00:00:00 --- log: started forth/03.12.30 00:02:04 wb 00:12:49 wow. MINIX is slow on this palmtop(HP 200LX) 00:13:19 DOS isn't much faster. actually Elvis on DOS is like 1/4 speed of Elvis on MINIX 00:16:05 for the love of god, why? 00:19:59 --- quit: I440r (Remote closed the connection) 00:34:55 Never used Elvis in DOS... But I don't think it's slow in Minix. 00:35:02 It's my favourite editor there :P 00:38:33 --- join: I440r (~mark4@12-160.lctv-a5.cablelynx.com) joined #forth 01:03:01 --- join: schihei (~schihei@p5085DB02.dip.t-dialin.net) joined #forth 01:27:45 --- join: ayrnieu (~julian@206.61.132.131) joined #forth 01:39:25 hi ayrnieu :) 01:39:32 welcome to #heretic 01:39:35 erm i mean #forth :) 01:40:04 hello I440r =) 01:40:20 You probably know me as 'cleverdra' if you don't know me as this nick. Sorry for getting impatient with you the other day. 01:41:12 lol thats ok 01:41:43 i knew i knew you - from in here - i just couldnt remember WHO you were heh 01:50:25 hi cleverdra ;)) i was missing you long time ! 01:50:35 happy new year ! 01:50:39 Serg - oh, hello =) Happy new year. 01:50:56 * I440r is tired 01:51:10 * I440r needs to write the assembler for isforth too 01:51:22 * Serg started in a new hobby - photo 01:51:24 but i cant get back into coding it... my mind wont focus on it 01:51:44 I440r: so close the aperture tight ;))) 01:51:50 heh 01:52:05 * I440r closes his apetures and goes ZZZzzz..... snort! 01:52:06 :) 01:52:10 nite nite 01:52:22 g'night. 01:52:35 : sleep 01:52:38 bed go tuck 01:52:38 (we all eventually take our eyes and become unconscious) 01:52:42 light off ; immediate 01:53:04 ooops, ~4 AM ;((((( 01:53:09 at I440r 01:57:59 Hi 02:13:57 --- join: schihei_ (~schihei@p5085DDAE.dip.t-dialin.net) joined #forth 02:21:30 --- quit: schihei (Client Quit) 03:40:28 --- quit: hovil (Read error: 60 (Operation timed out)) 03:51:42 --- join: aktnot_ (ident@233.80-202-65.nextgentel.com) joined #forth 03:59:31 --- quit: aktnot (Read error: 60 (Operation timed out)) 04:20:08 --- join: hovil (~hovil@CommSecureAustPtyLtd.sb1.optus.net.au) joined #forth 04:23:09 --- join: qFox (C00K13S@cp12172-a.roose1.nb.home.nl) joined #forth 04:26:25 --- join: crc (~crc@ACAFE2E1.ipt.aol.com) joined #forth 05:01:41 --- quit: Serg () 05:06:22 --- quit: hovil ("Leaving") 05:28:43 --- join: tathi (~josh@pcp02123722pcs.milfrd01.pa.comcast.net) joined #forth 06:22:40 --- quit: crc ("I was using TinyIRC! Visit http://www.tinyirc.net/ for more information.") 06:33:26 --- quit: tathi ("leaving") 06:59:10 Moo. 07:05:19 --- join: tathi (~josh@pcp02123722pcs.milfrd01.pa.comcast.net) joined #forth 07:10:53 --- nick: chrisrw -> asiammyself 07:10:58 --- nick: asiammyself -> chrisrw 07:12:50 --- nick: chrisrw -> asiammyself_hate 07:13:00 --- nick: asiammyself_hate -> chrisrw 07:48:06 anybody awake? 07:49:36 Nope. 07:50:23 nkay 07:50:35 what do you think the minimum required floating point ops in a CPU are? 07:50:45 0 07:50:59 er 07:51:07 well, if FPU is to be present. 07:51:16 lemme rephrase this question: 07:51:47 i need the 8 floating point instructions (operating on regs) which will create a complete floating point environment 07:52:30 Uh... hmm I don't know. 07:52:51 multiply, div, add, sub ... 07:53:02 and some compares? 07:53:14 well, i dont need to worry about branches. 07:53:42 well, i do, but you dont 07:53:44 I honestly don't know. 07:53:51 I've not done that kind of thing. 07:53:53 i got add, sub, mul, div 07:54:02 8 more 07:54:06 I440r! 07:55:29 i dont need to worry about signs, they can take care of that themselves (as the top bit is the sign :) ) 07:56:33 a simple (and/or/xor) 1 would work wonders there. 08:00:47 mod? 08:01:13 well you can do it with div actually.. 08:01:31 mod and div are together 08:01:33 fdivm 08:01:54 fdivm dividend, divisor, result, remainder 08:04:02 * qFox nods ... 08:05:14 If I'm not insane wouldn't you do and/or/xor using the normal ops? Why would they need specific fp versions? 08:09:13 tbh i cant think of any other floating point specific instructions, then again i dont really know much about instructions anyways :) 08:10:49 --- quit: warp0b00 ("Ce soir, ce soir, assanation d'un rock and roll star.") 08:12:26 yeah, you don't need mod. Actually, PPC doesn't have a mod instruction 08:12:43 you have to do div mul sub 08:13:48 oops. 08:14:01 * tathi should have finished reading the conversation first 08:19:47 :p 08:20:32 --- quit: qFox ("if at first you dont succeed, quit again") 08:23:01 --- join: qFox (C00K13S@cp12172-a.roose1.nb.home.nl) joined #forth 08:34:38 tathi: well, what else does PPC have? 08:34:47 --- join: aktnot (ident@233.80-202-65.nextgentel.com) joined #forth 08:35:02 tathi: i give you 8 fp instructions to fill as you please.. 08:39:49 --- quit: aktnot_ (Read error: 60 (Operation timed out)) 08:40:39 chrisrw: Without compare ones, I think you really only need 4 dude. 08:41:11 well, currently, i got 08:41:23 add, neg, mul, reciprocal, sqrt 08:41:47 if you have sqrt then you need pwr 08:42:19 also how about log? 08:42:23 that can be easily done in software. 08:42:26 log? er.. 08:43:17 --- quit: aktnot ("leaving") 08:43:30 --- join: schihei__ (~schihei@pD9E5CC2F.dip.t-dialin.net) joined #forth 08:48:03 * chrisrw is away: work 08:51:08 yeah, you really don't _need_ much. 08:51:23 you can get away with add, neg, mul, and reciprocal. 08:51:55 I've never done much fp stuff 08:52:17 I think if you're doing the usual fp stuff (matrices, vectors, and so on) the multiply-add instruction is very useful. 09:00:52 --- quit: schihei_ (Client Quit) 09:04:46 --- quit: schihei__ (Client Quit) 09:20:53 --- quit: skylan (Read error: 60 (Operation timed out)) 10:01:15 --- join: aktnot (ident@233.80-202-65.nextgentel.com) joined #forth 10:06:58 --- join: kc5tja (~kc5tja@66-91-231-74.san.rr.com) joined #forth 10:06:58 --- mode: ChanServ set +o kc5tja 10:21:27 * tathi is really liking this "test everything" concept. 10:22:17 test everything? 10:23:02 unit testing, whatever 10:23:23 have a test framework to run tests automatically 10:23:32 write the tests before you write the actual code 10:23:38 what kind of tests? 10:23:53 then once you write the code, you _know_ whether or not it actually works 10:24:54 well, in this case tests for all the subroutines that are going into my new colorforth implementation 10:24:55 i'm not sure what you mean, but i usually write my code on trial-and-error base, rather then thinking ahead :) 10:25:43 :) 10:26:31 well, I'm writing asm code, and I got tired of spending most of my time debugging. 10:26:41 so far that hasn't been a problem in forth. 10:27:30 what i would like is some kind of mechanism that shows the stack while editing..... 10:27:42 :) 10:27:47 qfox err why ? 10:27:55 because i have a problem thinking ahead 10:27:59 that's a sign that you have too many things on the stack :P 10:28:03 isforth doesnt show whats on the stack but it tells you the depth 10:28:12 and i noticed that thats kind of important with teh stack and all, while coding sometimes 10:28:15 'morning. 10:28:16 in forth i mean 10:28:54 hi kc 10:29:02 re tathi 10:29:02 I440r> i can run into "problems" with 4 items on the stack... 10:29:05 * kc5tja just woke up. 10:29:30 right now i simply put the stack in comments behind the commands in those cases 10:29:34 Well, I woke up about 30 minutes ago, but I was eating breakfast. 10:30:03 but i wouldnt mind an editor where you can "give it an example begin-stack" when defining a word 10:30:06 qFox: That is good that you're starting to recognize this now. Many folks who write Forth code don't ever realize this as a problem. 10:30:36 qFox: However, I find an editor wouldn't help me, but rather factoring your code into smaller and smaller pieces is a substantial help. 10:30:36 and that the editor can give me at any point in the code, what the stack would probably look like at that point of the definition 10:30:52 yes but if there's an editor that can do it for me... ;) 10:30:55 it helps while coding 10:31:14 and ye, factoring in forth is as important in forth as it is in asm (and probably others) 10:31:51 I'd hazard a guess it's actually more important in Forth than in Assembly. 10:32:01 well i dont know about that 10:32:21 * tathi agrees with kc5tja 10:32:28 so be it :p 10:32:54 anyways, if an editor would do this, i'd be happy 10:33:06 Based on my experience writing OS kernels in assembly, and general applications in Forth, I find Forth is more factor-heavy than assembly. Even with x86, though you have limited amounts of registers, the ability to freely access them in any arbitrary order alleviates much need for factoring. 10:33:10 Anyway, it's a dead subject. :) 10:33:39 * qFox nod 10:34:47 :( 10:34:50 Exactly as I predicted. 10:34:58 The FCC is going forward with the roll-out of BPL. :( 10:36:03 AZ has an FCC temporary permit to use BPL. Frequencies covered range from 2MHz to 38MHz. And that just so happens to cover *every* HF amateur radio band except the 160m band, which of course, is next to impossible to use in congested city (or, heck, even suburbia) residences. 10:36:22 that sucks 10:36:40 you mean your going to be effectivly off the air from now on ? 10:36:42 Especially here in California, where 90% of the population is under severe antenna-related restrictions enforced by law. :( 10:36:58 bpl? 10:36:58 No, but our useful range is *severely* diminished. 10:37:05 Kiss goodbye to worldwide conversations. 10:37:12 why would they do that ? 10:37:17 $$$$ 10:37:27 The power utilities all want in on the Internet/broadband thing. 10:37:38 err the fcc is a government agency, not a profit making concern is it ? 10:37:52 I440r: Name me one government agency that isn't also a profit making concern. 10:38:07 err they are all LOSS making concerns if you ask me :) 10:38:09 Remember that the republicans (e.g., big oil and established utilities) are primarily in power. 10:38:31 So naturally, it follows that the will of the people be utterly and completely ignored in the matter. 10:38:42 Fuck the people, let's let the big monopolies decide. 10:38:59 >:( 10:39:09 At least we still have primary status on the relavent bands. 10:39:19 And that means if they interfere with us, we interfere with them. 10:39:24 All 1,500 WATTS of us. 10:39:43 And there is not a fucking thing (legally) they can do about it. 10:40:09 But even so, transmitting isn't the problem; it's *reception* that is always the problem with ham radio. 10:41:03 If you're trying to grab that station from Japan, or that rare Cuban station for those collectors items Cuban QSL cards, "Fuck you" says FCC, "You'll just have to deal with this thing we call `progress.'" 10:41:26 "or that rare Cuban station for those collectors items Cuban QSL cards" rofl 10:41:30 All you'll hear on the radio is the characteristic WHISH WHISH WHISH WHISH WHISH WHISH of broadband over power line. 10:41:37 qFox: I'm serious. 10:41:44 qFox: Cuban stations are rare. 10:41:52 and why would that be.. 10:41:53 qFox: And getting a QSL card from a Cuban station is even rarer. 10:42:09 i dont know what a QSL card is.. 10:42:42 qFox: Gee, the economic embargo we've had against Cuba for decades might have something to do with it. The fact that the Cuban government doesn't let a lot of people have ham radio licenses also might have something to do with it. *AND* the fact that their postal service monitors *every* piece of mail in or out of the country also has a lot to do with it. :) 10:42:45 but i guess you collect them? :p 10:42:57 qFox: If I got one, I'd frame it, certainly. 10:43:04 A QSL card is a written confirmation of a radio contact. 10:43:09 oh 10:43:52 and the why was more of a joke, sorry, but what would be so special about a cuban radio station? unless you're cuban yourself... 10:44:29 Ham radio is a two-way radio communication service. 10:44:41 i know you meant your rant, but the cuban example struck me as funny :) 10:44:47 On the HF bands, you can get world-wide communications with relatively modest equipment. 10:45:34 Most commercial stations have 50kW or 100kW shortwave stations, while we often can do the same things (albeit with reduced audio quality) with as little as 50W to 100W. Depending on propegation, we can even do it on milliwatts to watts. 10:46:19 It's a fun hobby. 10:46:30 But, ... it won't be much fun if you can't hear overseas stations with it. 10:47:06 But I digress. 10:47:14 well thing is, imo air frequencies are like ip-addresses 10:47:31 except, there arent so many of them, and i doubt they can "infinitly" increase those numbers 10:48:30 so if there's a station in cuba that airs on freq x, i'm not surprised that this frequency is used by something/somebody else in an area far far away from cuba. at least thats what i understood was or shall be the problem you mentioned above here... 10:49:17 its like FM frequencies in holland... only 10 of them or something, and radio stations pay big bucks for them to stay alive (god that was lame) 10:49:32 and holland is relatively very small 10:49:57 yet most stations cannot be received thruout all of holland 10:50:37 well thats just my 2 cents :) 10:55:38 You're correct, but you're describing a completely different problem. 10:55:48 What I'm saying is, we have frequencies allocated exclusively for amateur radio. 10:55:53 then i'm afraid i misunderstood you 10:56:00 aye... 10:56:05 These allocations are international in scope. 10:56:15 are these freq's assigned? 10:56:20 E.g., our 40m shortwave band is from 7.0 to 7.3MHz. 10:56:24 Yes! 10:56:28 That's what allocated means. 10:56:32 no i mean 10:56:36 specific 10:56:41 to a radio station 10:56:43 all over the world? 10:56:51 There is no single radio station. 10:57:00 eh.. 10:57:03 There are millions of radio stations, all of which share the same set of frequencies. 10:57:21 yes.. 10:57:29 thats sorta what i meant 10:57:35 But that isn't the issue. 10:58:03 The issue is, that a non-amateur radio related service will be polluting OUR radio spectrum with THEIR bullshit. 10:58:28 And because Arizona is MUCH closer than Cuba for me, it necessarily will sound much, much louder. 10:58:49 In short, the Cuban station will get drowned out in a sea of noise inpenetrable to distant stations. 10:59:59 This is true even if they're using much lower power levels. 11:00:20 AZ has an FCC temporary permit to use BPL. 11:00:29 then what does this fcc do? 11:00:36 Federal Communications Commission. 11:00:40 take in mind that i dont know this term 11:00:44 oh 11:00:53 They are the United States' legal body that allocates and enforces spectrum usage policies. 11:00:56 and they deal/assign frequencies? 11:01:04 For this country, yes. 11:01:16 ic 11:01:45 They are obligated to conform to the IARU (International Amateur Radio Union, a division of ITU, the International Telecommunications Union) with respect to ham radio frequencies though. 11:01:46 so basicly they're breaking international agreements about the amateur-frequencies? 11:01:57 At least on a local basis, yes. 11:02:07 so is there anything that can be done about it? 11:02:15 Two things. 11:02:50 well... its not the first time. i'm sorry but this once against sorta proves that usa will agree to anything untill they dont agree to it... 11:02:53 --- join: skylan (~sjh@vickesh01-4902.tbaytel.net) joined #forth 11:02:58 One, we can increase our power and concentrate on local conversations so much that the BPL folks just give up in disgust (remember, we have primary allocation status on these bands, so they must 'accept interference' from us). 11:03:10 right, I figured that one out. 11:03:10 Two, we can petition the IARU to help clarify the situation. 11:03:43 ah, FCC thinks they can slip BPL through some loophole or other? 11:04:12 tathi: I wonder how they will mitigate interference from adjacent states or, heck, from adjacent power lines in the same state. 11:04:29 Consider: power lines are long. Very long. And the wavelengths these folks are considering using are ... long. 11:04:31 Very long. 11:04:35 yeah 11:04:41 So, it follows that these things are big, giant, ANTENNAS. 11:04:45 yup yup 11:04:54 (which is why such little power on their part makes life pure misery for a ham radio operator in a city) 11:04:59 don't know a lot about radio, but I got that far. 11:05:32 seems like it would make more sense for them to use higher frequencies, but I suppose that makes the equipment substantially more expensive 11:05:38 Couple that with the fact that the HF spectrum propegates world-wide on certain bands, and at least hemispherically on the rest, and you can easily see that even a small signal has huge potential for interference with another. 11:06:23 tathi: The problem is, you can't use higher frequencies on the power lines due to losses. The characteristic impedance of the power lines is so high that the dB loss per 100 ft of cable would make it utterly unusable. 11:06:38 Power lines just plain suck for broadband. 11:06:40 Period. 11:06:55 oh. right. 11:07:04 The power utilities have ZERO technical merit to their situation, and they're using their brute-force in the industry to get the government to sway to their side. 11:07:21 just like everything else in this damn country 11:07:42 Meanwhile, those of us who have radios more sensitive than most radio telescopes are left out in the cold by the FCC, with no significant means of legal recourse. 11:08:45 more sensitive than most radio telescopes? really? 11:08:49 Yep 11:09:25 The average ham radio has an antenna sensitivity of 1uV across a 50-ohm load for its minimum discernable signal, and provides gains as high as 160dB. 11:09:42 Some, like the Yaesu FT-1000MP, have even better specs. 11:09:49 ok 11:09:57 :) 11:11:17 Some, like the pixie 2, are a lot worse. 11:11:37 Sure 11:11:46 But what do you expect from a DC receiver with no filters and amplification only in the audio stage? ;) 11:12:12 You can still get good dynamic range with such a receiver. 11:12:24 You just lose the selectivity of a superheterodyne type, that's all. 11:12:38 The question is, What do you expect from a homebrew receiver made with hobby-grade components? :) 11:12:54 I didn't expect it to work, to be honest. 11:13:04 I know my limitations as a radio constructor. ;) 11:13:22 Well, don't feel so bad. 11:13:33 I don't. :) It's cute. 11:13:37 Want to see a picture? 11:13:40 I built my +66dB AF amplifier for an experimental binaural receiver that I wanted to try. 11:13:51 http://robert.zizi.org/pub/projects/pixie2/ 11:13:56 The amplifier sounded so good that I decided to use it as a general purpose AF amplifier for other things. Oh, it was *SO* warm. 11:13:57 The pixie, and another 80m TX. 11:14:01 It reminded me of a vacuum tube amp. 11:14:47 I'm rather surprised that it works too -- those are awfully long leads for the components. :) 11:14:56 Deadbug style construction is quite common in homebrewers though. 11:15:54 Robert: So do you transmit on 80m often? 11:17:02 Also, do you do morse code or phone? 11:22:55 All I know is that I would like to hurry and get my Forth environment running. 11:24:20 I want to start making some kits, and selling them to make some additional income (which is much needed for me at the moment). 11:24:50 Whoa, I skipped a line or two on that one. :) 11:25:06 FS/Forth is going to be used to help lay out PC boards and maintain inventory and what not for my kit business idea. 11:25:35 Maybe with enough experience, I can get FS/Forth to also simulate electrical circuits for me as well. 11:26:11 It'd be nice to be able to lay out a PCB, place all the components on it (somehow), and run the simulator. :) 11:26:26 kc5tja: Er, sorry, went away for a while. 11:26:31 kc5tja: I haven't got my license yet :) 11:27:12 kc5tja: But I only have 80m crystals. 11:27:33 3.579545MHz is a very popular crystal here in the USA. :) 11:27:40 It's the NTSC colorburst frequency. 11:28:11 I'm sure PAL has a similar frequency that you can get crystals for very cheaply. 11:28:29 3.6MHz (roughly) seems to pop into my mind. 11:33:06 I have a few 3.6864 11:33:18 But I'll buy those 3.5795MHz ones too. 11:33:30 And some others I found at Mouser. 11:33:42 3.6864 is what I was thinking of. 11:33:46 Anyway, I gotta go. 11:33:46 Thinking about dividing down a signal and playing at the 136kHz band. 11:33:51 OK, bye! 11:33:52 * kc5tja has to get his car door's lock fixed. 11:34:05 * kc5tja is away: locksmith 11:57:15 --- quit: qFox ("if at first you dont succeed, quit again") 12:13:22 --- quit: aktnot ("leaving") 12:50:34 --- join: meme (~meme@216.184.11.30.unused.swcp.com) joined #forth 12:51:44 --- quit: I440r ("Leaving") 12:55:19 --- join: I440r (~mark4@12-160.lctv-a5.cablelynx.com) joined #forth 12:55:19 --- quit: I440r (Client Quit) 12:55:39 --- join: I440r (~mark4@12-160.lctv-a5.cablelynx.com) joined #forth 14:04:35 --- join: snowrichard (~richard@adsl-068-209-159-248.sip.shv.bellsouth.net) joined #forth 14:04:40 hi 14:04:42 --- join: aktnot (ident@233.80-202-65.nextgentel.com) joined #forth 14:04:48 hello aktnot 14:06:25 hwt snowrichard 14:06:27 hey 14:07:35 hi mur. 14:07:41 can you do a favor? 14:07:49 test my server 14:08:35 http://richardsnow.is-a-geek.net 14:09:25 ok 14:10:43 well works for me but you know... 14:12:43 post works ok 14:14:31 ok i reply 14:14:42 thank you 14:15:11 :) 14:15:35 tried this forum software today to answer somebodys question about changing the colors 14:15:40 pretty cool 14:15:47 easy to understand 14:16:23 --- join: qFox (C00K13S@cp12172-a.roose1.nb.home.nl) joined #forth 14:19:21 thank you. I think is is pretty cool. My son did not understand PHP-NUKE ui 14:19:31 maybe this one he can use 14:21:12 :) 14:26:35 Hi 14:28:08 --- join: richard_ (~richard@adsl-068-209-159-248.sip.shv.bellsouth.net) joined #forth 14:28:16 lost my window 14:28:37 --- quit: snowrichard (Nick collision from services.) 14:28:46 --- nick: richard_ -> snowrichard 14:28:50 --- join: richard_ (~richard@adsl-068-209-159-248.sip.shv.bellsouth.net) joined #forth 14:29:42 http://adsl-068-209-159-248.sip.shv.bellsouth.net 14:30:16 what an ugly url. 14:30:24 but it works :) 14:30:40 and is my static IP 14:31:40 static IP are nice 14:31:48 14.95 a month more. 14:32:42 what gets me is phone company invented unix what like 40 years ago and they don't support it as a client :)))) 14:33:29 AT&T's belongings are owned by bellsouth.. hmm.. bell labs 14:33:31 rings bells 14:33:32 ok 14:33:33 att bell labs 14:33:58 Kernighan and Ritchie C :) 14:34:20 :) 14:34:33 nice book 14:34:39 I memorized it I think 14:34:44 which one? 14:34:51 "The C programming Language" 14:34:52 " Kernighan and Ritchie C :" ? 14:34:58 by those fellows? 14:35:01 yep 14:35:56 check amazon for it :) 14:35:57 aha 14:36:11 * mur does nto have so much time for programming anymore 14:36:20 dsnto 14:36:26 rreal job or school? 14:36:44 --- join: networm (~networm@L0626P23.dipool.highway.telekom.at) joined #forth 14:36:52 gees I can't type today either 14:37:00 hi worm 14:37:20 hi 14:37:35 you want to try to crash my BBS? 14:37:42 richardsnow.is-a-geek.net 14:38:14 no.. i'm not a computer worm.. 14:38:19 just joking 14:38:30 university 14:38:33 heh 14:38:33 takes. all my time 14:38:41 school can be tough 14:38:44 my hacking skills are 0 in any case 14:38:53 class all day study at night no sleep 14:40:02 art school, that is where i study. mentally hard. we have such quick and demanding tasks 14:40:09 I have been programming since 1975. I use the "hacking" in the mit sense of the word 14:40:17 few days time to produce book or what ever. it takes lot of work and even more thinking 14:40:50 I am working on ebook about E-commerce with PHP 14:41:03 mysql 14:41:05 apache 14:42:07 wanted an easy to use forum software for tech support line 14:42:27 this one will do I think 14:44:00 brb 14:44:29 water 14:45:10 --- quit: tathi ("leaving") 14:47:17 just posted url to article about FESTIVAL speech synthesis 14:49:54 snagged the source for it 14:51:41 Festival is diphone synthesis, isn't it? 14:52:38 http://robert.zizi.org/f.html <-- in case anyone is interested. 14:57:54 yay 15:06:47 Hey, it's new years eve now 15:06:54 At least in Sweden..and of course in Finland. 15:07:17 www.dyndns.org 15:08:44 years eve 15:08:45 hmm 15:08:52 madgarden: What? 15:08:57 31.12. 15:11:38 * kc5tja is back (gone 03:37:32) 15:15:58 Hi :) 15:20:15 :) 15:20:25 It took 'em long enough. 15:20:35 The door still isn't 100% perfect, but it's quite servicable. 15:22:23 Neat 15:22:35 A neighbour hit our car 15:22:48 Not terribly damaged though 15:23:35 btw, what would be the best thing to improve in that simple CW transciever? Adding more amplification and a filter at AF? 15:24:46 Not owning such a transceiver, I don't know. 15:25:45 I'd hazard a guess that the receiver can always use more amplification to dig out those weak signals. But as soon as you do that, you'll start to need better filtering, notch filters for those pesky broadcast stations, etc. 15:25:49 But you could guess, from your general electronics knowledge :) 15:26:08 No, I can't even guess. 15:26:18 (and I do say that precisely because of my prior electronics knowledge.) 15:26:38 Though I hazarded a guess anyway. 15:27:17 Hehe, thanks. 15:28:02 I guess a one-frequency crystal filter could be a way... 15:44:21 * I440r just fixed the rear sight group on his ar-15. someone very kindly BENT my windage screw while i was away grr 15:44:40 luckilly the local gun store had a replacement part for $10 15:45:00 if they hadnt have had that i would have had to have replaced the ENTIRE rear sight group for $75 15:46:58 brb - going to video store 15:48:27 --- quit: ayrnieu ("system") 15:53:14 --- quit: networm ("Client exiting") 15:55:26 --- join: ayrnieu (~julian@206.61.132.131) joined #forth 15:56:32 snowrichard: Use www.dyndns.org to hide your fugly URL. ;) 16:12:20 lol 16:16:19 --- join: TheBlueWizard (TheBlueWiz@pcd2dn08.ppp.FCC.NET) joined #forth 16:16:19 --- mode: ChanServ set +o TheBlueWizard 16:16:24 hiya all 16:16:30 re TheBlueWizard 16:16:49 Hey kc5tja 16:16:51 hiya kc5tja 16:20:35 * chrisrw is back (gone 07:32:34) 16:21:09 hey kc5tja 16:21:12 hey TheBlueWizard 16:22:14 hiya chrisrw 16:23:50 TheBlueWizard, kc5tja, I440r, everyone else: What do you guys think the most essential floating point instructions are? (and how to possibly combine them). You can either have 4, 8, or 16 of them :) 16:24:02 and you dont need to include any loads, stores, or copies 16:24:08 or compares 16:24:12 just pure arithmetic etc. 16:25:33 that question is a bit subjective...no question that arithmetic instructions are a must. But there are numerous applications for which 'higher' fp commands are needed 16:25:50 so in short I can't give a good answer 16:26:06 + - * / sqrt sincos (yes, one instruction that generates both sin and cos functions) power ln 16:26:34 ln being log to (you choose) base, right? 16:26:50 how would you do tan? 16:27:06 No. 16:27:12 ln being natural logarithm. 16:27:27 oh, so log e 16:27:41 You can compute any arbitrary logarithm using either log_10 or log_e (where ln(x) = log_e(x)). 16:27:49 tan(x) = sin(x)/cos(x) 16:28:13 oh yeah, duh. 16:28:26 lets see ... thats 8 instructions. 16:28:28 sw33t 16:28:39 you da man, kc5tja 16:28:41 :) 16:29:26 are there any processors (that you know of) that accept only registers for instruction arguments except for those that set registers 16:29:27 ? 16:30:14 --- join: crc (~crc@ACAFF183.ipt.aol.com) joined #forth 16:30:16 um, what about exponentiation? it should be included with ln as a natural complement 16:30:33 oops, I saw the word "power"...never mind 16:31:10 chrisrw: I do not understand your question. 16:31:48 kc5tja: like, my CPU idea's only instruction that accepts an immediate (as opposed to a register) is set 16:31:58 which is to set a register to something. 16:32:24 chrisrw: Every RISC architecture CPU I can think of is built the same way. 16:32:35 kc5tja: erm, no? 16:32:43 kc5tja: MIPS isn't like that. 16:32:47 Sure it is. 16:32:51 PPC has rlwinm, so it doesn't count either. 16:33:18 Instructions can take only registers or immediate values -- that's in. 16:33:22 that's it even 16:33:35 er 16:33:38 well, thats not what I meant. 16:33:58 i meant that ALL instructions take ONLY registers, except for the one that sets the values of registers 16:34:37 Well, that is certainly possible to do, but I fail to see the advantage of that. 16:34:52 It's still considered a LOAD/STORE architecture in that case. 16:35:06 yeah. 16:35:14 the reason I wanna do that is to make things easier :) 16:39:25 --- quit: aktnot ("leaving") 16:41:58 * warpzero is away: bee to the are to the bee 16:42:07 --- quit: warpzero (Read error: 54 (Connection reset by peer)) 16:42:20 --- quit: crc ("I was using TinyIRC! Visit http://www.tinyirc.net/ for more information.") 16:42:45 --- join: warpzero (~warpzero@dsl.142.mt.onewest.net) joined #forth 16:43:01 should fsincos take radians or degrees 16:43:02 ? 16:45:26 for math reason, radians is usually used 16:45:38 ok makes sense :) 16:45:48 ditto with ln and e^x 16:45:54 that means i need to keep an internal version of pi 16:46:11 yep 16:47:57 Well, ln and a^b don't take radians. 16:47:57 :) 16:48:12 But fsincos definitely should take radians. 16:48:37 I am talking about e needed for ln and e^x as vs. 10^x and log10 16:48:57 sorry for typing it out of context 16:49:00 The problem is, when using something like e^x, the value of 'e' is hardwired in that operation. 16:49:09 no 16:49:20 I prefer to load 'e' as a constant, then use a generic power function. 16:50:20 errgnphhhbl 16:50:26 it isn't so...e^x can be written as 1 + x + x^2/2 + x^3/6 + x^4/24 + ..... and there is no 'e' value anywhere in it 16:50:27 so what to do? 16:50:27 :) 16:50:31 However, for logarithms, I agree log-base-e is superior. 16:51:01 TheBlueWizard: Given only the hardware function e^x, compute any arbitrary power to any arbitrary base. E.g., compute 3^21. 16:51:23 (multiplying 3 times itself 21 times is not acceptable. :)) 16:51:25 yes...I know 16:51:36 * TheBlueWizard is a bit annoyed with kc5tja 16:51:54 lol 16:52:08 Well, no offense, but that isn't my problem. We're talking about a hardware implementation of math coprocessor functionality. 16:52:15 Not a vague algebraic equivalency. 16:52:33 Of course I know that e^x can be written as a Taylor expansion. 16:52:46 kc5tja: you can use various tricks to speed up 3^21, just like you would with multiplication (e.g. 7*11 can be accomplished by 11<<3 - 11) 16:53:23 TheBlueWizard: That's fine, as long as you know a priori you'll be multiplying a number precisely by 7. 16:53:34 What if you don't, though? 16:53:55 When I design hardware, I design for the biggest average bang for the buck. 16:54:08 I had only eight instructions to choose from (well, OK, 16, but still). 16:54:30 I can compute any arbitrary log given only ln, but I cannot do the converse given e^x. 16:54:48 At least, not in any way that would reduce clock cycle counts. 16:55:21 I think I made a reasonable decision given my set of constraints. 16:56:33 Anyway, I'll shut up now. 16:56:41 It seems I just serve to piss people off. 16:58:24 * TheBlueWizard is afk 16:58:52 kc5tja: naah, I value your input muchly :) 16:59:14 kc5tja: well, for my simulator, I'm not gonna implement floating point until later on 16:59:26 especially since I myself pretty much never use floats anyway 16:59:36 Well, TheBlueWizard has a good point too. But he was getting upset with me because he was talking about one problem, and I was trying to solve another. *shrug* 17:02:04 * TheBlueWizard is back 17:02:17 yeah....generic function requires generic algorithms. But in case of integral powers, you can use bits to speed up the work...for ex, 3^21 = 3^10101, which is 3^(16 + 4 + 1) so you compute square of square of 3, save it, then compute square of square of that number, then multiply it by the saved number, then multiply it by 3...done...this is similar to Russian multiplication method 17:02:56 I was upset because of kc5tja's rather obstinate stance on certain points 17:04:36 * kc5tja never did get a full understanding of the Russian multiplication method. 17:04:46 * kc5tja will have to research that more fully in the future. 17:05:18 that method is pretty neat in its simplicity and is easy to understand 17:07:41 The Egyption method pretty much is binary multiplication. 17:07:46 http://www.pballew.net/old_mult.htm 17:09:58 same method :) 17:11:39 http://mathworld.wolfram.com/RussianMultiplication.html 17:13:38 It turns out that I use Russian method quite often, without realizing it. :) 17:14:31 so do you understand how it works then? 17:17:48 Yeah. 17:17:58 The last time I saw it explained to me it was rather complex. 17:18:06 But it turns out I do that kind of math all the time in my head. 17:18:51 I'm constantly doing things like, "What's 128 times 6?", and then going "256x3", and from there it gets sticky (I interpolate), "512+256", to finally get "768." :) 17:22:34 lol 17:22:48 i figured out 256*256 yesterday in my head 17:22:49 by going 17:23:17 512*16 = 1024*8 = 2048*4 = 4096*2 = 8192*1 17:25:24 chrisrw: what about using the logarithms? like 256=2^8 so 256*256=2^(8+8)? 17:28:28 the Russiam Multiplication is really a binary multiplication in disguise...the condition where you don't add certain numbers corresponds to the zeroes in the binary representation of one of the number...that's why it works that way 17:28:30 chrisrw: The correct answer is 256*256 = 512*128 = 1024*64 = 2048*32 = 4096*16 = 8192*8 = 16384*4 = 32768*2 = 65536. :) 17:31:31 er, yeah, whatever 17:31:33 lol 17:31:36 * chrisrw wasn't thinking 17:31:42 but i did figure it out right yesterday 17:37:01 --- quit: qFox ("if at first you dont succeed, quit again") 17:37:56 :) 17:38:33 It's all pretty fascinating. 17:38:44 The one thing that does fascinate me a lot is the Chinese abacus. 17:38:58 There's a tool that people have used for millenia, still use to this day, and will continue to use virtually forever more. 17:39:06 Some people are lightning fast with those things. 17:40:28 kc5 - have you read Feynman's account of an arithmetic battle with an abacus-seller? 17:40:49 Yes, but I don't recall the specific details. 17:41:05 I know that the Abacus seller was able to beat the calculator-using person on several accounts though. 17:46:24 kc5 - you don't recall the general details, then. Nobody used a calculator in that account. 17:46:37 Different account then. 17:46:51 bank account? 17:46:52 :) 17:47:49 There was an experiment conducted shortly after WWII (e.g., within 10 years) that pitted an American accountant (with his adding machine/calculator) against a Japanese accountant, armed with only a Chinese/Japanese abacus. 17:48:13 The Japanese devasted the American in all but two cases (both cases involved computing the powers of numbers). 17:48:47 I wish I could recall more details than that, though. Otherwise, I don't have enough info to find the reference. 17:55:30 * chrisrw wants an abacus 17:55:32 they're cool. 17:59:38 kc5tja: do i have to provide specialized FP compares, or would the regular compares work? (assuming IEEE) 18:00:26 Yes, you need to provide specialized compares. 18:00:36 However, '-' is sufficient for this task. 18:00:46 Then you need only check to see if a number is positive, 0, or negative. 18:19:10 --- join: proteusguy (~proteusgu@216.27.161.121) joined #forth 18:39:39 * kc5tja is away: Tandy Shack 18:39:58 --- quit: I440r ("brb") 18:45:04 --- join: I440r (~mark4@12-160.lctv-a5.cablelynx.com) joined #forth 18:45:38 hi madgarden 18:45:39 er 18:45:40 Mark 18:45:41 lol 18:45:49 hiya I440r 18:53:32 --- quit: snowrichard (Nick collision from services.) 18:53:51 --- join: snowrichard (~richard@adsl-068-209-159-248.sip.shv.bellsouth.net) joined #forth 18:53:53 --- join: richard__ (~richard@adsl-068-209-159-248.sip.shv.bellsouth.net) joined #forth 18:53:53 hello 18:54:17 hello 18:54:24 hi dick :P 18:54:34 richardsnow.is-a-geek.net 18:54:52 test web server 18:55:04 hi tbw 18:55:07 wathcin a dvd 18:55:08 :) 18:55:26 I got classic rock stuff tape playing 18:56:19 server works 18:57:07 --- quit: snowrichard (Client Quit) 18:58:42 --- quit: richard__ (Client Quit) 18:58:44 --- quit: richard_ ("Leaving") 19:01:42 Microsoft gives you windows, linux gives you the whole house 19:01:45 :) 19:03:35 no...Microsoft sells you an expensive window that is easily breakable, linux gives you an earthquake resistant stone house :) 19:14:24 :)) 19:16:30 lol Tandy Shack 19:16:32 i lol 19:18:23 tandy shack? 19:34:45 --- quit: fridge ("Client exiting") 19:34:59 lol 19:35:04 I'm reading DPANS 19:35:18 and they dedicate an entire chapter on data types 19:35:19 lol 19:35:19 lol 19:35:19 lol 19:36:48 --- join: snowrichard (~richard@adsl-068-209-159-248.sip.shv.bellsouth.net) joined #forth 19:36:51 hello 19:37:39 hiya snowrichard 19:37:54 * kc5tja is back (gone 00:58:15) 19:38:24 the wonderful wiz there was..... 19:38:24 wb kc5tja 19:38:30 how are ou 19:38:31 you 19:38:33 kc5tja: fsincos in one cycle. possible? 19:38:55 With advanced pipelining techniques, yes. 19:39:15 snowrichard: I could be better, but overall, I'm doing OK. 19:39:29 kc5tja: well, its a 3 step pipeline. 19:39:38 http://richardsnow.is-a-geek.net abuse it I want a stress test :) 19:40:16 blog 19:40:44 or bbs or forum ... 19:41:28 I had a good christmas got to spend time with my 13 year old boy 19:41:49 snowrichard: !!!!!! 19:41:52 snowrichard: dude, thats wrong 19:41:54 lol 19:41:55 who is bigger than I 19:42:02 * chrisrw is 5'7" 19:42:03 :( 19:42:17 193 pounds, size 13 feet 19:42:27 yeah riiiight 19:42:36 he is 19:43:00 my son 19:43:03 oh 19:43:04 lol 19:43:11 i thought he was 13 feet tall 19:43:12 lol 19:43:14 * chrisrw has 13W 19:43:27 at about the same pounds 19:43:30 * chrisrw is a chubby kid 19:44:18 so load test my server. :)) 19:44:55 Well, right now, I'm trying to get my ham rig to work with my sound card. 19:45:18 radio with sound card? why not just use shoutcast and broadcast on internet :)) 19:45:33 I would like to have been able to afford ham stuff 19:45:40 when I was kid. 19:45:55 kc5tja: how many pipelines do you think I should have? 19:46:43 I have been dirt poor most of my life. :( 19:47:07 chrisrw: No idea. 19:47:15 poor poor pitiful me ..... lay my head on the railroad track. 19:47:17 Never designed an FPU, never was particularly interested in them. 19:47:25 my idea: have 4, each with its own IP. 19:47:40 snowrichard: I saved up six years for his ham rig. 19:47:56 I bet you did 19:48:12 (Well, I was originally going to get a Yaesu FT-1000MP, but then the Kenwood TS-2000 came out, and that offered a vastly superior price/feature-set ratio) 19:48:46 I know nothing about them but always liked to screw around playing DJ with my tape cassettes 19:49:11 and my dad was in AM radio in 60's 19:50:15 I have set up a shoutcast server it is piece of cake 19:51:10 you could set up a web page that queues requests and have web server control playlist for your ham broadcast 19:52:12 i have 3 sound cards in this machine 19:52:52 video 1 sound and lan on mb 19:53:00 2 cmipci too 20:20:57 --- quit: snowrichard ("Leaving") 20:22:32 --- quit: chrisrw (Read error: 104 (Connection reset by peer)) 20:56:59 * warpzero is away: THIS IS IT 20:58:12 --- join: hovil (~hovil@CommSecureAustPtyLtd.sb1.optus.net.au) joined #forth 21:03:00 --- join: chrisrw (~chris@wbar8.lax1-4-11-099-104.dsl-verizon.net) joined #forth 21:13:50 gotta go ... bye all 21:14:28 --- part: TheBlueWizard left #forth 21:14:29 night The 21:14:31 doh 21:14:34 Oh well. 21:14:47 Well, got the ham rig piping audio into the computer right now. 21:15:03 gmfsk has already been successful at decoding PSK31 and RTTY signals. 21:18:33 * chrisrw is gonna get to net-install netbsd on a sparc20 box tomorrow 21:27:48 * kc5tja nods 21:28:02 * kc5tja is just plain happy that he can use his guitar with his computer -- that is dang sweet. :) 21:28:29 oh 21:28:30 lol 21:28:36 i've been doing that occasionally 21:28:37 its fun. 21:31:25 * kc5tja is downloading Audacity now. 21:32:58 hehe 21:33:01 have fun, my friend. 21:33:14 i think I might just do the same. 21:33:19 well, i already have audacity 21:33:27 unFUCKINGbelievable. 21:33:36 Debian's site is down, SourceForge is down, . . . 21:33:54 I can't build Audacity without wxWindows, and of course, that's only on SourceForge now. >:( 21:34:19 ... 21:34:19 lol 21:34:28 i could shoot ya an RPM or such. 21:34:46 Sourceforge is being such a shit lately 21:34:59 No. I found a source for 2.4.2, even later than the so-called "official" 2.3.2 release. 21:35:25 :) 21:35:33 serisouly 21:35:49 I'm considering ditching sourceforge and going savannah or something. 21:35:57 sourceforge is constantly down 21:36:02 and they still don't have a decent search 21:36:27 host it yourself 21:36:54 I've noticed everybody is getting a lot tighter with hosting 21:38:09 Well, in my case, I can't host it myself, because the web server ports are blocked by the cable company. 21:39:36 i dont want to. 21:39:41 i dont want to pay money. 21:39:50 but savannah would work, i guess? 21:39:59 savannah got hacked recently 21:40:01 I think' 21:40:09 !!!!! 21:40:15 ddddaaaammmm=mmmmmiimtitao'aotg 21:40:21 what the fuck is this workld coming to 21:42:04 Yes, Savannah was rooted. 21:42:12 But they should be back up now. 21:42:45 Yep, back up they are. 21:43:15 its a risk you run when you operate so many services 21:44:56 I personally disagree with that. 21:45:05 When a box is rooted, that implies buggy server software. 21:45:21 would they have full-time salaried staff looking after it? 21:45:55 If such software is written correctly, debugged thoroughly, then root-hacks wouldn't ever occur. 21:46:14 In the case of SourceForge, yes, they DO have salaried people working and looking after the site. 21:46:28 I'm moderately sure GNU does too, though maybe not nearly as many. 21:46:45 But I'm positive that SF does, because they sell SourceForge (the software) as a commercial product. 21:47:08 You can't possibly run a service like Savannah or SourceForge without having full-time staff. 21:47:53 ugh 21:48:00 how the hell did savannah get rooted? 21:48:03 i mean 21:48:13 ssh root should be fuggin disabled 21:48:35 Probably, the folks who rooted Savannah realized, "Gee, Savannah and SourceForge run the same basic software, so . . ." and proceeded to use the exact same hack that they used on GNU. 21:48:55 chrisrw: Gaining root via ssh is not a root hack. 21:49:18 yes, but 21:49:23 isnt that what they did!? 21:49:31 they got the password, then used ssh. 21:49:32 A root hack occurs when you hand-craft a small machine language snip of code, and place it into something like an URL string or some other packet of data. 21:50:14 You then send that packet, which is sufficiently large enough to (hopefully) overflow the stack-allocated buffer of a root-level process. 21:50:39 By making it large enough, you can overwrite the stack frame, including the return address, to point to your own custom code, which can grant you all sorts of God-like powers over that server. 21:51:06 oh 21:51:13 was savannah running a 2.4.21 kernel? 21:51:21 (unpatched, i should say) 21:51:33 They didn't say. 21:51:49 likely 23:11:56 --- quit: kc5tja ("THX QSO ES 73 DE KC5TJA/6 CL ES QRT AR SK") 23:59:59 --- log: ended forth/03.12.30