00:00:00 --- log: started forth/02.08.16
02:07:58 --- join: Speuler (~l@p3E9B8EB9.dip.t-dialin.net) joined #forth
02:08:16 <Speuler> 'morning
02:09:52 <onetom> hi
02:10:11 <Speuler> hi 1tom
02:10:26 <Speuler> how's your bot doing ?
02:14:53 <Robert> Morning 1tom, Speuler. )
02:14:59 --- join: Soap` (~flop@202-0-42-22.cable.paradise.net.nz) joined #forth
02:15:11 <Speuler> god dag the_rob
02:15:25 <Robert> :)
02:15:40 <Speuler> good mood today ?
02:15:48 <Robert> Just woke up :P
02:15:57 <Robert> But yeah, I'm always in a good mood when I wake up.
02:18:06 * Speuler will turn to CaffeineJunkie mode shortly
02:20:29 <Speuler> also only awake for half an hour or so
02:21:12 <Robert> :)
02:21:20 * Robert will go eat a bit.
02:25:35 * Speuler will go beat it
02:26:07 --- nick: Speuler -> CaffeineJunkie
02:45:39 --- quit: CaffeineJunkie (Operation timed out)
02:59:22 --- quit: futhin (Ping timeout: 14400 seconds)
03:21:14 --- quit: sif (sendak.openprojects.net irc.openprojects.net)
03:21:16 <onetom> speuler: dont have time 2 have an eye on the bot
03:21:19 --- join: sif (~sifforth@ip68-14-9-225.ri.ri.cox.net) joined #forth
03:22:18 <onetom> speuler: so ive stopped it. uknow our netconnection is so kludgy.
03:23:21 <onetom> and if it parts from the net the bot eats up all the cpu cycles
03:52:39 --- quit: sif (sendak.openprojects.net irc.openprojects.net)
03:52:56 --- join: sif (~sifforth@ip68-14-9-225.ri.ri.cox.net) joined #forth
04:25:26 --- quit: Soap` (Read error: 104 (Connection reset by peer))
05:40:12 --- join: Serg_penguin (~Z@nat-ch1.nat.comex.ru) joined #forth
05:40:26 --- quit: Serg_penguin (Client Quit)
05:58:16 --- join: cleverdra (julianf@0-1pool37-37.nas2.florence1.sc.us.da.qwest.net) joined #forth
06:57:51 --- join: Serg_penguin (~Z@217.10.35.226) joined #forth
06:58:39 --- quit: Serg_penguin (Client Quit)
08:49:58 <onetom> hi cleverdra 
08:50:15 <onetom> any news?
08:51:02 <cleverdra> I've settled on a Scheme, I think.  You?
09:08:57 <onetom> A Scheme? cant understand that A
09:09:44 <onetom> not much, coz i have 2 do the usual sysadmin tasks
09:09:58 <onetom> seting up some backup mechanism
09:09:58 * cleverdra doesn't understand "can't understand that A"
09:10:25 <onetom> writing some picture slicer script & the like
09:10:52 <cleverdra> Slice pictures how?
09:11:09 <onetom> ah, and the most boring stuff is playin ping-pong w huge movie files, recompress them & write them 2 cd
09:11:34 <onetom> slice just along a grid into small pieces
09:11:52 <onetom> i havent started it yet, but probably its not 
09:12:01 <onetom> as ez as i thought 4 the first time...
09:12:08 <cleverdra> So what did you mean by "cant understand that A"?
09:12:30 <onetom> coz the orig picture is a ~2000x2000pixels 1
09:12:46 <onetom> Scheme is a lang, right?
09:13:07 <onetom> then what does it mean: a Scheme?
09:13:25 <onetom> A SPECIFIC Scheme?
09:13:30 <onetom> A SPECIFIC Scheme implementation?
09:13:45 <cleverdra> Yes, of course.
09:14:07 <cleverdra> Well, yes anyway.  Like 'a Forth'.
09:14:16 <cleverdra> I have a Forth named IsForth =)\
09:14:30 <onetom> ahha :)
09:15:09 <onetom>  but u havent said any more specific after
09:15:27 <onetom> like A Scheme VERSION/IMPLEMENTATION or the like
09:15:44 <onetom> it has confused meg a bit :)
09:15:49 <onetom> s/meg/me
09:18:34 <onetom> on the other hand i still researching the possibilities of a forth os
09:19:29 <onetom> eg, we have started an exploration w tcn on the topic "4th-vs-C execution cost"
09:20:27 <onetom> but we still have 2 find a good example for proving 1is better that the other
09:20:38 <cleverdra> What does proving that have to do with a Forth OS?
09:21:47 <onetom> it could b an undoubtable prove that a 4thOS would b more effective (and so faster on the proper hw) than a C
09:22:08 --- join: futhin (~thin@h24-64-175-61.cg.shawcable.net) joined #forth
09:22:36 --- quit: futhin (Client Quit)
09:22:38 <onetom> so not only the different ideology/methodology of forth makes the difference between the two
09:22:41 <onetom> brb
09:23:18 * cleverdra shrugs.
09:50:21 --- join: I440r (~mark4@1Cust231.tnt1.bloomington.in.da.uu.net) joined #forth
10:03:23 <Robert> Hey :)
10:04:45 <I440r> hi
10:45:01 --- join: proteusguy (~irc@24-197-147-197.charterga.net) joined #forth
10:51:22 --- join: futhin (thin@h24-64-175-61.cg.shawcable.net) joined #forth
10:51:36 <futhin> onetom: you there?
10:52:50 <I440r> not all :)
10:54:53 <futhin> i440r: are you spending lots of time with isforth or is your job keeping you busy?
10:55:54 <I440r> no my job is semi finished, on hold till next month
10:56:20 <I440r> im plodding my way thru reorganizing the opcode tables into a LOGICAL order
10:56:31 <I440r> instead of having similar encoding scattered all over the place
10:57:50 <futhin> are you happy that your job is on hold? :P
10:57:56 <futhin> so that you can spend quality time with isforth
10:58:00 --- join: proteus (~irc@24-197-147-197.charterga.net) joined #forth
10:59:15 <Robert> What kind of logical order?
10:59:15 <Robert> Most commonly used instuctions first, to gain speed?
10:59:15 * Robert thinks he's lagged.
10:59:36 * futhin wonders if there are any instances where recursive algorithms are faster than iterative? and how do they compare on x86 and forthchips?
11:00:29 <futhin> robert: yeah, you might be lagged.. i'll let you know when my ping gets back
11:00:57 <futhin> or maybe they were swallowed up into the abyss!
11:01:04 <I440r> robert no. all like opcoedes together instead of scattered all over the place
11:01:18 <I440r> makes it simple for me to create a function to actually assemble LIKE opcodes
11:03:25 <cleverdra> futhin - well, what do you mean by recursive?  Recursive algorithms do not necessarily create recursive processes.  Algorithms that generate linear recursive processes tend to be, anyway, easier to create and slower to run -- see Fibonacci numbers for a classic example.
11:04:01 <futhin> okay, well then s/algorith/proccess
11:04:03 <I440r> futhin ive NEvER seen a recursive fucntion that couldnt be coded better as an itterative function (ackerman doesnt count, it has NO useful purpose)
11:04:30 <futhin> are recursive functions less costly on a forthchip because of the return stack ?
11:04:49 <futhin> do iterative functions still beat out recursive functions on a forthchip
11:05:03 <Robert> Why waste return stack space?
11:05:07 <futhin> why do people code recursive functions?  lazy bums
11:05:16 <cleverdra> futhin - an iterative process would tend to beat a recursive process, because of what Robert just said.
11:05:54 <cleverdra> futhin - not just stack space, but space.  Iterative processes work in fixed memory.
11:06:44 <futhin> why does anybody support recursive functions? :/
11:07:55 <cleverdra> futhin - because, again, 'recursive function' ne 'recursive process'; because recursion is useful, and sometimes less-costly for special cases -- say, one of the troff state-machines that I was going to send you.
11:08:03 <cleverdra> futhin - BTW, what's your email?
11:08:51 <futhin> cleverdra:  jbrobey@yahoo.com
11:08:52 <cleverdra> um, gcd in Forth, if you've seen it, is an nice recursive function that generates an iterative process.
11:09:07 <Robert> Hrm.. How would I solve a = b^x, when I know a & b but not x?
11:09:13 <cleverdra> Of course you need optimized tail-recursion.
11:09:19 --- quit: proteusguy (Read error: 110 (Connection timed out))
11:09:25 <Robert> er
11:09:30 <cleverdra> Robert - logarithms
11:09:33 <Robert> a = b^kx
11:09:38 <Robert> Well, I can't get it to work.
11:09:38 <futhin> cleverdra:  do i recall you asking somebody on here for my email address a short while ago for sending me something else ?
11:09:59 <Robert> I know a and b (a is below b), and a = b^8k
11:10:00 <cleverdra> ln(a) = ln(b^kx) = ln(b)*ln(kx)  ... or something like that.
11:10:06 <Robert> So, k is negative.
11:10:15 <Robert> Hrm...
11:10:18 <cleverdra> futhin - yes.
11:10:27 <Robert> I though ln(b) * 8k.
11:10:59 <cleverdra> Robert - eventually you reach a point where you raise to a power of e, to 'undo' the ln() and get an answer.
11:11:22 <cleverdra> ln(ax)=ln(a) + ln(x)
11:11:35 <cleverdra> ln(a^x) = x*ln(a) .. ah, I'm remembering.
11:11:40 <Robert> Yeah.
11:11:42 <Robert> So...
11:11:50 <Robert> ln(a) and ln(b) are both negative.
11:11:59 <futhin> that's correct
11:12:09 <Robert> So, ln(a) = 8*k*ln(b)
11:12:15 <futhin> yeah
11:12:17 <cleverdra> ln(a) = ln(b^kx) = ln(b)*kx => x = ln(a)/(ln(b)*k) -- which you can compute
11:12:19 <Robert> But k has to be negative.
11:12:37 <Robert> Hmm...
11:12:46 <cleverdra> oh, you don't know k either?
11:12:55 <Robert> Well, I know x, but not k.
11:12:59 <cleverdra> Oh.
11:12:59 <Robert> Well, same thing.
11:13:02 * cleverdra nods.
11:13:06 <Robert> Anyway, x is 8.
11:13:11 <Robert> I don't know k.
11:13:16 <Robert> ln(a) and ln(b) are both negative.
11:13:16 <cleverdra> k = ln(a)/(ln(b)*x) -- get a calculator =)
11:13:32 <Robert> ln(a)/(ln(b)*8) <--- that would get positive.
11:13:39 <Robert> But k should be negative.
11:13:45 * cleverdra nods.
11:13:46 <Robert> a = f(8)
11:13:49 <Robert> b = f(0)
11:13:53 <cleverdra> Why should it be negative?
11:14:15 <proteus> none of this is RPN - I'm ashamed of you all...
11:14:18 <Robert> f(x) = b^kx
11:14:19 <cleverdra> You said that   a = b^kx, and that k must be negative.
11:14:35 <cleverdra> proteus - because RPN is stupid, here.  When I do the calculation I'll do it on my RPN calc.
11:14:36 <futhin> a ln b ln x * /
11:14:38 <Robert> Yeah, since a is below b.
11:14:43 <proteus> :-)
11:14:47 <futhin> :P
11:15:01 <cleverdra> Robert - kx could be fractional
11:15:08 <futhin> blame the education system for brainwashing cleverdra and robert ;)
11:15:08 <Robert> If k was positive, a should be above b, right? Since a = f(8), and b = f(0), and f(x) = b^kx
11:15:19 <Robert> proteus: RPN?
11:15:30 <Robert> cleverdra: Fractional? :)
11:15:34 <cleverdra> Mathematics is more concisely expressed in infix syntax.
11:15:36 <proteus> Reverse Polish Notation - the forth way Robert...
11:15:42 <Robert> proteus: Ah... Heh.
11:15:43 <cleverdra> Robert - you know... square root of x is x^(1/2)
11:15:45 <futhin> RPN is the natural way to do math! :P
11:15:53 <proteus> it is indeed.
11:15:55 <cleverdra> futhin - to *do* it, maybe, but not to express it.
11:16:08 <futhin> :P
11:16:11 <Robert> cleverdra: Well, hrm.
11:16:22 <Robert> I'm dumb.
11:16:23 <Robert> Well, brb.
11:16:29 <proteus> cleverdra - depends on how deep your brain's stack is... mine's pretty shallow I'm afraid...
11:16:31 * Robert understands 'food'.
11:16:39 <Robert> cleverdra: And yeah, it is fractional.
11:17:36 --- quit: proteus ()
11:17:38 --- join: proteusguy (~irc@24-197-147-197.charterga.net) joined #forth
11:17:55 <cleverdra> Robert - curiously, how did you come upon this problem?
11:18:06 <futhin> yay proteusguy! your internet connection triumphs again!
11:18:11 <proteusguy> yep...
11:18:17 <proteusguy> bites
11:18:49 <futhin> proteusguy: you could install linux :P
11:18:58 <futhin> and use ircii  or xchat in x win
11:20:30 <cleverdra> Or connect to futhin's machine and use GNU screen with ircii
11:20:38 <proteusguy> I have linux in dual boot on this box... just in something that requires Windows at this point. I do have cygwin running however...
11:20:47 <futhin> cleverdra:  heheh, i don't have linux installed on my comp currently
11:21:08 <futhin> but i've actually been thinking of switching because nobody else is using my computer these days
11:21:21 * futhin is reading ESR's cathedral bazaar essay right now.. interesting :)
11:21:47 <cleverdra> Funny thing, I am snarfing ESR's writings right now.
11:22:00 <cleverdra> I read CatB a long time ago, but I'd like a refresher.
11:22:04 <futhin> proteusguy:  you use something that requires windows?  games?  watching videos? 
11:22:16 <proteusguy> ESR is a python guy... which is a cool language. I like his thoughts.
11:22:43 --- join: Kitanin (~clark@h-209-91-66-234.gen.cadvision.com) joined #forth
11:22:48 <proteusguy> futhin - nope - no games or videos (can watch DVDs under linux better)... writing C++ code right now that has win32 target.
11:23:17 <cleverdra> For a counter-Python view, see http://www.paulgraham.com/power.html "Succinctness is Power"
11:23:31 <proteusguy> checking it out now...
11:24:27 <futhin> if succintness is power, then forth is the most powerful language! :P
11:24:30 <futhin> what the hell
11:24:48 <futhin> if all coders were thinking "succintness is power" then wouldn't more look at forth? 
11:24:58 <futhin> forth is not just _succint_
11:25:18 <futhin> it also encourages succint (sp?) solutions
11:25:25 <cleverdra> futhin - more might build large languages like Perl, or more specialized languages that are very concise for a particular field of tasks.
11:25:34 <futhin> s/succint/succinct/
11:25:52 --- join: kc5tja (~kc5tja@user-24-214-86-42.knology.net) joined #forth
11:25:54 <futhin> cleverdra:  forth is easily specializable tho :P
11:26:08 <cleverdra> I describe that URL as counter-Python because it goes against popular Python rhetoric, not because it's counter-Python, even though it starts off with Python =)
11:26:10 <cleverdra> futhin - true.
11:27:08 <cleverdra> (well, not really goes against...  Nevermind.)
11:27:37 <cleverdra> futhin - sometimes Forth can be so specializable that you find yourself almost programming in another language =)
11:27:47 <futhin> yes, exactly
11:29:03 <futhin> this article has made me realize why i think that forth as a high level scripting/general purpose lang of an os would be very nice and hopefully easily usable by non-programmers
11:29:23 <futhin> WHY..  the potential of forth to be succinct..
11:29:39 <cleverdra> kc5tja, if you're curious:  http://www.paulgraham.com/power.html "Succinctness of Power"
11:29:45 <cleverdra> also, hello kc5tja
11:29:56 <futhin> i should probably finish the article before saying anything more heh
11:30:00 <proteusguy> The article focuses solely on language and ignores environment. I find python very succinct for the applications I use it for. I like C++, Python, and FORTH very much. No one pays me for forth however...
11:30:55 <futhin> pay yourself for forth
11:31:11 <futhin> you'll get rich!
11:31:14 <proteusguy> I intend to some day - maybe with a cool PDA project.
11:31:19 <kc5tja> Paul is pretty much blinded by Lisp.  While he's an intelligent writer and knows his stuff, he refuses to acknowledge the existance of any other language besides Lisp.  I'd wager he feels threatened by Forth.
11:31:35 <futhin> i'm blinded by forth
11:31:37 <futhin> i freely admit that
11:31:52 <futhin> i have grown increasingly bigotted against C coders and any other language
11:31:56 <cleverdra> proteusguy - certainly.  a library called 'a' that had a function called 'b' that, when invoked, ran a production server and juggled cakes would give Python a concise way to say "Run a production server and juggle cakes."
11:32:00 <proteusguy> I've found lisp to be a generally useless language when it comes down to getting serious work done. 
11:32:04 <futhin> i still have _some_ respect for lisp
11:32:17 <cleverdra> proteusguy - do you mean Scheme or Lisp?
11:32:29 <cleverdra> proteusguy - and how familiar are you with Common Lisp propaganda?
11:32:43 <proteusguy> Well both - a little more than I wanted to be.
11:33:00 <cleverdra> kc5tja - oh, I got a response from Paul.  Do you want me to forward it to you?
11:33:06 <futhin> proteusguy:  lisp has a fast development cylce like forth, and also has more high-level features and libraries and stuff
11:33:10 <proteusguy> I've done lisp development in the past... always have to end up using another language to get the final bits done.
11:33:12 <futhin> cleverdra: what kind of response ?
11:33:26 <futhin> proteusguy: what kind of projects did you use lisp on?
11:33:28 <cleverdra> Common Lisp is a huge language; I wonder how it could be generally useless.
11:33:31 <kc5tja> If you wish.  sfalvo at hifn dot com
11:33:39 <futhin> jbrobey@yahoo.com
11:33:47 <proteusguy> futhin - you mean its got a neat development environment. 
11:34:01 <cleverdra> futhin - an explanation.
11:34:01 <cleverdra> kc5tja - OK.
11:34:39 <proteusguy> It's useless in that it doesn't map well to the abstractions that my designs create. Takes too much effort for what I could accomplish in many other languages.
11:34:40 <cleverdra> also, futhin, I'll send you that tiny (unfinished, but it allowed me to read some simple manpages) troff with its example state-machines
11:34:40 * futhin doesn't get much spam at all at his email account.. clog's logs have robots.txt so google doesn't surf clog's logs
11:35:15 <cleverdra> proteusguy - well, OK.
11:36:36 <futhin> cleverdra:  what kind of response did you get from paul  ?  or forward it to me please, i'm interested
11:37:10 <cleverdra> yeah, I just sent it to both of you.
11:38:07 <kc5tja> Personally, I find that response unacceptable.
11:38:09 <Robert> Hrm.. what is ln(a*b^(c*d)) ?
11:38:10 <cleverdra> (futhin - this is for gforth, and its state-machines are implemented using CASE structures.)
11:38:21 <cleverdra> kc5tja - how so?
11:38:22 <kc5tja> But he does have a point.  His code is, to a limited extent, polymorphic.
11:38:40 <cleverdra> Robert - c*d*[ln(a)+ln(b)]
11:38:47 <kc5tja> But he very clearly implied integers in his code.  At least, that's how I understood it.
11:39:09 <Robert> cleverdra: What do the [ and ] mean?
11:39:20 <cleverdra> kc5tja - perhaps he did.  At the top of his page he (now) says that he means 'numbers', not 'integers'
11:39:21 <kc5tja> Outter parentheses
11:39:27 <Robert> Oh.
11:39:29 <Robert> Just for clearity?
11:39:32 <Robert> Uhm
11:39:33 <cleverdra> Robert - what kc5tja said.  Yeah, for clarity.
11:39:34 <kc5tja> Yes.
11:39:36 <Robert> clearness* 
11:39:37 <Robert> Or something.
11:39:42 <Robert> clarity. :)
11:39:43 <futhin> forth doesn't do polymorphism ???  surely it does
11:39:43 <Robert> Thanks.
11:39:46 <kc5tja> cleverdra - I see.  
11:40:02 <kc5tja> Forth doesn't do polymorphism.  Everything is bound as early as possible.
11:41:29 <cleverdra> futhin - OK, sent the troff.
11:42:18 <cleverdra> futhin - note the RECURSE in the second state-machine =)
11:43:18 <futhin> cleverdra:  does your smudcl do anything yet?  like connect to another server?
11:43:44 <futhin> kc5tja:  forth _could_ do polymorphism right?   "coding in forth is creating a new language"
11:43:51 <cleverdra> futhin - hm?  I stopped working on that.
11:44:15 <futhin> cleverdra:  yeah, because you are waiting for isforth to have dynamic memory management
11:44:48 <kc5tja> Yes, but now you're not coding in Forth anymore; you're programming in some domain-specific language.  And even so, will it be as concise as the Lisp equivalent?
11:45:34 <futhin> probably not because you would have to include the code for making it polymorphistic in the first place?
11:45:46 <futhin> unless you renamed the "language" to something other then forth :P
11:45:58 <futhin> or called it ANS Forth ver. 2
11:46:33 <cleverdra> futhin - connecting to another server is actually very very easy (caveat re hostnames); and my intended design for smudcl was so simple that there wasn't much else to it, either.  The reason I stopped work is that I couldn't keep with the 'simple' requirement.
11:46:36 <kc5tja> No, because your domain-specific language is almost certainly not nearly as concise as Lisp is.  There are tons of OOFs around, and absolutely none of them can do run-time type checking for intrinsic types on the data stack.  So you always have to qualify them.
11:46:56 <cleverdra> futhin - and it's the non-simple mud client that I stopped work on because IsForth didn't have dynamic memory management.
11:47:35 <cleverdra> You can make an everything an object, though, as in RPL.
11:49:29 <cleverdra> I may fall off.  There's a storm here.
11:53:29 <futhin> cleverda:  why not create a simple sockets package so that morons like me can use them to build the irc clients and mud clients and ftp clients ;P
11:55:05 <cleverdra> futhin - why don't you write a quick networking program to show me the kind of interface you'd like?  'cause I was learning toward dumb-primitives and immediate specialization.
11:58:09 --- join: proteus (~irc@24-197-147-197.charterga.net) joined #forth
12:00:54 --- join: ayrnieu (julianf@0-1pool37-37.nas2.florence1.sc.us.da.qwest.net) joined #forth
12:00:59 <I440r> futhin thats what im doing with isforth. sorta
12:01:02 --- quit: cleverdra (Read error: 104 (Connection reset by peer))
12:01:07 <I440r> i STILL need to fix dns queries tho
12:01:16 --- nick: ayrnieu -> cleverdra
12:03:35 <cleverdra> I440r - ah, are you approaching that problem too?  I think I detailed it in that smudcl source.  Do you want some assembler for getaddrbyname?
12:06:59 <I440r> cleverdra actually, i wouldnt mind it - but remember, ill be doing dns queries by formatting the query and transmitting it to the server via socketcall syscall. i WONT be using any LIBRARIES to do the dns queries
12:07:28 <cleverdra> I440r - there's more to it than that; you want to look at /etc/hosts
12:08:59 <I440r> i know
12:09:23 <I440r> thats why ive not coded this yet hehe
12:09:55 <cleverdra> (do a strace on a networking program)
12:10:05 --- quit: proteusguy (Read error: 110 (Connection timed out))
12:11:28 <cleverdra> actually, I can't find the assembler.  Sorry.
12:11:47 --- nick: kc5tja -> kc-food
12:12:08 <I440r> well if you find it send it to me :)
12:26:58 --- join: tcn (tcn@tc2-login6.megatrondata.com) joined #forth
12:27:04 <tcn> hey
12:27:42 <cleverdra> hello tcn
12:27:44 <I440r> tcn!
12:29:05 --- quit: cleverdra ("Leaving")
12:32:24 <tcn> don't everyone talk at once
12:33:34 <I440r> lol
12:33:43 <I440r> we were chatting but then you came in and it went quiet :)
12:35:14 <I440r> u looked at that assembler ?
12:35:51 <tcn> asm.f? yeah, it was only 1 page and it didn't do anything :)
12:36:05 <tcn> it's a virtual assembler
12:36:17 <I440r> it can assemble SOME instructions already
12:36:20 <I440r> ohhh rite
12:36:26 <I440r> i hadnt got that far when i sent that to you
12:36:29 <I440r> dns me
12:36:34 <I440r> look in my ftp
12:39:55 <tcn> ok.. now write code for MOV and the ALU instructions
12:42:11 <I440r> im working on that
12:42:29 <I440r> im going through all the instruction encodings and reordering them into a more LOGICAL order
12:42:47 <I440r> so that all similarly encoded opcodes are next to each other
12:43:32 <I440r> but its taking time because its boring :P
12:45:52 <I440r> i gotta take the siberian poodle for a walk :P
12:45:54 <I440r> brb
13:09:56 --- quit: proteus (Read error: 110 (Connection timed out))
13:09:57 --- quit: Kitanin (Read error: 104 (Connection reset by peer))
13:12:44 <onetom> the better example is:
13:14:04 <onetom> x(a,b,c,d,e) { y(a,b,c); z(d,e); }
13:14:28 <I440r> back
13:14:36 <onetom> a b c d e x
13:14:48 <onetom> : x z y ;
13:16:01 <onetom> tcn: so, think about it hard! ;)
13:16:35 <tcn> that's not realistic
13:16:48 <onetom> i think u can save the extra preparation of the 2 stack frames 4 y & z
13:17:47 <tcn> i would never pass so many parameters in an inner loop
13:18:03 --- join: proteusguy (~irc@24-197-147-197.charterga.net) joined #forth
13:18:03 <onetom> just by using the proper execution/parameter order
13:18:09 <onetom> loop?
13:18:14 <onetom> who is talkin about a loop
13:18:15 <onetom> ?
13:18:50 <onetom> lets look 4 a similar but realistic example
13:20:03 <tcn> graphics primitives would be a good test of the CPU design
13:23:29 <onetom> k, got a 1st example
13:23:38 <onetom> imagine an object oriented system
13:23:52 <onetom> and lets take a look at constructors
13:24:36 <onetom> init (attrib1, attrib2 .... :integer; newattrib:string)
13:24:37 <onetom> begin
13:24:51 <onetom>   inherited init (attrib1m attirb2 ...);
13:25:11 <onetom>   _newattrib := newattrib;
13:25:16 <onetom> end;
13:25:26 <onetom> its a very common construct ithink...
13:25:36 <onetom> but now i gotta go
13:25:43 <tcn> ok, see ya
13:25:58 <onetom> but plz explain that graphics primitive idea
13:26:14 <onetom> im gonna read back the logs tomorrow (or 2 nite :)
13:26:23 <tcn> circle drawing is sorta register-intensive..
13:26:32 <tcn> gotta make sure we have enough regs :)
13:26:33 --- nick: kc-food -> kc5tja
13:27:01 <tcn> i tried writing a circle algorithm in forth and finally gave up!
13:27:10 <tcn> in C, Asm, etc. it's easy
13:28:03 <I440r> tcn show me the c and asm code
13:28:17 <I440r> breshenham circle code ?
13:28:59 * tcn nods
13:29:18 <I440r> show me the c code :)
13:29:23 <I440r> email it to me
13:30:21 <onetom> and also 2 me, plz :)
13:30:40 <onetom> let us see how can we adapt it 2 4th
13:30:47 <onetom> gnight
13:30:55 <tcn> ok i'll dig it up
13:31:15 <onetom> thx in advance
13:53:53 <I440r> anyone got a forth 32 bit crc function ?
14:10:54 --- quit: proteusguy (Read error: 110 (Connection timed out))
14:35:15 --- quit: tcn ("Leaving")
14:46:01 --- join: Kitanin (~clark@h-209-91-66-234.gen.cadvision.com) joined #forth
14:48:40 <futhin> onetom: don't you think that even if C code is as efficient on x86 as Forth is on forthchips, forth still results in smaller solutions because it is a coding methodology and encourages simplicity?
15:50:23 --- join: proteusguy (~irc@24-197-147-197.charterga.net) joined #forth
15:50:24 --- quit: Kitanin (Read error: 104 (Connection reset by peer))
15:51:31 --- quit: I440r ()
16:12:00 --- quit: proteusguy (Connection timed out)
16:50:40 --- join: proteusguy (~irc@24-197-147-197.charterga.net) joined #forth
16:56:20 --- join: kp2 (kp2@salm-m2-21.wvi.com) joined #forth
16:56:23 <kp2> hey all
16:58:36 <kc5tja> re
17:22:32 --- quit: kp2 (Read error: 113 (No route to host))
17:25:34 <kc5tja> Off to aikido.
17:25:43 --- quit: kc5tja ("THX QSO ES 73 DE KC5TJA/6 CL ES QRT AR SK")
17:28:27 --- join: tcn (tcn@tc1-login10.megatrondata.com) joined #forth
17:43:31 --- join: tom (tcn@tc3-login8.megatrondata.com) joined #forth
18:04:24 --- quit: tcn (Read error: 110 (Connection timed out))
18:09:16 --- quit: proteusguy (Connection timed out)
18:09:48 --- join: CrowKiller (punkrules@Ottawa-HSE-ppp3653644.sympatico.ca) joined #forth
18:11:33 * CrowKiller tried to flash his laptop's BIOS with a bootable cdrw, it failed miserably
18:12:42 <tom> killed it?
18:12:52 <CrowKiller> yes
18:13:15 <CrowKiller> I hope the flash isnt soldered into the motherboard
18:13:25 <tom> heh
18:13:34 <CrowKiller> I'll try to reprogram it myself with the specs
18:13:44 <tom> flash is a dumb idea
18:13:49 <CrowKiller> if I can find them ;p
18:14:24 <futhin> tom: don't you think that even if C code is as efficient on x86 as Forth is on forthchips, forth still results in smaller solutions because it is a coding methodology and encourages simplicity?
18:15:22 <tom> in certain cases
18:15:35 <tom> in other cases it's just frustrating
18:15:51 <CrowKiller> I'll actually try to finish my Forth system on this machine (friend's computer) so I can program the flash using a 100% homemade solution
18:16:10 <futhin> the frustrating part would be the fact that you are starting from low-level ?
18:17:04 <tom> no, being forced to use RPN
18:17:27 <futhin> umm..  my perspective is that there isn't any RPN
18:17:44 <tom> CK: you'll have the kind of BIOS you ought to.. just a simple bootloader
18:18:03 <futhin> but then, my perspective is that there isn't a data stack.
18:18:25 <tom> are you still talking about forth?
18:19:18 <futhin> i think the whole point of the stack..  is not because it's there, but because it lets you do implicit data passing..   i think forthers are supposed to forget about the stack, and pretend that "implicit data passing" is the main feature, not the side effect
18:19:59 <CrowKiller> maybe I could hack myself a minimal BIOS that could fetch bytes in memory using the EPP Pport and then do a jump at the start when finished downlaoding
18:20:17 <CrowKiller> futhin: nameless data storage is crucial point in forth
18:20:46 <futhin> RPN is really only occuring in isolated cases..  1 2 + .    with a different perspective (or a higher level one)  you are able to create language close to english. 
18:21:00 <CrowKiller> "over naming" is considered a bad programming practice as much as "over factoring"
18:21:07 <tom> yeah
18:21:13 --- join: TheBlueWizard (TheBlueWiz@ip-216-25-202-149.vienna.va.fcc.net) joined #forth
18:21:18 <TheBlueWizard> hiya all
18:21:18 <futhin> : WASHERMACHINE  FILL SPIN EMPTY SPIN ;
18:21:20 <CrowKiller> hi TBW
18:21:21 <futhin> or whatever
18:21:25 <TheBlueWizard> no I440r in sight....
18:21:33 <TheBlueWizard> hiya CrowKiller
18:21:42 <futhin> crowkiller:  over factoring?  wtf?
18:21:54 <CrowKiller> defining too much words
18:22:03 <TheBlueWizard> futhin: why the second SPIN in the definition?
18:22:09 <tom> i kinda like how C string operations, for example, are done with a few pointer operations, not a bunch of factored out words..
18:22:18 <futhin> thebluewizard: to dry it..
18:22:33 <tom> ie, c = *p++;
18:22:35 <futhin> thebluewizard: it spins a bit while hte water drains out
18:22:55 <futhin> crowkiller:  most forth coders under factor in the first place, over factoring is not a concern
18:23:06 <futhin> crowkiller: i  highly doubt theres any such thing as over factoring
18:23:15 <futhin> factoring a lot is good programming
18:23:35 <CrowKiller> tom: do you know about the "A register"?
18:23:35 <tom> i know you hate C with a passion :)
18:23:45 <tom> yes CK
18:23:49 <futhin> tom: me? :)
18:24:16 * tom nods
18:24:46 <TheBlueWizard> futhin: I see re: needing to spin a bit
18:25:29 --- join: skylan (sjh@Riverview35.tbaytel.net) joined #forth
18:25:36 <tom> I haven't "proven" it yet, but I have a hunch that Forth will lose because of DUP, SWAP, ROT, etc.
18:25:38 <CrowKiller> +! is the same thing as c = *p++
18:25:55 <futhin> tom:  i haven't coded enough forth to determine how easy it is to get past the rought spots of forth..  
18:25:55 <CrowKiller> rot isnt needed
18:26:00 <CrowKiller> swap sometimes
18:26:07 <tom> put * and ++ can work on any variable not just the A register
18:26:15 <tom> s/put/but
18:26:15 <futhin> tom: if everyone codes like chuck moore, is there any problem?
18:26:39 <CrowKiller> tom: its an abstraction
18:27:49 <tom> it makes it easier to read.. it's consistent
18:28:37 <futhin> forthers should be pretending the stack doesn't exist
18:28:55 <futhin> all begining forthers seem to try to treat it like an array or just put wayyy too many items on the stack
18:29:08 <tom> right
18:29:16 <tom> they should be using more variables
18:29:20 <futhin> no!
18:29:21 <futhin> eww
18:29:31 <futhin> heh
18:29:42 <futhin> no, they should be coding like chuck moore ;P
18:29:48 <futhin> chuck moore sez:
18:29:48 <CrowKiller> "young" forthers need a lot of GOOD examples, and there's isnt any around usable enough
18:29:57 <tom> what are you gonna say, they should take shortcuts?
18:30:16 <futhin> a) a word should be 2 lines of code or less
18:30:26 <futhin> b) don't pass more than 2 parameters by stack
18:30:33 <futhin> c) don't mess with more than 3 stack items at a time
18:31:04 <tom> i think forth had it's place, in machines with small RAM that had to be interactive
18:31:04 <futhin> d) stack depth of 6 (or was it 8? or 16?)  is effectively infinite
18:31:08 <CrowKiller> tom: I wrote a C18 virtual machine compiler for x86 with a footprint of 400 bytes, and it should be quite speedy
18:31:26 <tom> what's C18?
18:31:39 <futhin> there are 25 C18 cores in a 25X forthchip
18:31:50 <CrowKiller> this is about 32 instructions
18:31:56 <tom> oh that :)
18:32:48 <tom> i'm thinking a single 100mhz CPU ought to be enough for anything if the software isn't bloated
18:32:58 <futhin> i agree absolutely
18:33:04 <futhin> my computer is a p133 mhz with 32 megs of ram
18:33:13 <CrowKiller> its intended to work in a "living" system, ex: compiling an interupt routine table each time its needed on the fly
18:33:24 <futhin> and i think it's ridiculous when windows and i.e runs a little slowly slowly on it
18:33:34 <futhin> my previous computer was a p75mhz with 16 megs of ram too
18:33:38 <tom> i don't see much reason for parallel processing.. i'm sure there are applications for it but there's not enough of a market for me to rack my brains messing with it :)
18:33:52 <tom> i have no passion for it :)
18:35:32 <futhin> tom:  for parallel stuff?   well who really cares, the 25X chip is mostly in the "interesting" realm of things..  but it's got about 2500 mips available to it, so it might be useful somehow..  and a good performance for the price
18:35:41 <tom> heh.. i'd like to build a homebrew machine sometime, without surface-mount components.. that means, 16-bit..
18:36:07 <futhin> tom: yeah i've been talking with kc5tja about building a forthchip desktop computer
18:36:43 <futhin> patriot science sells shboom chips (350 mips) for 15 bucks, so i was thinking it would be interesting to build a forthchip desktop computer using shboom chips..
18:37:04 <tom> oh, the price has come down
18:37:21 <tom> can you get small lots for that price?
18:37:23 <futhin> F21 chips are more expensive (100 bucks maybe)  but they would be good for the video card :)
18:37:31 <futhin> tom:  small lots ?
18:37:52 <tom> less than say $100 worth of shbooms?
18:38:12 <futhin> tom:  yeah, they sell them individually, i emailed them asking about that
18:38:20 <futhin> it's 15 bucks plus shipping and handling
18:38:24 <tom> sometimes prices are $5 ea for 1000, $500 ea for 1
18:38:37 <futhin> naw, it's 15 bucks for 1 :)
18:38:42 <tom> direct or from a distributor?
18:38:45 <futhin> direct
18:38:56 <CrowKiller> do they ship in canada? ;pp
18:39:01 --- quit: skylan (Connection reset by peer)
18:39:02 --- join: skylan (sjh@Rockcliffe83.tbaytel.net) joined #forth
18:39:08 <futhin> crowkiller: they'll ship to canada, sure
18:39:13 <futhin> 15 bucks + shipping and handling costs
18:39:30 <futhin> you send them your credit card number
18:39:40 <futhin> thru fone or email
18:40:03 <tom> their web site sucks
18:40:10 <futhin> yup :)
18:40:32 <tom> i wonder if there's a 16-bit 6502 descendant
18:41:05 <tom> there's also a 6816 which would be a good candidate..
18:41:25 <futhin> ?
18:41:35 <tom> for homebrew..
18:41:43 <TheBlueWizard> tom: there is a 65816...a 16-bit chip
18:41:45 <futhin> tom:  so you built a 2 stack machine with 20 ttl parts ?
18:42:11 <TheBlueWizard> it is currently being made by Western Digital
18:42:39 <tom> i designed it
18:42:51 <tom> thanks tbw
18:43:16 <TheBlueWizard> np :)
18:43:49 <TheBlueWizard> but I am not exactly thrilled with 65816 architecture...well, google it and you can judge that for yourself :)
18:44:00 <CrowKiller> tom: is the design posted somewhere?
18:44:07 <tom> heh.. so that's what the IIgs and the SNES used
18:44:29 <tom> CK: nope.. but Brad Rodriguez posted a similar design, which he actually built :)
18:44:42 <CrowKiller> thanks
18:45:10 <tom> i think mine was a little simpler.. but would it work? ;)
18:46:26 <tom> heh.. that 65816 sounds like what the 386 is to the 8086
18:47:23 <TheBlueWizard> in some ways, yeah...though minus the protection stuff, of course
18:48:08 <tom> hmm.. 24 address lines.. that's enough
18:48:51 <TheBlueWizard> 24 bits --> 16 meg
18:50:54 <tom> but w/ the 6502 you could use a little glue logic to do bank switching..
18:52:46 <TheBlueWizard> yeah...kinda hackish...and is obviously design specific
19:02:58 <tom> heh.. i guess Small C looks more appealing after learning Forth
19:03:23 <tom> everything is an 'int' pretty much.. it doesn't check types
19:05:35 <TheBlueWizard> hehe...actually it supports two types: int and char (in one implementation [x86 I think] it has to distinguish between function and data pointers)
19:08:29 <futhin> tom: how much coding have you done in forth?   how big is retro and was it coded in forth?
19:09:34 <tom> i forget.. it was pretty small
19:09:47 <TheBlueWizard> gotta go...bye all
19:09:50 <tom> maybe a 20k kernel
19:09:51 <tom> see ya
19:09:54 <futhin> tom:  coded in forth or asm?
19:09:55 <CrowKiller> bye
19:10:00 <TheBlueWizard> bye tom
19:10:03 <TheBlueWizard> bye CrowKiller
19:10:04 <futhin> bye tbw
19:10:12 <TheBlueWizard> bye futhin
19:10:26 --- part: TheBlueWizard left #forth
19:10:56 <tom> bah. there's an x86 Small C on sourceforge, with nothing released in over a year of existence
19:11:04 <futhin> tom:  coded in forth or asm?
19:11:13 <futhin> i'm not trying to attack you here or anything
19:11:49 <tom> about half and half
19:11:56 <futhin> what do you perceive as the problems with forth besides RPN?
19:12:44 <tom> the explicit stack is the main problem
19:13:17 <futhin> it's only explicit if you think of the stack
19:13:24 <futhin> i dunno
19:13:36 <futhin> i think it's a matter of perspective :P
19:13:51 <tom> i mean, it works great for little things.. you get into modest sized programs and it's tough to manage
19:13:52 <futhin> i try to avoid thinking about the stack, and just pretend implicit data passing is the main feature
19:14:05 <futhin> but that's what factoring is for
19:14:15 <tom> what have YOU written in forth?
19:14:31 <futhin> small crappy stuff hehe
19:14:46 <futhin> my biggest project was in postscript
19:15:25 <futhin> never actually finished it, but i was struggling with coming up with the simplest solution and factoring out all the words properly
19:15:35 <futhin> it was quite hard to make small words
19:15:38 <futhin> and factor everything out
19:15:48 <futhin> i haven't mastered that yet
19:15:54 <tom> so i see :)
19:16:20 <tom> you should really try C again
19:17:08 <futhin> or learn lisp :P
19:17:24 <CrowKiller> futhin: or write you own forth
19:17:41 <CrowKiller> and then code in it the way you like
19:17:52 <futhin> yes
19:17:53 <futhin> that too
19:18:04 <futhin> i don't think i'm going to make much progress with forth until i code my own forth
19:20:22 <CrowKiller> start now ;p
19:20:27 <tom> well, it would teach you something about compilers without really dealing with parsing
19:20:50 <futhin> maybe i should learn TCL too :P
19:21:15 <futhin> tom:  yeah, there's no parsing in forth..   i have no idea how to parse strings in forth..
19:21:19 <futhin> heh
19:21:21 <tom> if i hadn't written forth compilers i would never have thought about writing an incremental C compiler
19:21:29 <futhin> suppose i have a string
19:21:37 <futhin> how do i check that the first character of the string is a < 
19:21:50 <CrowKiller> cmp ;p
19:21:56 <futhin> and that there is an enclosing > somewhere down the string?
19:22:00 <tom> over c@ '< = if ... then
19:22:19 <tom> HTML parser? :)
19:22:24 <futhin> nope
19:22:25 <futhin> irc
19:23:52 <futhin> what do you have OVER for heh
19:24:08 <futhin> you are pedicting something about the items on the stack already? :P
19:24:18 <tom> you said there's a string..
19:24:27 <futhin> yeah ok
19:25:00 <futhin> over c@ < = if  <code to keep checking for the enclosing ">"> then ..
19:25:01 <futhin> hrm
19:25:03 <futhin> what a pain
19:25:31 <tom> welcome to parsing
19:25:33 <futhin> is there any forth libraries/code that provide enhanced parsing stuff :P
19:25:47 <tom> you might search for FTRAN
19:25:50 <CrowKiller> dont need to
19:26:21 <futhin> crowkiller: ?
19:26:34 <CrowKiller> why would you need strings, even if you deal with the "outside world"? dont like strings, think of something better for YOUR application
19:27:11 <futhin> crowkiller:  lots of useful applications need to parse strings..  i need to parse strings for many reasons
19:27:31 <futhin> they aren't avoidable unless perhaps EVERYTHING in the computer world was forth
19:27:39 <futhin> from the irc servers to the os..
19:27:42 <futhin> perhaps
19:28:28 <futhin> i need to parsing for an irc client, for a web browser, for other stuff..
19:28:32 <tom> or if you do like Chuck, and deal with nothing but graphics
19:29:01 <CrowKiller> he didnt wanted to deal with ascii strings because he saw no sue for them
19:29:35 <CrowKiller> in HIS application, he didnt need them
19:29:45 <CrowKiller> my compielr dont do parsing
19:30:25 <CrowKiller> it process bytewide huffman encoded packets, or "tokens"
19:30:30 <tom> well.. for most real world tasks you need to deal with text
19:30:41 <tom> like what we're doing now, IRC
19:30:47 <futhin> yes
19:31:06 <tom> it's one of forth's weak point
19:31:07 <futhin> unless the protocol was radically different and supported forth ;)
19:31:27 <futhin> tom:  yeah, forth should come with parsing tools..
19:31:35 <tom> you know, a lot of things are easier with 0-terminated strings
19:31:39 <futhin> words that make parsing simple..
19:31:51 <futhin> what's wrong with strings & string lengths ?
19:32:05 <tom> in C you're only dealing with an address, not a count
19:32:07 <futhin> not all strings are terminated with some character
19:32:30 <CrowKiller> pack the strings, count them, and you'll get something better; eliminate them you'll be in heaven ;p
19:32:34 <futhin> in C all strings are 0-terminated ?
19:32:35 <tom> if it's < do this, " do that, 0 quit, anything else, print it..
19:32:49 <tom> futhin: yes
19:33:01 <tom> they don't have to be but it's a convention
19:33:50 <tom> there's one thing where it's a drawback.. counting the string length
19:34:39 <tom> for most things, it's one less register you have to use in an inner loop
19:34:47 <CrowKiller> dont terminate it by zero, count it at the very begining instead
19:34:51 <futhin> in ASM do you do 0-terminated strings?  or is it easier to count string length?
19:35:13 <tom> i dunno.. people do both sometimes
19:35:21 <futhin> i think i recall it's easier to count string length in asm
19:35:27 <futhin> than to cmp all the time
19:35:37 <tom> you don't cmp..
19:35:48 <futhin> like for printing out the string, it's realy easy to just print out as many chars as the count is
19:35:49 <CrowKiller> the BIOS utilize 0 termianted strings in its display routines
19:36:02 <CrowKiller> futhin: LEARN ASM NOW ;p
19:36:03 <futhin> tom: you have to cmp to find the end of the string..
19:36:10 <tom> nope
19:36:11 <futhin> crowkiller:  i know some asm
19:36:18 <tom> just check the zero flag
19:36:19 <futhin> it's a little rusty that's all
19:36:37 <CrowKiller> futhin: know it all, code your compiler, exercuise yourself, sharpen your brain
19:37:03 <futhin> tom:  you have to check the flag tho?
19:37:04 <tom> yeah, write a little forth compiler, even if C is better
19:37:38 <futhin> in asm its much more common to have the string length, it makes it easier to display a string iirc
19:37:40 <tom> LODSB;  OR AL,AL;  JZ END
19:38:07 <tom> hmm
19:38:28 <tom> you'd put the string pointer in ESI, location on the screen in EDI..
19:38:47 <tom> and color in AH.. then,
19:39:12 <tom> .1: LODSB; OR AL,AL; JZ END;  STOSW; JMP .1
19:39:55 <tom> with counted strings you put the count in ECX and LOOP instead of JMP
19:39:57 <CrowKiller> on the pentium, test AL, AL gives a better pairing (faster)
19:40:07 <tom> oh
19:40:16 <CrowKiller> little parenthesis ;o)
19:40:21 <tom> thanks
19:41:35 <tom> I coded printf() in asm.. that was pretty fun
19:42:32 <tom> except i bet you'd get the same code writing it with a switch() in C
19:42:36 <futhin> .1: movsb stosb loop .1     ??
19:42:52 <tom> movsb = lodsb stosb
19:43:02 <futhin> mov cx, stringlength
19:43:11 <CrowKiller> when I code in asm I almost always llok at this page and the overall site: http://www.quantasm.com/optxref.html
19:43:22 <CrowKiller> asm for x86, that is.
19:43:32 <tom> you use lodsb/stosb if you have to do something in between.. REP MOVSB to move a big chunk..
19:43:39 <futhin> jnz .1
19:44:15 <futhin> mov cx, stringlength .1: lodsb stosb jz .1 
19:44:25 <futhin> is that analogous?
19:44:38 <futhin> is that comparable to your 0-terminated code?
19:45:42 <tom> mov cx, stringlength  rep movsb
19:46:18 <tom> (counted string)
19:46:44 <futhin> how many lines of code do you think a quicksort in asm would be?
19:46:54 <tom> .1: lodsb  test al,al  jz .done  stosb  jmp .1
19:47:03 <tom> (0-terminated string)
19:47:25 <futhin> looks like the counted string one is a little shorter :P
19:47:35 <tom> for the simplest case, yeah
19:47:40 <CrowKiller> counted strings rulez ;p
19:47:46 <tom> try parsing with them
19:47:52 <CrowKiller> packed+counted rulez even better
19:47:56 <futhin> the count could be abstracted out in forth
19:48:13 <futhin> you could have words dealing with the strings & count and you don't have to know about the count
19:48:21 <futhin> you could abstract it out in parsing words
19:48:30 <futhin> and the parsing stuff could be included with forth
19:48:37 <futhin> and newbies could come along and see that parsing is easy
19:48:46 <futhin> and the newbies take over the world!
19:49:26 <CrowKiller> parsing is dead easy already
19:49:39 <CrowKiller> depends on how much you want to hurt yourslef as a programmer ;p
19:50:55 <tom> heh.. just as I suspected.. LODSB etc is slower on the 486 & up.. these chips are becoming more RISC-like
19:51:10 <CrowKiller> slower but faster
19:51:26 <CrowKiller> doh
19:51:27 <CrowKiller> lol
19:51:29 <futhin> if lodsb is slower, what instruction is used now ?
19:51:36 <CrowKiller> futhin
19:51:50 <tom> heh.. look at that web pages
19:52:01 <CrowKiller> the previous one I gave on the chan
19:52:03 <tom> http://www.quantasm.com/optxref.html
19:52:05 <CrowKiller> its all there
19:52:26 <CrowKiller> its one byte so its much better, you get 4-5x object code compression
19:53:51 <tom> personally, i never bother with pentium optimizations.. don't need the extra speed.
19:54:27 --- join: proteus (~irc@24-197-147-197.charterga.net) joined #forth
19:54:41 <tom> but I might make a compiler do some optimizations
19:56:57 <tom> oh well.. time for me to call it a night
19:57:20 <futhin> tom: wait
19:57:36 <futhin> tom: tell me what you think it would take for forth to make you happy?  parsing?  anything else?  
19:57:45 <futhin> do you really hate RPN and the stack that much? :P
19:59:00 <tom> yeah, it's a waste of time for "substantial sized" programs
19:59:17 <tom> makes a good programmable calculator though
20:00:00 <futhin> :(
20:00:15 <CrowKiller> it encourages you to make small programs in fact, the stack is the ultimate multibitwide I/O component, but its small
20:00:39 <futhin> tom: a CAD program has been coded in forth.. www.forthcad.com
20:01:05 <futhin> crowkiller: yeah, substantial sized programs aren't supposed to exist in forth :P
20:01:22 <futhin> crowkiller:  i think you comment about "over factoring" is TOTAL PURE BULLSHIT :P
20:01:51 <CrowKiller> think of a system where the call/ret overhead slow down the program
20:02:04 <CrowKiller> that what I was refering to
20:02:12 <futhin> bah, only losers use x86
20:02:16 <futhin> real men use forthchips :P
20:02:32 <futhin> the over head is less in forth anyways
20:02:36 <futhin> forth has the return stack
20:02:42 <futhin> that's how it tracks "calls"
20:02:43 <CrowKiller> i knwo all that, thanks ;p
20:02:47 <futhin> it doesn't have to have call
20:02:49 <futhin> it can use jmps
20:03:17 <CrowKiller> ?
20:03:50 <CrowKiller> call exist but take one "cycle"
20:04:07 <futhin> indirect threading, direct threading, subroutine..
20:04:27 <futhin> some forths don't make "CALL" instructions at all..
20:04:34 <futhin> or rarely..
20:04:42 <CrowKiller> they do calls
20:04:49 <futhin> no, some do JMPs
20:05:08 <CrowKiller> they do pushRetStack Jumpafterwards
20:05:08 <futhin> ITC does jumps i think..
20:05:34 <futhin> yeah, they aren't making x86 asm CALL
20:05:41 <CrowKiller> its the same as doing all the instruction of lodsb instead of using lodsb
20:06:06 <CrowKiller> the behavior is the same
20:07:01 <futhin> CALL is more resource intensive than just JMPing i think
20:07:28 <futhin> subroutine forth  uses CALLs
20:07:38 <futhin> : test dup blah ;
20:07:52 <futhin> compile into test:  CALL dup  CALL blah
20:08:05 <futhin> but ITC does  JMP dup  JMP blah or whatever
20:08:28 <futhin> people should code editors in asm!
20:08:35 <futhin> for fun
20:13:19 <CrowKiller> im coding an editor
20:13:22 <CrowKiller> in asm
20:13:28 <CrowKiller> its the second part of my system
20:13:33 <CrowKiller> code one for me
20:13:44 <CrowKiller> i have no idea how a good editor must be made lol
20:14:20 <CrowKiller> kc5tja's scene renderer could be the way to go
20:15:07 <futhin> scene graph renderer
20:18:17 <CrowKiller> chuck didnt liekd the video hardware for text rendering so he wrote his own "virtual card" as a second task in his colorforth
20:18:39 --- quit: tom (Read error: 110 (Connection timed out))
20:18:49 <CrowKiller> an editor that could use the textmode legacy device present in every pc would not be portable but would be very small
20:21:09 <futhin> you know, i'm gonna learn lisp
20:21:22 <futhin> i want to see what the hell lisp macros are all about
20:21:33 <futhin> i suspect forth can do lisp macros
20:21:40 <futhin> but i should learn lisp
20:29:02 <CrowKiller> A killer editor is what's needed to make my x86 forth popular
20:29:22 <CrowKiller> I think i'll write it in javascript or something trabslations will follow if tehre's enough interest
20:32:37 <futhin> ??
20:33:04 <futhin> an editor in javascript?
20:33:14 <futhin> running on top of your 400 byte forth compiler?
20:33:34 <CrowKiller> no the editor runs online
20:33:40 <CrowKiller> its completely parametrizable
20:33:45 <CrowKiller> usable from almost everywhere
20:33:52 <CrowKiller> even from a windows box
20:34:08 <CrowKiller> it outputs let say base64 chars
20:34:35 <CrowKiller> cut and paste into an hex editor and voila
20:34:47 <CrowKiller> i skipped a step here
20:34:59 <CrowKiller> anyway you get the point ;p
20:35:21 <CrowKiller> you can print your program for future use or anything i dont know
20:35:52 <CrowKiller> maybe a windows hack would be better
20:36:13 <CrowKiller> resizing an Edit box to accept "length of editor" bytes
20:36:24 <CrowKiller> paste it in there
20:36:40 <CrowKiller> send a timer message
20:36:48 <CrowKiller> that does calback and voila
20:37:18 <CrowKiller> i saw a paper on exploiting windows messages for privilege escalation on cracked systems or local systems
20:37:47 <CrowKiller> the guy used 4 or 5 widnwos message to implement the code
20:37:49 <CrowKiller> very small
20:42:20 <CrowKiller> I dont want to code the editor in asm, I'll try to come up with something
20:42:50 <CrowKiller> I dotn have any editor, so maybe i,ll consider writing the byte stream for the compielr myself ;p
20:43:23 <CrowKiller> keybaord handling is still a problem, i'll try to experiment with those
20:43:43 <CrowKiller> like a less than 1$ parallel port keyboard, if you already have the connector
20:45:17 <CrowKiller> made of a single pcb or maybe small diameter copper wires with a plate at the end, acting as a key, anyway i'm out to sleep
20:45:20 <CrowKiller> see ya everyone
20:45:43 --- quit: CrowKiller ("Bom bom gigolo")
21:11:30 --- quit: proteus (Connection timed out)
21:34:16 --- join: kc5tja (~kc5tja@ip68-8-206-226.sd.sd.cox.net) joined #forth
21:46:08 <futhin> hello kc5tja :)
21:46:22 <kc5tja> Hello
21:49:27 <futhin> tcn doesn't like forth anymore..  because of RPN and lack of parsing it seems :(
21:50:04 <kc5tja> Those are the very reasons to like it though.
21:51:16 <futhin> really?
21:51:40 <futhin> i was thinking that forth should have some parsing words to make it easier for people to parse strings..
21:52:00 <kc5tja> I like Forth because it is brutally simple.
21:52:04 <futhin> like if a person wants to make a "useful" application, like an irc client or whatever..  chances are he's gonna need parsing
21:52:26 <kc5tja> Well, C doesn't have any parsing facilities of its own either.
21:52:43 <kc5tja> I'm not entirely too sure what is meant, in this context, by parsing.
21:54:41 <futhin> he mentioned "c = *p++;"
21:54:45 <futhin> me
21:54:51 <kc5tja> OK.
21:54:56 <kc5tja> I hardly consider that parsing though.
21:55:00 <futhin> i'm just thinking of parsing a string and finding  the first <  and the enclosing >  and the stuff in between
21:55:10 <kc5tja> That's just iterating the characters of a string, which Forth can do just fine.
21:55:25 <futhin> i'm not really sure of how to parse strings in forth..  
21:55:46 <futhin> like finding  the first "<"  and the enclosing ">"  and extracting the stuff in between
21:56:15 <kc5tja> You'd have to do it in Forth the exact same way you would do it in C -- roll your own.
21:56:52 <futhin> over c@ < = if begin over i + c@ > = if ..... then ..... then  
21:56:54 <futhin> ?
21:57:09 <kc5tja> If you write code like that, you wil quickly get mired down.
21:57:36 <kc5tja> DUP C@ '< = IF Found'<' ELSE ... THEN
21:57:49 <kc5tja> s/DUP/OVER/ (sorry)
21:58:13 <kc5tja> : Found'<'   2DUP '> PARSE HandleXMLTag ;
21:58:43 <kc5tja> This is just one of a whole myriad of ways you can do this sort of thing in Forth.
22:01:13 <futhin> forth could come with some parsing vocabulary to make things easier for us :P
22:01:36 <kc5tja> But, what is parsing??
22:01:46 <kc5tja> Parsing is different depending on what you're trying to do.
22:01:46 <futhin> well i mean, string handling
22:01:51 <kc5tja> XML parsing is nothing like parsing a C program.
22:01:57 * kc5tja shrugs
22:02:02 <kc5tja> I like it just the way it is.
22:02:14 <kc5tja> At the most primitive level, it's not really any different from C.
22:02:28 <kc5tja> If you want, you can make implementations of C's string handling functions for Forth.
22:03:03 <futhin> what C string handling functions are you thinking of?
22:03:18 <kc5tja> strtok() and stptok()
22:03:28 <futhin> heh, i don't even know those
22:03:34 <kc5tja> But even then, *none* of C's functions support "nesting" of things at all.
22:03:47 <futhin> nesting is forth's advantage?
22:03:48 <kc5tja> You need an abstract grammer of some kind, along with a tool like Bison or Flex.
22:03:54 <kc5tja> Forth already has GREY.
22:04:09 <kc5tja> No, Forth's primitives don't support nesting either.
22:04:11 <futhin> GREY?  hmm time for me to google for that
22:04:32 <kc5tja> GREY is a parser generator along the same lines as Flex/Bison.
22:09:55 <futhin> what do you mean about nesting?
22:10:33 <futhin> btw, how do you recommend i handle a string where i want to find "<" and the next ">"  and extract the stuff in between?
22:11:16 <kc5tja> ( this is a ( valid comment ) BUT THIS ISN'T )
22:11:27 <futhin> ah
22:11:41 <kc5tja> PARSE does that for you.
22:11:48 <futhin> yeah
22:11:53 <kc5tja> Given the start of a string, and a character to look for, it'll returning the containing string.
22:12:03 <futhin> hrm?
22:12:24 <kc5tja> Look up the ANSI Forth definition of PARSE.  I can't explain it any more succinctly than I just have.
22:13:27 <futhin> heh
22:14:03 <futhin> okay.. so:   over c@ < = if parse > then 
22:14:07 <futhin> something like that
22:14:22 <kc5tja> OVER C@ '< = IF 2DUP '> PARSE ..etc.. THEN
22:14:33 <kc5tja> PARSE is like WORD
22:14:46 <kc5tja> Except it returns a real string, and doesn't modify the source string at all.
22:16:49 <futhin> what's with the ' before the character ?
22:17:15 <kc5tja> That's just my shorthand for [CHAR].  So if you have '> it's the same as [CHAR] >
22:17:22 <kc5tja> Pygmy supports this style natively.
22:17:29 <kc5tja> (Pygmy doesn't have PARSE, however)
22:18:15 <futhin> so none of the forths support '>  ?   i have to type [char] >   ?
22:18:38 <kc5tja> I can't answer that; that's a Forth-specific issue.
22:18:55 <kc5tja> All I know is that pygmy supports it; I have no idea if any other one does.
22:18:59 <kc5tja> If you want, you can do this:
22:19:14 <kc5tja> [CHAR] < CONSTANT '<
22:19:19 <kc5tja> [CHAR] > CONSTANT '>
22:22:51 <futhin> so pygmy  defined a whole bunch of words '> 'a 'b 'c etc?
22:23:02 <kc5tja> No.
22:23:22 <kc5tja> I said Pygmy supported the 'A syntax natively.
22:23:33 <kc5tja> I do not know if other Forths do or do not. 
22:23:38 <futhin> yeah
22:23:50 <futhin> pygmy's parser supports that..  k
22:25:07 <kc5tja> It's understood to be a special variant of a number.
22:28:18 <kc5tja> Well, I think I'm going to hang with my roomie and watch a movie.
22:29:06 <futhin> ok, ttyl
22:50:41 --- join: proteusguy (~irc@24-197-147-197.charterga.net) joined #forth
22:51:03 --- quit: skylan ("Reconnecting")
22:51:07 --- join: skylan (sjh@Riverview18.tbaytel.net) joined #forth
22:52:58 --- quit: kc5tja ("BitchX: TASTES GREAT! LESS FILLING! TASTES GREAT! LESS FILLING!")
23:14:46 <futhin> anybody alive?!
23:59:59 --- log: ended forth/02.08.16