27:55 yeah, you were part of it 21:28:05 * gilbertdeb is confused still. 21:28:15 http://www.taygeta.com/forth/dpans6.htm#6.2.1850 21:28:45 It depends on which servers split. 21:28:45 ah I see now. 21:28:45 the dictionary is a PITA when deallocating 21:28:46 For example, gilbertdeb was here for me. However, jdamisch and Klaw are still gone. 21:28:47 I am on the same server . 21:28:57 the chains involved can become QUITE complexs 21:28:59 --- join: jdamisch (jdamisch@207.191.240.207) joined #forth 21:29:06 MrReach: ? 21:29:07 jdamisch: wb 21:29:11 How do you mean? 21:29:15 :^) 21:29:32 for anything that chains in the dictionary, there must be a word called to unchain it 21:29:37 i celebrated Yuri Gugarin Day last Saturday 21:29:47 why him? 21:29:48 i drank vodka and ate caviar 21:29:54 why not tsiolkovsky? 21:30:04 or marker words or ANEW will cause invalid pointers 21:30:09 i don't know, they just have a Yuri day, and i was there 21:30:09 wodka :D 21:30:12 MrReach: I'm not convinced that is necessarily true. But the number of independent chains that exist must be maintained, and I'll admit that is a PITA. 21:30:19 But I've never had to write "destructor" code in Forth ever. 21:30:38 usually because one compiles a program, runs it, then bails out 21:31:09 we used to say Forget instead of FOO 21:31:15 can i use forth cgi even when apache is being used? 21:31:25 yes, FORGET is now deprecated 21:31:33 somesite.org/index.html is handled by apache 21:31:33 thin, yes 21:31:38 It does become a huge issue when you support multiple wordlists and/or vocabularies though, since each vocabulary is a wholly independent chain. 21:31:58 I don't like MARKER personally. 21:31:59 mrreach: how? 21:32:20 I also link .DLL/.so references and library call points in the dictionary 21:32:23 I'm not exactly sure what to replace it with yet, but I do know that MARKER seems clumsy to me. 21:32:43 MARKER is much better from the system designer's point of view 21:33:00 thin ... 21:33:06 Yes, I'll agree, because it gives the proper opportunity to snapshot the system state. 21:33:09 oh nevermind 21:33:16 mrreach: never mind i think i got it 21:33:22 first, apache can call any number of executable progs depending on the URL 21:33:30 yeah 21:33:59 once an executable is running, it need only peruse the evironment for the specifics of the web call 21:34:03 I just don't like it from a usability stand point. I prefer chuck's EMPTY solution more. 21:34:17 and return the document on stdout (including headers) 21:34:27 kc5tja: how does EMPTY work? 21:34:36 yes, EMPTY/FENCE is very similar to MARKER 21:34:50 --- quit: ChanServ (Shutting Down) 21:35:03 bwb: ColorForth maintains a fixed set of variables. FENCE tells the Forth where to "fencepost" the system -- e.g., "Take a picture of the dictionary HERE, and remember it." EMPTY restores the dictionary to that state. 21:35:11 MARKER is nearly identical to FENCE ... 21:35:22 My question, though, is does Chuck's system allow for nested fences? 21:35:24 --- join: Klaw (chuck@ip68-4-77-247.oc.oc.cox.net) joined #forth 21:35:26 I don't think it does. 21:35:29 EMPTY is nearly identical to executing 21:35:35 Yep. 21:36:14 hmmm ... perhaps a 'toolset' should have been specified for unloading the dictionary, like for wordlists 21:36:24 --- quit: natty (leguin.freenode.net irc.freenode.net) 21:36:24 --- quit: Klaw (leguin.freenode.net irc.freenode.net) 21:36:24 --- quit: MrReach (leguin.freenode.net irc.freenode.net) 21:36:24 --- quit: bwb (leguin.freenode.net irc.freenode.net) 21:36:25 --- quit: onetom (leguin.freenode.net irc.freenode.net) 21:36:25 --- quit: skylan (leguin.freenode.net irc.freenode.net) 21:36:25 --- quit: Robert (leguin.freenode.net irc.freenode.net) 21:36:25 --- quit: thin (leguin.freenode.net irc.freenode.net) 21:36:25 --- quit: TreyB (leguin.freenode.net irc.freenode.net) 21:36:25 --- quit: gilbertdeb (leguin.freenode.net irc.freenode.net) 21:36:25 --- quit: kc5tja (leguin.freenode.net irc.freenode.net) 21:36:27 FORGET is easiest for beginners to grasp imo 21:36:39 lol 21:37:28 --- join: Klaw (chuck@ip68-4-77-247.oc.oc.cox.net) joined #forth 21:37:28 --- join: bwb (~bwb@ip68-4-121-108.oc.oc.cox.net) joined #forth 21:37:28 --- join: MrReach (~mrreach@209.181.43.190) joined #forth 21:37:37 --- join: thin (~thin@sc232-12.cariboo.bc.ca) joined #forth 21:37:45 lol 21:38:02 --- quit: thin (Client Quit) 21:38:06 --- quit: MrReach (leguin.freenode.net irc.freenode.net) 21:38:06 --- quit: bwb (leguin.freenode.net irc.freenode.net) 21:38:06 --- join: thin (~thin@sc232-12.cariboo.bc.ca) joined #forth 21:38:14 --- join: ChanServ (ChanServ@services.) joined #forth 21:38:14 --- mode: leguin.freenode.net set +o ChanServ 21:38:24 --- quit: thin (Client Quit) 21:38:39 --- join: natty (~n1ywb@155.42.84.139) joined #forth 21:38:40 --- join: thin (~thin@sc232-12.cariboo.bc.ca) joined #forth 21:38:40 --- join: Robert (~snofs@h138n2fls31o965.telia.com) joined #forth 21:38:40 --- join: skylan (sjh@Sprint3031.tbaytel.net) joined #forth 21:38:40 --- join: onetom (~tom@novtan.bio.u-szeged.hu) joined #forth 21:38:40 --- mode: ChanServ set +o thin 21:38:41 --- quit: thin ("Leaving") 21:38:42 --- join: thin (~thin@sc232-12.cariboo.bc.ca) joined #forth 21:38:48 --- quit: thin (Client Quit) 21:38:49 --- join: thin (~thin@sc232-12.cariboo.bc.ca) joined #forth 21:38:54 --- quit: thin (Client Quit) 21:38:55 --- join: thin (~thin@sc232-12.cariboo.bc.ca) joined #forth 21:39:02 maybe we should all just come back a little latter, this is silly 21:39:19 bah i wonder what server kc5tja and mrreach are on 21:39:32 ?? 21:39:45 --- join: bwb (~bwb@ip68-4-121-108.oc.oc.cox.net) joined #forth 21:39:45 --- join: MrReach (~mrreach@209.181.43.190) joined #forth 21:40:19 --- join: kc5tja (~kc5tja@ip68-8-206-137.sd.sd.cox.net) joined #forth 21:40:19 --- join: gilbertdeb (~gilbert@fl-nked-ubr2-c3a-29.dad.adelphia.net) joined #forth 21:40:19 --- join: TreyB (~trey@cpe-66-87-192-27.tx.sprintbbd.net) joined #forth 21:40:23 jdamisch: there are a whole bunch of servers that are all interconnected 21:40:27 OK, it shouldn't take too long. Just traffic time, really. 21:40:44 i see that now 21:40:47 Klaw: yer in oc? 21:40:49 if a router goes down then two networks could exist 21:40:54 yes 21:40:55 whereabouts? (just curious as to where other forthers are) 21:40:55 brb folks -- going to the supermarket. 21:41:02 nice of you guys to join us. 21:41:06 splitty 21:41:06 yay! my USB device driver compiles! 21:41:08 --- quit: MrReach (leguin.freenode.net irc.freenode.net) 21:41:08 --- quit: bwb (leguin.freenode.net irc.freenode.net) 21:41:08 --- quit: Klaw (leguin.freenode.net irc.freenode.net) 21:41:25 grats 21:41:43 how often does it get this way? 21:41:56 oh good a floodsplit 21:41:57 not often 21:41:57 --- join: MrReach (~mrreach@209.181.43.190) joined #forth 21:41:57 --- join: bwb (~bwb@ip68-4-121-108.oc.oc.cox.net) joined #forth 21:41:57 --- join: Klaw (chuck@ip68-4-77-247.oc.oc.cox.net) joined #forth 21:42:10 oc? 21:42:10 MrReach: I didn't get everything -- netsplit. Can you resend what you were saying regarding EMPTY and MARKER and such? 21:42:10 * kc5tja is in Oceanside, California (San Diego county, just south of Orange County) 21:42:32 oklahoma city. 21:42:51 MARKER is nearly identical to FENCE ... 21:42:55 EMPTY is nearly identical to executing 21:42:58 hmmm ... perhaps a 'toolset' should have been specified for unloading the dictionary, like for wordlists 21:43:11 and then [utting MARKER in the optional wordset 21:43:14 putting 21:43:23 [end] 21:43:34 Thanks. I can go to the store with peace of mind now. :) 21:43:48 Yeah, I've been thinking about MARKER and such, because eventually I'll need it myself for FS/Forth. 21:43:51 * MrReach sticks his tongue out at kc5tja 21:43:52 Anyway, brb. 21:44:02 uhh 21:44:04 why? 21:44:09 * kc5tja is confused? 21:44:18 for resuming a conversation and then bolting out the door 21:44:22 * MrReach laughs 21:44:34 Well, I just ate a can of plain tuna fish, and now I'm thirsty, and water ain't doing it for me. 21:44:35 :) 21:44:42 ack! 21:44:52 Plus all that talk above about Captain Crunch made me hungry...for Cap'n Crunch. :D 21:44:52 * MrReach offers kc5tja some mayo 21:44:59 Iick. No way. 21:45:03 Plain tuna for me. 21:45:08 just eat the stick of butter 21:45:09 I'd rather have raw tuna, but I digress. I'm cheap. 21:45:28 in that case eat the raw package of top raman 21:45:31 ok, breaded tuna plank is good, too 21:45:46 Yes; tuna steak is wonderful. 21:45:53 Sashimi is my favorite form of tuna though. :) 21:45:56 my two tuna recipes. tuna salad: tuna, celery, and a dash of mayo tuna surprise: egg, cheese, tuna, microwave and stir it all up and microwave it lots 21:46:01 Anyway, I'm out. BBIAB. 21:46:07 not me, generally only eat cooked meats 21:46:15 see you soon 21:46:49 parsing CGI in forth is nasty, too 21:47:03 in fact, I can't think of anything that's EASY in forth 21:47:18 elegant and efficient, yes 21:47:22 not easy, though 21:47:29 string stack looks interesting for helping parsing words 21:48:03 just create new stacks for every problem you want to deal with.. floating point? no prob, make a floating point stack. string parsing? no prob, make a string stack :P 21:48:34 if you recall, I wrote a string library that worked similar to MS Basic string heap awhile back, speuler helped me design it 21:48:41 easier than gobbing everything onto the paramater stack and then going nuts with stack gymnastics 21:48:44 unfortunately, I lost it with a HD crash 21:49:01 bogus 21:49:17 make backups, theres no excuse now that we have CD-RW 21:49:28 oddly enough, coming up with an elegant specification was the hardest part 21:49:33 just keep all of your code separate in one folder and backup that one folder once a day 21:49:34 There's a message to be learned in this... *sigh* 21:49:38 Forget the trip to the store... :/ 21:49:47 oh? store closed? 21:49:56 No. The car won't start -- battery is dead. 21:50:06 not a good thing ... why? 21:50:15 do you have a recharger? 21:50:37 I must have hit the headlight switch while getting out of the car today, because I found it was on "twilight mode." 21:50:41 Not with me. 21:50:47 But I will come tomorrow... :) 21:51:10 maybe it will recharge itself by tomorrow enough to start the car???? 21:51:17 Doubt it. 21:51:18 ok 21:51:22 ha! not likely 21:51:36 It's been powering the yellow lights and tail lights for about 8 hours already. 21:51:48 stickshift??? 21:51:51 tisk tisk 21:51:52 Auto 21:52:14 I want to switch the tranny for stick in the future though. I abhore that damned automatic transmission. 21:52:20 mrreach: specification/design is ALWAYS the hardest part, espescially with Forth :) 21:52:33 I won't drive anything BUT auto ... but I have hemophilia 21:52:34 kc5tja: this is the new rx-7? 21:52:44 thin: Yeah. 21:52:56 MrReach: I don't see how hemophilia has anything to do with auto versus manual... 21:52:59 you're old rx-7 was auto? 21:53:05 thin: No, it was manual. 21:53:12 s/'re/r 21:53:19 with the rotary engine? 21:53:22 Yes 21:53:23 I've often got a leg that doesn't work properly ... or an arm 21:53:42 MrReach: Ooooh... I didn't know that about hemophilia. What causes it? A clogged blood vessel? 21:53:43 that's too bad 21:53:56 bleeding inside the joints 21:53:58 (sprains) 21:54:04 Ouch... :( :( :( 21:54:10 That really sucks dude... Sorry to hear that. :( 21:54:12 it is a lack of blood cloting ability 21:54:28 it's a damn nuisance, especially if all you have is a stickshift 21:54:52 My primary dislike for automatics stems from two things: fuel consumption (it's worse than a stick), and it's sluggish going up hills (more of the engine power goes into spinning the tranny fluid than being transmitted via the torque converter) 21:55:23 yep, older autos have 5-7% loss on level roads 21:55:41 Even the newer ones aren't up to my preferences. :/ 21:55:41 newer auto are identical to stickshifts at highway speeds 21:55:49 and it's not as much fun, and it doesn't give you as much control, and you can't do fancy things like double-clutching rev matching ;) 21:56:17 Well, in California, we have a ton of hills, and most of my driving is city on top of that. 21:56:18 :/ 21:56:29 So it's a double whammy for me. 21:56:33 yep, used to live in Cali 21:56:39 The sooner I can make enough to swap the tranny, I will. 21:56:52 THen it'll be the sports car it has always wanted to be. :D 21:56:55 heh, again I prefer auto for city driving ... 21:57:00 (actually it's not bad now as it is. But...) 21:57:16 less movement of limbs allows more attention and surrounding hazards 21:57:35 I've avoided more accidents with stick, because I can upshift immediately to accelerate or engine-brake as needed. 21:57:40 and=for 21:57:42 With an auto, I'm screwed. 21:57:51 * MrReach nods 21:57:58 Switching from drive to 2nd takes a whole two seconds in my car. 21:58:11 with auto, your brakes *MUCT* be perfect 21:58:15 (1.5 to realize, "Hey, he really DOES mean to do this!", and another 0.5 seconds to change gears) 21:58:23 * kc5tja nods 21:58:42 stickshift is more fun, though 21:58:50 kc5tja: how much do you think it'll cost to swap the tranny? 21:59:02 Probably about $3000 or so, easily. 21:59:04 will you pick up the stickshift tranny from a wreckyard? 21:59:11 umm 21:59:21 heh 21:59:25 because not only will a new transmission be needed, but they need to remove the old gear select handle and put in the stick. 21:59:35 of course he will, but then have it rebuilt 21:59:46 You presume too much. :) 21:59:53 that adaptation can be difficult 21:59:59 Not for the RX-7. 22:00:12 The engine sits so low and inline with the drive shaft, so it's all plug-n-play. 22:00:20 an RX7 with auto trans? 22:00:24 nifty! 22:00:25 * kc5tja nods 22:00:38 yeah, I bet you DO hate it 22:00:45 Rotary engines are the damn coolest engines ever invented, second ONLY to the gas turbine. 22:01:07 Rotary engine neato keeno 22:01:17 have they managed to impove efficiency any? 22:01:23 Amazingly enough, yes. 22:01:37 The RX-8's 13B RENESIS has an EPA fuel economy of 20/30MPG. 22:01:42 good 22:01:51 well, the milage is only only measure 22:01:57 Considering it's a 250HP engine naturally aspirated, with a 10,500 RPM redline, that's damn sweet. :) 22:02:00 only one measure 22:02:28 actually, come to think of it ... 22:02:50 I'm surprised they aren't using rotaries for HD pickups and deliverly lorries 22:03:05 where efficiency isn't really expected and power is all that counts 22:03:12 The rotary lacks low-end torque (it has a lot of the behavioral characteristics of a gas turbine in that respect) 22:03:30 All its torque comes at the high end, as well as its power. 22:03:40 * thin is on the look out for a Corolla GTS (years 1985 to 1987) in his area :P 125HP with 6800 redline, weighs 1000 kg (or 2200 lbs) 22:03:41 * MrReach nods 22:03:50 i havn't seen a rotary coupled to an electric assit.... electric gives good torque 22:04:09 it should be easy to increase the hp on the corolla gts with some simple mods like a better intake 22:04:16 thin: My roommate has a Corolla GTS. It weighs 3000 pounds, has a 130HP engine, and a 7000RPM redline. 4AG engine, AE86 chassis. 22:04:27 the high RPMs would work beautifully with auto trans, though 22:04:35 jdamisch: Yup, I agree. It'd be ideal for a rotary hybrid vehicle. 22:05:07 MrReach: It actually doesn't. The redline for the auto-tranny RX-8 is only 7500 (about the same as mine). Going much faster runs the risk of cooking the tranny fluid. 22:05:09 how did they improve efficiency? fuel injection? 22:05:19 They've been fuel injected for a long time. 22:05:22 Since 83. 22:05:27 yes, I know 22:05:31 The efficiency comes from a lot of different factors. 22:05:40 I'm not familiar with their history, though 22:05:42 --- join: flyfly (~marekb@ip164.ktvprerov.cz) joined #forth 22:05:44 back 22:05:46 First, the intake and exhaust are done on the side housings, and not through the rotor housing. 22:06:06 and I can't think of how to improve eff except maybe improve friction at apex and side seals 22:06:11 This gives better port placement, so they can eliminate overlap between intake and exhaust (a major problem with the 12A and 13B prior to the RENESIS design). 22:06:21 Next, the seals are different in composition. 22:06:21 * MrReach nods 22:06:26 Next, the rotors are 14% lighter. :) 22:07:06 that shouldn't make TOO much difference, rotating mass and all ... oh, sports, never stable 22:07:08 Next the cooling around the rotor housing is less in the intake, and more in the exhaust, so the intake charge is "pre-heated" somewhat before compression, but after the intake port closes. 22:07:34 The mass of the rotors is very much a critical limiting factor. That's what permits it to spin at 10500RPM without damaging the stationary gears. 22:07:36 ok, I follow that 22:07:53 I thought the apex seals were the limitation? 22:08:04 Remember that the rotors are filled with oil while the engine is in operation too -- the rotors are oil cooled. :D 22:08:05 or was that from flex in the carrier? 22:08:36 didn't know that, either 22:08:41 That problem was solved years ago when the 13B came out in the 1986 RX-7. They switched from single-backed spring-loaded, two-piece seals to double-backed, three-piece seals. 22:09:11 ok 22:09:11 The seals were steel, but ion implanted to make them ceramic-like. 22:09:25 For the RENESIS, that's still true, but they have a slightly different compound for the steel. 22:09:30 moving the ports to the side, though, forces a two-rotor configuration, at most 22:09:37 Nope. 22:09:41 You can get three rotor as well. 22:10:14 I thought that was the case too, but if you study three- and four-rotor designs, they're usually arranged as multiple pairs of rotors. That is, the housings between each of the pairs is thicker and stronger than the other housings. 22:10:20 do you mean to say that they are still on the rotor housing, but more seperation between intake and exhaust? 22:10:29 Because of this extra thickness, you have enough room for the necessary plumbing. 22:10:35 Nope. 22:10:38 oh, ok 22:10:49 The intake and exhaust ports are in the side housings. The peripheral ports are no longer. :) 22:11:02 so there's a plenum type thing between rotors 2 and 3? 22:11:06 one word. twin cam 16valve engine :P 22:11:06 (well, for street legal use anyway; peripheral intake and exhaust is still *THE* king to pump mad HP out of the engine) 22:11:31 thin: One word: 13B bridgeport. 22:11:37 right, but have backpressure probs 22:11:59 No cams, no valves, no lifters, no timing chain, no problems. Just pure 350HP N/A power. With turbo, bump that up to 500HP easy. 22:12:12 I'd very much like to see rotories improved and being more commonly used 22:12:23 MrReach: People do it all the time. :) 22:12:51 I attended SevenStock 5, and there were tons of 400HP engines even without the peripheral ports (most were bridgeported, with forced induction of some kind). 22:13:11 Someone even managed (for a pro racing team) to fit an R26B engine in an FC!! Good grief!! 22:13:31 Nothing like having an 800HP engine under the hood of a car weighing at most 2800 pounds. 22:13:47 on the pickup, I wouldn't mind rotory, as long as I could get the 400HP normal aspiration 22:14:01 really don't care for blowers or turbos 22:14:29 but it would be *WAY* too much work and adaptation 22:14:41 and god knows if it would pass smog or not 22:15:25 If Mazda made another REPU, then it'd pass smog, I'm sure. 22:15:52 heh, but modifying to get into a pickup might screw that up 22:16:02 But they'd probably use a 20B instead of a 26B; cheaper, and a RENESIS variant would give around 350HP or so. 22:16:22 actually, it probably would, my p/us are classified as farm vehicles ... reduced smog controls 22:16:30 REPU = Rotary Engine Pick-Up -- a real truck they made back in the early 70s. :) 22:16:54 cool 22:17:04 Despite the low low-end torque, I've heard nothing but good things about it. But my sources are doubtlessly biased in this area. :) 22:17:13 * MrReach nods 22:17:24 SOmeone did adapt a Toyota T100 pickup with a 20B-REW at SevenStock 5 though, and yes, it was used for hauling stuff around. :D 22:17:35 a planetary immediatly behind the engine would reduce most of those probs 22:17:36 I thought it was the coolest thing. 22:18:12 because the torque converter wouldn't start to grab until the engine was at 3000 or so 22:18:15 I think a 8-speed (yes, 8-speed) manual tranny would work great. I don't like planetary gearboxes. Too little freedom. 22:18:26 nonono 22:18:27 Mine grabs right away. :D 22:18:46 just a gear reduction ... planetary config for strength 22:18:53 OOoh...I see. 22:19:16 Yeah, that makes sense. 22:19:17 that's the point, yours does grab right away ... and its a bit weak in low-end torgue 22:19:46 keep the engine between 4-10,000 22:19:50 * kc5tja nods 22:20:56 My 13B has max torque at 4500RPM and max power at 6500RPM (150HP with stock exhaust); max torque on the RENESIS Is like 2800RPM now, and max power is at 9500RPM. It's an incredible advancement in the performance of the engine. 22:20:57 another 2% loss in torgue to the wheels though 22:21:16 I'm not too sure on the 2800RPM stat though. But still, I do know it's less than mine. LOT less. :) 22:21:24 yes, and that looks very much like port tuning 22:22:12 do you think they'll EVER get the efficiency up to or higher than reciprocating engines? 22:22:48 Yes, but not with 80mm wide rotors. 22:22:49 the rotories would provide substantial weight savings in the hybrid vehicles 22:22:54 The volume to surface area ratio is just too low. 22:23:05 They need to get those rotors much wider before any further improvements will occur. 22:23:23 and then you're battling with carrier flex again 22:23:43 I don't know why people are afraid to put bearings around the eccentric, like they do with a cam shaft. 22:23:46 It just doesn't make sense. 22:23:53 Well, it will make the engine bigger, I suppose. 22:24:15 i need to leave, its been nice chatting 22:24:21 you mean ball bearings, rather than babbit? 22:24:28 ok, jdamisch, be well 22:24:28 --- part: jdamisch left #forth 22:24:32 There are a lot of rumors flying around that Mazda will resurrect the 15A engine for the fourth generation RX-7. If true, then it'd be interesting to compare fuel consumption with that engine. 22:24:47 Whatever type of bearing is used in most engines. 22:24:53 I think a fluid (oil) bearing is often used. 22:24:55 ha! 22:25:03 I haven't a clue what rotories use 22:25:27 know that recip use babit metal with presurised oil 22:25:30 They use a roller bearing as a thrust bearing, but the shaft bearing I think is a fluid type bearing. 22:25:56 * kc5tja nods -- that's a fluid bearing. 22:25:58 ooops, no they don't, I've seen needle bearings 22:26:11 * kc5tja can't imagine a 300HP V8 with a needle bearing. :) 22:26:14 guess it depends on the size of the engine 22:26:23 no, the big ones need it 22:26:27 Really? 22:26:31 How are the "needles" then constructed? 22:26:40 Wouldn't the needle get busted or bent? 22:26:46 babit is relatively soft ... gets hammered out of shape 22:27:04 but it holds an oil film really well 22:27:08 A fluid bearing has zero contact metal to metal, except when starting. 22:27:17 in theory 22:27:25 (most gas turbines actually use an air bearing now-a-days) 22:27:44 in a recip engine, there's gonna be contact when the fuel explodes 22:27:57 Not if the bearing is thick enough. 22:28:10 If it still touches, then the oil pressure isn't high enough. :) 22:28:18 * MrReach grins. 22:28:27 Here's what I want to know though... 22:28:33 How does one keep the oil film in the bearing? 22:28:41 I should go tear apart a couple of trashed engines to learn about new tech 22:28:43 If you've got all that oil pressurized, why doesn't it just seep out the bearing? 22:29:07 two reasons ... there is less clearance at the edges 22:29:24 and the babit metal holds oil like copper and gold hold mercury 22:29:45 * kc5tja didn't know that gold and copper could "absorb" mercury. 22:29:53 they don't 22:30:00 they have an afinity for it 22:30:15 the mercury will cling to copper/glod 22:30:18 gold 22:30:29 I'd like to see how these bearings are made. Because I just can't visualize that the Babit bearing just "doesn't" leak oil. 22:30:38 Ahh, like solder to copper. 22:30:51 if you try to take it off, you can't ... you can only make the layer thinner and thinner 22:30:58 yes, similar to that 22:31:11 of course it leaks oil 22:31:24 I do know that the Babbit bearing is made of a material not unlike solder actually. Lots of lead and tin compounds. 22:31:32 that's why there's a gallery in the crankshaft to supply it 22:31:51 Yeah, but I'm talking about it *spewing* oil. 22:31:59 The pressure to maintain those bearings is a LOT. 22:32:11 --- join: ramnull (~nicad@12-241-145-39.client.attbi.com) joined #forth 22:32:22 yes, one would think it would certainly be higher than 40-60PSI 22:32:28 wb, ramnull 22:32:37 * kc5tja has an idea -- electromagnetic bearings! :D 22:32:49 they are used on occasion 22:33:04 In commercial engines? I've not heard of it being used in that application. 22:33:06 not real good for high-torgue situations, though 22:33:35 hmmm 22:33:43 I think that the turbine powered diesel locamotives use an electromagnetic suspension 22:33:45 What's up? 22:34:20 Gotta take a break. IsForth is pissing me off at the moment. 22:34:26 Heheh 22:34:31 * kc5tja knows that feeling all too well. 22:34:36 That's why it's taking this long to complete FS/Forth. 22:34:40 is that I440r under a different nick? 22:34:45 ALmost 6 months, and it still doesn't interpret a single word yet. 22:34:47 nope 22:34:57 Not a chance. 22:35:00 But I'm close. 22:35:01 ok 22:35:15 kc5tja: why does pygmy/fsforth piss you off sometimes? 22:35:19 what element is it? 22:35:25 I440r kicks my ass in coding. 22:35:26 the interpreter I wrote from scratch took me about 20 days to usable system 22:35:27 poor development enviroment? 22:35:57 the s>z word isn't s>z ing. 22:36:04 pygmy? that's what I used to metacompile my system with 22:36:13 --- quit: gilbertdeb (Read error: 104 (Connection reset by peer)) 22:36:26 what is s>z ??? 22:36:30 I check the new address returned by the word, and it's still a counted string. 22:36:47 MrReach: Changes a counted string to an ascii terminated string. 22:36:59 errr...zero terminated. 22:37:02 where is it stored temporarily? 22:37:11 thin: You get burnt out coding, and suddenly you start making stupid mistakes that you aren't aware you're making. Suddenly, you're yelling and screaming at the computer for not doing what it's "supposed" to do, despite you telling it, right here!, what it should be doing. 22:37:17 The string or the address? 22:37:29 heh ... 22:37:38 MrReach: Well, so far, I ahve about 10 days of coding time invested total in FS/Forth, despite spending 6 months on it. :D 22:37:44 do you mean to say that it converts it in place??? 22:37:53 ah! ok 22:37:56 It makes a new string and returns the address on the stack. 22:38:12 where does the new memory come from? ALLOT ? 22:38:13 MrReach: Funny; I'm using Pygmy to implement the first target compiler for my own Forth. :D 22:38:17 ALLOCATE ? 22:38:31 pygmy is awesome 22:38:52 Yeah, but I'm kinda tired of its quirkiness. That's my primary goal for replacing it. 22:38:55 over + 0 swap c! 22:39:13 huh? 22:39:32 Thats the whole word. 22:39:38 For example, <# # # #> will *print* a 2-digit number, not return the address of the number's string. 22:39:38 that only appends a null byte to the end of the existing string 22:39:48 Most annoying. 22:39:56 yep, pygmy is definately quirky 22:40:07 --- join: karingo (karingo@239.portland-13-14rs.or.dial-access.att.net) joined #forth 22:40:16 ramnull: you're storing 0 to the address?? 22:40:41 No. 22:40:46 It was working last night. 22:40:53 Must have been doing something different. 22:41:06 erm ... 22:41:25 ok, is the region returned by the word the same region suppied to it? 22:41:51 if not, where does the new region come from? 22:42:24 Alright, same address. But it's still a counted string. 22:42:30 ramnull: factor your words more so that you don't have to use OVER ;) 22:42:40 ok, gotcha 22:42:54 thin: I use OVER all over the place. I don't think there's anything really wrong with it. 22:43:09 if you're returning the same region, you're gonna have to MOVE all the characters down one 22:43:20 otherwise, you have no place to store the null 22:43:32 I'll do this " create foostring ," bar" " and then " foostring 3 s>z " 22:43:43 kc5tja: yeah but i try to minimize it 22:43:57 kc5tja: i felt really bad that i couldn't avoid it with the bresenham circle algorithm in forth :P 22:44:07 the "3" is unnecessary and redudant 22:44:36 The common string usage is to represent a string with a double-cell on the stack: (address length). 22:44:54 * kc5tja finds that method is a lot more useful than either NUL-terminated or counted strings. 22:45:00 if you're pointing to a counted string, then you use only the address 22:45:38 Jeez! Propegation is *really* good tonight. I'm hearing stations from Vermont and Massechussettes(sp?) on the radio. 22:45:48 you will run into probs if the count on the stack disagrees with the embedded count for some reason 22:45:56 Well, the docs tell me to pass the number of chars to it also. 22:46:10 what docs are these? 22:46:11 MrReach: That's assuming ," stores the count of the characters. 22:46:33 It segfaults if I just pass the address. 22:46:35 BRB -- I'm going to try and hook a QSO on 40m CW... 22:46:42 he said he was starting with a counted string, and was converting it to a null-terminated string in place 22:46:51 ok, good look 22:46:57 good luck 22:47:08 damn rented fingers, can't do a thing with them 22:48:18 ok, I'll try this as a one-liner, off the top of my head, no testing ... 22:48:44 I got it. 22:49:06 I'm gonna fix that fucking thing. 22:49:18 Fixed it. 22:49:29 Now it works. 22:49:39 : s>z ( a -- ) count 2dup over 1+ swap move ( a len ) + 0 swap c! ; 22:49:55 what "docs" were you refering to??? 22:50:20 The little help files included with the IsForth tarball. 22:50:48 I440r: used ( a len ) to specify a counted string??? 22:50:52 shame on him! 22:51:29 : s>z c@ 1 + over + 0 swap c! 22:51:31 my version moves the entire string into the count to make room for the zero at the end, then places the zero 22:51:57 oh hell! I got an "off by one" error in my code 22:52:16 that doesn't elimate the count, though 22:52:27 eliminate 22:53:48 heh, he's doing something odd ... wonder what it is, though 22:54:15 MrReach: Youre looking at it? 22:54:26 nope, don't have his code 22:54:36 but the specification makes little sense 22:54:49 By the way it looks to me, he should be getting a stack underflow. 22:54:50 ok, let's do a graphic ... 22:55:17 ...4ABCD... 22:55:40 the dots are what comes before and after the counted string in the dictionary or memory region 22:56:06 the "4" is an integer byte containing the number 4 22:56:16 the "ABCD" is the text 22:56:35 back 22:56:38 when finished, you want to end up with ...ABCDx... 22:56:50 right? (where "x" is the null byte) 22:56:50 Correct. 22:57:15 ok, then you can see that the "ABCD" *MUST* all be moved down one byte 22:57:35 so that there's room to append the null byte at the end 22:57:37 Or, the address could be moved up one byte. 22:58:03 ramnull: That works only if you always remember to allocate n+1 characters in the counted string. 22:58:05 no, because the "..." at the end might be valuable data that would be overwritten 22:58:25 Good point. 22:58:58 so you're gonna need a MOVE ... and then something to store the 0 at the end 22:59:34 I'd code the above like this: 22:59:44 the spec for MOVE is ( a-src a-dst num -- ) 23:00:02 : TerminateIt + 0 SWAP C! ; 23:00:15 * ramnull nods 23:00:26 kc5tja: without moving the string itself? 23:00:32 I'm not done yet. 23:00:36 ah! 23:00:36 I have to think about this. 23:01:15 : s>z ( a -- ) count 2dup over 1- swap move ( a len ) + 0 swap c! ; 23:01:18 The spec for s>z is "s>z ( a1 n1 -- a1 ) " 23:02:05 : TerminateIt + 0 SWAP C! ; 23:02:07 --- quit: karingo () 23:02:14 : MoveIt OVER 1- -ROT MOVE ; 23:02:25 : s>z COUNT 2DUP MoveIt TerminateIt ; 23:04:04 Whoops... 23:04:17 : MoveIt OVER 1- SWAP MOVE ; <-- that is correct. 23:04:36 Alright, heres my version... 23:04:45 would MoveIt or TerminateIt have uses elsewhere? 23:05:03 : toz c@ 1+ over + 0 swap c! ; 23:05:05 MrReach: I factor relentlessly, whether or not they would have uses elsewhere. 23:05:21 I do not optimize anything unless I have a critical need to do so. 23:05:45 I generally don't think about optimisation either 23:06:06 ramnull: Remember that C@ consumes the address it has on the stack; therefore, to call it you'd need to invoke it with multiple copies of the string address already on the stack. 23:06:29 Use COUNT instead; it converts a counted string into an explicit string reference on the stack. 23:06:42 I would prob write it like this ... 23:06:49 : COUNT ( equivalent definition: caddr -- caddr+1 u ) DUP C@ SWAP 1+ SWAP ; 23:07:27 : s>z ( a -- ) 23:07:27 count ( a u ) 23:07:27 2dup ( a u a u ) 23:07:27 over 1- ( a u a u a' ) swap 23:07:27 move ( a u ) 23:07:27 + 0 swap c! 23:07:29 ; 23:08:03 I prefer one line per definition, and more telegraphic code. I find excessive comments to be quite distracting. 23:08:12 Gotcha. 23:08:44 actually, those particular comments were to help me keep the stack straight in my head 23:09:10 But if I give the address a dictionary definition, the point becomes moot. I only need to worry about COUNT if I'm working with a buffer yes? 23:09:38 no, the other way around 23:09:57 it doesn't matter when there is a buffer becuase there's usually free space at the end 23:10:02 I see. 23:10:25 it's CRUCIAL not to overwrite data in the dictionary 23:10:43 Gotcha. 23:10:44 what is s>z supposed to do? 23:10:48 HOWEVER ... 23:11:00 Hmmm...or I could assign the address to a lookup table. 23:11:29 you mentioned earlier that the string originated with ," just prior to s>z being called ... is this guaranteed to always be true? 23:11:43 nevermind 23:12:08 MrReach: Well, for this little app, yes. 23:12:32 then just use ," blah blah blah" 0 C, 23:12:45 Actually I was doing create foostring ," barstring" 23:13:03 create foostring ," barstring" 0 C, 23:13:31 MrReach: That does simplify things. 23:13:59 then, when you need the adress of a zstring, use "foostring 1+" 23:15:01 is there a version of ," that doesn't store a count? (in some systems its called PLACE) 23:15:49 oooops, no it isn't 23:18:13 might be called z," or z, 23:19:07 well, I think I'm gonna go play on yahoo chat 23:19:24 --- nick: MrReach -> _MrAway_ 23:19:55 Be back in a bit. Gotta go get a pizza. 23:20:03 --- quit: ramnull ("This isn't Happy Hour!") 23:33:33 kc5tja: got any urls to forth design? (besides moving forth) 23:34:22 Nope. 23:34:24 They don't exist. 23:34:27 (as far as I can tell) 23:35:58 :/ 23:37:48 there's the 11th and 12th chapter to Starting Forth available online but that doesn't really qualify 23:39:54 i liked reading it 23:40:12 Thinking Forth is the best forth design book 23:40:16 except its pretty old 23:40:22 and lacks a lot of stuff 23:41:08 i was just looking for the concepts 23:42:45 Well, what's needed before we can amass a large body of information on the design of Forths is a set of publications and articles on the topic. 23:42:49 This is where thin's site comes in. 23:43:30 yep 23:43:38 soon i'll have the forum and wiki on the site 23:44:02 I've read brodie's books 23:44:08 thin: you have a site? 23:44:16 yeah, look at the topic 23:44:19 forth.bespin.org 23:44:32 ah 23:45:18 Well, getting a good signal from a chap in Missouri. 23:45:43 i suppose i need to start asking for articles.. 23:46:15 it's not going to be a magazine, more of a collection of great articles.. 23:46:20 there's probably some good articles out there 23:46:27 that i could ask the author for permission to put on the site 23:46:45 Definitely. 23:46:56 And plus we might write a few ourselves as time permits. 23:47:10 * kc5tja needs to develop a cirriculum vitae anyway. 23:49:11 hmm forth programmer's handbook is free? 23:49:13 never knew that 23:49:21 ?? 23:49:26 I paid for mine, damnit! :) 23:49:29 its available for download 23:49:34 Heh. 23:49:36 with swiftforth's demo 23:49:38 Cool. 23:49:38 :p 23:49:48 I like deadtree documentation better than online docs anyway. 23:49:53 --- join: deluxe (~deluxe@pD9E4EB7E.dip.t-dialin.net) joined #forth 23:49:53 amen 23:50:10 hmm 23:50:15 bon jour! 23:50:25 guten abend! 23:50:36 Buon journo... 23:53:43 OK, I'm going to bed now. 23:53:48 night 23:53:52 good night 23:53:57 --- quit: kc5tja ("THX QSO ES 73 DE KC5TJA/6 CL ES QRT AR SK") 23:59:59 --- log: ended forth/03.04.17