00:00:00 --- log: started forth/05.01.02 00:15:12 --- join: SeaForth (~SeaForth@c-24-1-126-202.client.comcast.net) joined #forth 00:55:53 --- join: qFox (C00K13S@82-169-140-229-mx.xdsl.tiscali.nl) joined #forth 01:01:39 Hey :) 01:12:58 yo 01:13:00 wazap 01:13:18 --- quit: SeaForth (Read error: 104 (Connection reset by peer)) 01:13:53 hm, #forth is about 1/3rd in size since htel last time i looked 01:13:59 :/ 01:14:01 did war break out? 01:14:07 I think so. 01:14:10 hrm 01:14:13 There was a big colorforth rebellion. 01:14:16 well i'll be gone next week 01:14:20 you dont say 01:14:23 And some people got tired of...some other people 01:14:25 Aww 01:14:28 What will you do? 01:14:40 party 01:14:42 basically 01:14:44 :) 01:15:01 but i wont be home for the week 01:15:34 Heh 01:15:40 Well, have fun then 01:16:03 --- quit: Sonarman (Read error: 110 (Connection timed out)) 01:20:01 i will 01:20:02 :) 01:40:32 --- join: Sonarman (~snofs@adsl-63-196-0-75.dsl.snfc21.pacbell.net) joined #forth 02:41:10 --- quit: Sonarman ("leaving") 03:54:26 --- join: bbls (~bbls@80.97.121.133) joined #forth 04:05:36 Hi 04:05:47 hi Robert 04:08:24 are you using win32? 04:08:46 Not right now, but I have a win98 computer. 04:08:56 ah 04:10:14 http://www.wirednewyork.com/skyscrapers/citigroup/images/citigroup_center_601lexington_18may02.jpg 04:10:23 http://www.thecityreview.com/citi3.gif 04:10:34 i don't know why, but i find this building impressive :) 04:11:47 It's...large 04:12:00 let me search another one 04:12:49 another picture.. 04:15:28 hmm, i can't find it.. 04:15:49 oh 04:15:51 here it is 04:15:59 http://data.greatbuildings.com/gbc/images/cid_2066624.jpg 04:16:07 Robert? 04:16:50 * Robert checks 04:17:07 :) 04:17:41 it is really impressive 04:17:56 Yes, 04:27:05 http://www.colinfahey2.com/2003nov_usa/07_new_york_city_midtown_1368.jpg 04:27:08 same building 04:28:32 Such large buldings are unforthish. 04:28:37 Forth is more like a concrete bunker. 04:28:39 haha :) 04:32:47 well, some people like running a 1kb kernel on a 64kb machine 04:33:02 i preffer running a 4kb kernel on a 4Gb machine :) 04:36:19 I'm thinking about the amount of memory I should build. 04:36:33 Is 16 bits enough? 64 bits could be a little too much I think. 04:37:00 what do you mean? 04:37:20 I'm thinking about starting my computer project by building some memory. 04:37:51 static memory? using flip flops? 04:40:08 Yeah. 04:41:04 Unless you have a better idea 04:41:26 well for memory of this size, this is the best way (and the easiest too) 04:41:28 I'm open to suggestions, as SRAM requires a lot of components. 04:41:48 also the fastest memory 04:41:56 just that is very expensive 04:41:59 :) 04:42:13 but since you need only a few bits, it's ok 04:42:18 Right. 04:42:29 you run into trouble only when you want to go for several Mb of static mem :) 04:42:34 Hehe 04:42:40 I'd prefer to stay off the FETs 05:00:55 --- join: tathi (~josh@pcp01375108pcs.milfrd01.pa.comcast.net) joined #forth 05:02:00 Hi 05:02:09 Hi Robert 05:02:19 What's going on? 05:02:45 Minimalism and large buildings. 05:04:37 ah. I see. 05:04:56 http://www.wirednewyork.com/skyscrapers/citigroup/images/citigroup_center_601lexington_18may02.jpg 05:04:59 http://data.greatbuildings.com/gbc/images/cid_2066624.jpg 05:05:02 http://www.colinfahey2.com/2003nov_usa/07_new_york_city_midtown_1368.jpg 05:05:06 :)) 05:05:15 yeah, I just checked the logs 05:06:23 Robert: how wide of a processor are you going to make? 05:06:25 Hmm... about RTL logic - how does one determine which value of resistors to use? 05:06:50 tathi: Until it works or I realize it's a stupid project. 05:06:58 Which means that it'll be pretty small and simple. 05:07:39 cool 05:07:49 * tathi wants to build a transistor computer one of these days. 05:08:11 I'm young and bored. Seems like a good time. 05:08:20 http://www.play-hookey.com/digital/electronics/rtl_gates.html 05:08:28 Yeah, I'm there already. :) 05:08:36 Nice page 05:08:49 But as far as I can see the value isn't discussed. 05:09:20 Oops, DTL logic was what I meant. 05:09:29 oh, DTL 05:09:53 Should I go as low as possible without melting anything? 05:10:22 hmm, i have a question 05:10:53 what about using \n as word definition terminator instead of ; ? 05:11:11 Forcing one-liners on people are you? ;) 05:11:29 bbls: well...I still find I need definitions longer than one line sometimes. 05:11:39 Robert: I was thinking it should be fairly high. 05:11:44 tathi then you should consider refactoring :) 05:12:00 bbls: no, I mean even once everything is factored as far as makes sense. 05:12:04 also using a non-standard character for definition starting 05:12:17 like some highly visible ascii char 05:12:28 tathi: Why? 05:13:28 in diode logic, the resistor is just to hold it in a stable state when it's "off", right? 05:13:52 so...when your inputs force it to switch to the other state, then the lower the resistor, the more current draw you're going to have 05:14:46 But when we're switching in the other direction, won't a high value resistor slow down? 05:14:56 I.e. when the transistor starts conducting again. 05:15:55 that I don't know. 05:25:37 * tathi really needs to learn more electronics theory 05:25:43 Hehe 05:25:54 I just found that info on another page 05:26:02 (still on play-hookey) 05:26:07 http://www.play-hookey.com/digital/experiments/dtl_inverter.html 05:45:13 ah. 05:58:23 --- topic: set to 'Forth: Two stacks, a dictionary of word names and their definitions.' by crc 05:58:47 I demand the right to an extra stack! 05:58:59 Write one. 05:59:45 Or build one into your CPU :) 05:59:54 Heh. 06:00:05 Nah, stacks are for losers. 06:00:30 * crc wonders if he should kick robert out for blasphemy... 06:01:29 :D 06:02:49 * crc is wondering why most forths use \ as the comment character 06:03:11 Because they hate Swedes. 06:03:23 Our kb layout makes \ a bitch to type. 06:03:35 Not as bad as $, though. 06:03:41 is | easier for you to type? 06:03:57 No, that's even worse. 06:04:00 (I use | as the comment char in retroforth) 06:04:01 hmm 06:04:16 get a better keyboard layout. 06:04:18 :D 06:04:21 Thanks. ;) 06:04:52 Robert why don't you use an US kb? 06:05:04 I'd recommend dvorak or qwerty (in that order) 06:05:05 Because I suck. 06:05:18 (disclaimer: I still use qwerty most of the time...) 06:05:22 And once in a while I use Swedish. 06:05:31 here in romania there were several trials to introduce localised keyboards, but all of them failed. people simply preffer to use a standard keyboard 06:05:42 I tried dvorak for an hour before I decided that I'll never learn how to type. 06:05:54 then try qwerty :) 06:06:04 bbls: Here you can't even get US keyboards 06:06:10 crc: I'm on qwerty. 06:06:15 heh 06:06:30 \ isn't a pain to type on my qwerty keyboard 06:06:42 Not on a US qwerty kb, no. 06:06:43 | is just as easy ;) 06:06:46 hmm 06:07:04 Use a US qwerty keymap then 06:07:12 Robert maybe your language is not easly read when is written in english alphabet 06:07:14 I could. 06:07:29 bbls: It works, but doesn't look very good. 06:07:33 and switch to a swedish one when you need to use swedish 06:08:04 Uh 06:08:11 Are you serious? 06:08:12 well, that might be one good reason, romanian is perfectly readable with english alphabet (i hardly see any computer text using romanian alphabet) 06:08:30 Perhaps if I could switch very quickly. 06:08:35 Hmm 06:08:40 Because I switch between English and Swedish all the time. 06:08:47 * crc used to switch quickly... 06:09:05 I haven't explored keymaps under Linux yet though 06:09:22 I could probably set up some easy switch 06:09:31 But it wouldn't matter much. 06:09:54 I still can't type. Especially not on this laptop 06:10:35 use an external keyboard 06:11:23 That's not so easy 06:11:33 why not? 06:11:44 No room here 06:11:47 hmm 06:12:02 And I should use the other box more. 06:12:05 sit the keyboard on top of the laptop keyboard :) 06:12:10 Heh 06:12:11 or use the other box 06:26:44 --- join: snowrichard (~richard@adsl-69-155-177-153.dsl.lgvwtx.swbell.net) joined #forth 06:32:48 bbl 06:32:55 bye 06:33:10 Hi snowrichard 06:33:14 And bye crc 06:35:30 I spent new years eve at my mom's and we stayed up till 2:00 AM watching "The Green Mile" 06:36:15 Nice movie 06:36:40 they got a dvd player for christmas 06:37:39 I hooked it up for them 06:38:24 there was 3 years worth of dust on the top of the tv that my mom wiped off. :) 06:38:34 We got a DVD player last christmas, but we're not really using it. We have a very large video tape collection (and an ancient - 15 years or so) VCR 06:38:40 Hah 06:39:01 had to move the tv out of the cabinet to get to the rear jacks. 06:41:21 I just bought a vcr 06:42:20 * Robert watches a lot less TV these days 06:43:17 I have the dish but lots of the channels are pretty useless. I mean how many shopping channels does one need? 06:45:35 Hehe, right 06:47:15 I got CD ripping and burning added to my gui 06:47:58 What do you use for burning? 06:48:03 k3b 06:48:30 its the default on kde from mandrake and it works 06:52:39 Ah, I use that too. So how does your GUI interface that? 06:53:08 I just have a button that says "burn CD" and it runs system("k3b & "); 06:54:18 Hehe, OK. Probably not worth duplicating their code 06:54:56 I don't know how to use any of the console CD burning programs, but one idea would be to use one of those. 06:55:05 I might get more advanced when actually have track data like title and artist. CDRDAO can put that stuff on the CD 06:55:19 cdrdao? 06:55:22 a .toc file 06:57:44 it is used to make audio cd's (referenced in CD-Burning howto 06:58:26 Don't make many of these anyway...one gets spoiled by the number of hours a HD with mp3s can take 06:59:07 If I blow away Windows I have 20 GB on that hard drive. 06:59:34 /dev/hda5 31G 30G 648M 98% /home 06:59:41 That is the sad story of my HD 06:59:57 almost full 07:00:09 Yeah...no matter how much I remove it always is 07:00:19 --- nick: Robert -> lscd 07:12:42 my sister wants me to take her to church so I'll be back later 07:14:06 OK, bye! 07:20:04 --- quit: snowrichard ("Leaving") 07:28:13 --- join: Topaz (~top@spc1-horn1-6-0-cust217.cosh.broadband.ntl.com) joined #forth 07:29:14 Hi 09:29:54 --- quit: Topaz (Remote closed the connection) 09:30:51 --- join: Topaz (~top@spc1-horn1-6-0-cust217.cosh.broadband.ntl.com) joined #forth 09:31:52 back 09:32:00 wb 09:32:23 hmm, there is a site with free photos 09:32:25 * crc sees that he didn't miss much :) 09:32:36 but i don't remember the name 09:33:53 ah, this is it http://www.sxc.hu/ 09:39:53 --- join: T0paz (~top@spc1-horn1-6-0-cust117.cosh.broadband.ntl.com) joined #forth 09:42:08 --- quit: saon (Remote closed the connection) 09:52:48 --- join: saon (1000@c-24-129-90-197.se.client2.attbi.com) joined #forth 09:56:44 --- quit: Topaz (Read error: 110 (Connection timed out)) 10:36:42 --- join: bbls1 (~bbls@80.97.121.133) joined #forth 10:56:54 --- quit: bbls (Read error: 110 (Connection timed out)) 11:17:50 --- join: Teratogen (~leontopod@intertwingled.net) joined #forth 12:22:31 --- quit: tathi ("Lost terminal") 12:23:07 --- join: tathi (~josh@68.81.165.175) joined #forth 12:25:16 --- join: slava (~slava@CPE00096ba44261-CM000e5cdfda14.cpe.net.cable.rogers.com) joined #forth 12:25:28 does anybody have that link that gives a quick rundown of the most important 386 instructions in octal? 12:27:11 http://tom.bespin.org/src <-- there's one or two on there that I use 12:27:29 http://tom.bespin.org/src/opc386.txt 12:27:51 http://tom.bespin.org/src/opcodes.html 12:28:01 Those are the two that I use most of the time 12:28:02 thanks 12:28:07 np 12:29:12 * slava wonders if BOUNDS is faster for array bounds checking than manual index tests 12:29:32 no idea there 12:29:38 * crc doesn't obsess over speed 12:35:34 slava that's what i wondered too :) it might be very CPU dependant 12:35:58 i know that LOOP is slower on pentium4 than DEC/JNZ because LOOP is micro-coded 12:37:05 Yes, when you're coding for P4, forget common sense. 12:37:14 Forget anything else that's common or sensible. 12:37:29 heh 12:37:32 well i'm coding for any x86 really 12:37:44 i'm not worried about instruction scheduling and such yet, i just want to generate code that is not completely absurd 12:37:48 Which means pentium+ these days anyway? 12:37:53 :) 12:37:54 yeah 12:38:17 Well, except for freaks like me and a large part of the Forth community. 12:38:29 i don't plan on using any pentium-specific opcodes 12:38:44 CMOVE and such might be handy, but they're not critical 12:38:55 i will require a 387 though 12:39:14 what's the first cpu with cmove? 12:39:23 ppro 12:41:36 i might directly jump and require SSE in my case :) 12:41:52 387 fp stack is convinient for me 12:41:59 sse would require a register allocator 12:42:20 which i plan eventually, since stacks are too slow, but not just yet 12:54:02 --- join: Raystm2 (~Ray@adsl-69-149-51-211.dsl.rcsntx.swbell.net) joined #forth 14:38:26 --- join: TheBlueWizard (TheBlueWiz@modem-137.nyc-tc03a.fcc.net) joined #forth 14:39:01 --- join: Herkamire (~jason@h000094d30ba2.ne.client2.attbi.com) joined #forth 14:51:22 --- join: Sonarman (~snofs@adsl-209-233-52-1.dsl.snfc21.pacbell.net) joined #forth 16:04:35 --- join: p-Imperator (CapStone@cs24160141-160.satx.rr.com) joined #forth 16:04:35 --- quit: Testament (Read error: 104 (Connection reset by peer)) 16:22:57 --- quit: Sonarman (Read error: 110 (Connection timed out)) 16:50:38 --- quit: crc (Remote closed the connection) 17:02:44 --- join: snowrichard (~richard@adsl-69-155-177-153.dsl.lgvwtx.swbell.net) joined #forth 17:06:45 --- quit: Raystm2 ("User pushed the X - because it's Xtra, baby") 17:20:22 * swsch is away: going ... going ... gone. 17:23:31 --- quit: skylan (Remote closed the connection) 17:24:35 --- join: skylan (~sjh@dialup-216-211-5-21.tbaytel.net) joined #forth 17:36:35 --- quit: T0paz (Remote closed the connection) 17:36:39 --- join: aum (~aum@60.234.138.239) joined #forth 17:40:03 yee-haaa, got JH locals working on EPICforth 17:40:31 no more brain-mutilating stack gymnastics 17:40:43 now you're just doing brain-mutilating local gymnastics! 17:40:52 try to write your code so taht stack manipulation is minimized 17:41:28 when you're manipulating more than 3 items at a time, it's pretty difficult 17:41:39 that's why you don't manipulate more than 3 items at a time 17:41:44 that is, without using some kind of off-stack storage 17:41:44 I just make objects if I have complex data 17:41:59 which oo impl are you using? 17:42:11 objects.fs? mini-oof.fs? oof.fs? 17:42:15 no 17:42:27 i'm developing factor.sf.net, which is OO at all levels 17:42:43 * aum takes a peek 17:42:47 so everything on the stack is an object? 17:42:59 yes, and knows its type at runtime 17:43:06 integers are not heap allocated though 17:43:10 there's a trick to get them on the stack directly 17:43:11 gc? 17:43:16 yup 17:43:34 refcounting, or true java-style gc? 17:44:06 true gc, yes. java style? hmm... :) 17:45:16 * aum speculates on the wild drunken party that ended with forth, joy, lisp and python ending up in bed together 17:46:10 --- join: swsch_ (~stefan@swsch.sustaining.supporter.pdpc) joined #forth 17:46:36 i'm happy with my locals implementation - efficiently uses 'return stack frames' 17:48:40 --- quit: swsch (Read error: 110 (Connection timed out)) 17:49:37 what's wrong? my code worked first time, i'm not used to that 17:49:57 must have wrote it in my head as i slept last night 17:54:37 purists be damned, locals rule 17:56:19 --- join: Sonarman (~snofs@adsl-64-160-166-189.dsl.snfc21.pacbell.net) joined #forth 17:59:34 aum, i think objects allow for easier factoring than locals 17:59:47 locals tend to encourage long words 18:00:05 yes, but EPIC forth has savage RAM limitation 18:00:17 yeah 18:00:32 runs on PIC16/PIC18, and (apart from stacks) is limited to 180 bytes sram 18:01:31 w00t! google and ye shall find - just downloaded 2 C-to-forth translators 18:01:38 heh 18:02:26 takes me 4x as long to write something in C compared to python, also takes me 4x as long to write in forth compared to C 18:03:31 forth is easier than C IMHO 18:03:33 --- quit: bbls1 (Read error: 60 (Operation timed out)) 18:03:46 depends on the task 18:04:00 for simple bitbashing, forth is easier 18:04:24 but for bigger stuff - manipulating complex data structures - i find C far easier, and python far easier again 18:04:31 GC helps 18:04:36 C is a pain 18:04:53 C has a perfectly good gc - it's called 'fred()' :P 18:05:00 sorry, 'free()' 18:05:03 heh 18:05:28 i'm amazed at how many C programmers dislike gc, then go off and write a refcounting implementation in their app 18:05:38 its really amusing that most don't realize a good generational GC is 10000x faster than refcounting or malloc/free 18:06:24 factor doesn't have generational GC yet, so a full collection of the heap takes 5ms or so 18:06:38 with generational GC it becomes virtually free 18:06:42 --- join: crc (~crc@pool-70-20-243-55.phil.east.verizon.net) joined #forth 18:06:49 --- mode: ChanServ set +o crc 18:06:57 hi crc 18:07:01 with generational GC, object allocation is a few cycles, and deallocation is free 18:07:24 hi aum 18:07:29 refcounting is not too bad, at least in python 18:07:34 GC uses more memory 18:07:44 because there is always some amount of garbage floating around 18:07:54 aum, python is pretty slow though, and its memory amanger is one reason 18:08:07 i do like refcounting, it's simple and sensible 18:08:31 it doesn't work with cycles and has high overhead though 18:08:38 i agree taht generational GC is not simple at all 18:08:41 slava: my answer to that is that python's C api is awesome, allows easy recoding of time-critical stuff in C 18:08:58 i prefer not recoding time-critical stuff in C, though. that's what a compiler is for ;) 18:09:01 psyco looks promising 18:09:29 psyco is intermittently good - sometimes gives 50x speedups, on occasion it can actually slow things down 18:09:44 heh 18:09:53 my favourite is pyrex - a python/C blend language with the best features of both langs 18:09:58 --- quit: tathi ("leaving") 18:24:01 bye all 18:24:10 --- part: TheBlueWizard left #forth 18:30:57 --- quit: crc (Nick collision from services.) 18:31:16 --- join: crc (~crc@pool-70-20-136-183.phil.east.verizon.net) joined #forth 18:53:57 goodnight 18:54:04 --- mode: ChanServ set +o crc 19:21:49 --- mode: ChanServ set +l 83 19:39:01 --- join: Sonarman_ (~snofs@adsl-64-169-92-228.dsl.snfc21.pacbell.net) joined #forth 19:41:54 --- quit: Sonarman (Read error: 104 (Connection reset by peer)) 20:02:00 --- part: slava left #forth 20:14:12 --- quit: Sonarman_ (Read error: 110 (Connection timed out)) 20:30:04 --- quit: skylan (Remote closed the connection) 20:30:18 --- join: skylan (~sjh@dialup-216-211-5-21.tbaytel.net) joined #forth 20:43:03 --- join: Sonarman (~snofs@adsl-64-171-254-236.dsl.snfc21.pacbell.net) joined #forth 20:55:14 --- join: adu (~adu@pcp04424087pcs.nrockv01.md.comcast.net) joined #forth 20:59:53 --- part: adu left #forth 21:05:44 --- nick: lscd -> Robert 21:07:03 --- quit: yumehito (Read error: 110 (Connection timed out)) 21:54:09 the worst disadvantage of locals is that they tend to make code readable 21:54:46 --- quit: aum () 22:02:17 Haha 22:02:37 What a summay of the Forth spirit. 22:13:32 --- join: Testament (CapStone@cs24160141-160.satx.rr.com) joined #forth 22:14:30 Hi 22:16:27 heh 22:17:09 the worst part about locals, is that they are usually used to make a tangled mess a bit more legible, instead of factoring properly 22:25:45 --- quit: Sonarman ("leaving") 22:30:56 --- quit: p-Imperator (Read error: 110 (Connection timed out)) 22:31:12 --- join: p-Imperator (CapStone@cs24160141-160.satx.rr.com) joined #forth 22:47:32 --- join: yumehito (yumehito@giguz.turbo.nsk.su) joined #forth 22:47:43 Hi 22:48:25 --- quit: Testament (Read error: 110 (Connection timed out)) 23:02:39 --- quit: Herkamire ("bed") 23:54:31 --- join: swsch (~stefan@swsch.sustaining.supporter.pdpc) joined #forth 23:59:59 --- log: ended forth/05.01.02