00:00:00 --- log: started forth/02.06.18
00:08:33 --- quit: rob_ert (Read error: 113 (No route to host))
00:13:00 --- join: davidw (~davidw@adsl-32-74.38-151.net24.it) joined #forth
00:59:19 --- quit: Stepan ("Client Exiting")
02:45:26 --- quit: Fractal (Read error: 110 (Connection timed out))
02:49:57 --- join: rob_ert (~robert@h237n2fls31o965.telia.com) joined #forth
02:52:53 --- join: Serg_penguin (~snaga_NOI@nat-ch1.nat.comex.ru) joined #forth
02:52:58 <rob_ert> Hello :)
02:53:03 <Serg_penguin> hi
02:53:36 <Serg_penguin> what do u work on now ?
02:54:40 <rob_ert> Not much, I'm going out :P
02:54:54 <Serg_penguin> i now play Arcanum and Dungeon Siege ;)
02:54:57 <rob_ert> Then I'll code some more on my file system code.
02:55:22 <Serg_penguin> just upgraded - P4 1600, GeForce 2 Pro, 256 DDRAM...
02:55:27 <Serg_penguin> what FS ?
02:55:40 <Serg_penguin> existing or original ?
02:56:25 <rob_ert> My own ;)
02:56:38 <rob_ert> I got to keep it simple, otherwise it gets too hard for me, hehe.
02:56:49 <rob_ert> And I want to make a really small implementation.
02:57:04 <Serg_penguin> i read Olafy book "networked OSes" - knew much new abt FS, RAM management, protection etc...
02:57:23 <Serg_penguin> what it will look like - fat, ext2, ntfs ?
02:58:51 <rob_ert> More like FAT.
03:00:09 <Serg_penguin> sux, better just implement fat driver
03:00:30 <Serg_penguin> it is VERY prone to get messed up
03:00:55 <Serg_penguin> it worth only storing NT swap...
03:03:11 <Serg_penguin> i was impressed by unix os'es as they described by Olafy
03:03:19 <Serg_penguin> unix FS'es err
03:03:50 <Serg_penguin> and by ntfs idea - anything is tagged atribute, inc. data
03:04:10 <Serg_penguin> but it too damn complex..
03:05:28 <Serg_penguin> bye, work...
03:05:30 --- quit: Serg_penguin ()
03:15:13 --- join: mur (ammu@baana-62-165-189-214.phnet.fi) joined #forth
04:25:43 <davidw> so, what cool things have you done with forth today?
04:26:26 <rob_ert> No forth today :)
04:51:12 --- quit: mur ("reboot")
04:56:59 --- join: mur (ammu@baana-62-165-189-214.phnet.fi) joined #forth
05:04:07 --- join: Serg_penguin (~snaga_NOI@nat-ch1.nat.comex.ru) joined #forth
05:04:07 --- quit: Serg_penguin (Client Quit)
05:06:51 --- quit: mur (Read error: 104 (Connection reset by peer))
05:23:05 --- join: Serg_penguin (~snaga_NOI@nat-ch1.nat.comex.ru) joined #forth
05:23:14 <Serg_penguin> hi
05:35:12 * Serg_penguin slaps rob_ert around a bit with a large trout
05:40:50 --- part: Serg_penguin left #forth
05:52:48 --- join: dsmith (firewall-u@cherry7.comerica.com) joined #forth
06:50:22 --- join: cleverdra (julianf@0-1pool37-1.nas2.florence1.sc.us.da.qwest.net) joined #forth
06:51:19 <rob_ert> Hi :)
06:51:50 <cleverdra> hello rob_ert
06:54:13 --- quit: Soap` (Read error: 104 (Connection reset by peer))
06:57:19 --- join: Serg_penguin (~snaga_NOI@nat-ch1.nat.comex.ru) joined #forth
06:57:27 <Serg_penguin> hi
06:58:22 <rob_ert> Hi.
07:16:15 * Serg_penguin slaps rob_ert around a bit with a large trout
07:21:11 --- quit: Serg_penguin ()
07:25:40 <cleverdra> What a *jerk*!
07:27:01 <davidw> who?
07:27:39 <cleverdra> This Serg fellow, who came in, said "hi", slapped rob_ert with a fish, and then left.
07:48:54 <rob_ert> cleverdra: He also messaged me.
07:49:10 <rob_ert> And I think he slapped me because I went away in the middle of the conversation :)
08:33:10 --- join: Herkamire (~jason@68.15.54.54) joined #forth
08:33:45 <rob_ert> Hey Herkamire :)
08:33:50 <Herkamire> hey :)
08:34:36 <Herkamire> I'm working on an a 3d wireframe viewer for anaglyph (red and blue) glasses :)
08:34:47 <Herkamire> (in forth)
08:35:05 <rob_ert> Wow :)
08:35:22 <davidw> was it you that was talking about anti-aliasing?
08:35:23 <rob_ert> I wouldn't even be able to do 3D stuff in BASIC :(
08:35:54 <Herkamire> somebody else (speuler?) brought up anti-aliasing. I was just whining about lines
08:36:37 <Herkamire> I don't know all the 3d stuff, but so long as the camera look directly along the z axis, it shouldn't be too hard
08:37:07 <Herkamire> I've almost finished 3dlinto ( x y z -- )
08:37:37 <Herkamire> I haven't even started on rotating a model
08:38:41 <davidw> are you going to release it?
08:39:09 * davidw thinks it's cool that someone here is actually using forth for something, rather than just talking about how it should be mroe popular;-)
08:39:49 <Herkamire> yeah :). It's in my svn repository: http://herkamire.homeip.net:3/svn/ana/
08:40:16 <Herkamire> It'll be GPL, although I haven't put copyright stuff in it yet
08:40:51 <Herkamire> davidw: Me too :) I think it's great that some people here are developing forth systems, but somebodys gotta write an application sometime. :)
08:41:20 <Herkamire> I may end up writing a game or some such instead of a viewer
08:41:53 <Herkamire> my current goal is to get it to rotate a cube at an interesting angle
08:42:30 <davidw> ouch, just goes straight to the frame buffer, eh?
08:42:53 <Herkamire> davidw: yes :) I don't want to learn Xlib or how to interface with shared libraries :)
08:44:11 <Herkamire> some porting required (not much... and it's documented :) :))
08:46:08 <davidw> why don't you just write directly to the graphics chip
08:51:13 --- join: tathi (~josh@wsip68-15-54-54.ri.ri.cox.net) joined #forth
08:52:07 <davidw> isn't that what The Leader tells us?
08:54:03 <Herkamire> davidw: I don't have permissions :) :)
08:54:46 <Herkamire> I got the address of the framebuffer on the graphics card from the X11 log, but I get a segfault when I tried to store there.
08:55:51 <davidw> you could unprotect that memory or something
08:56:15 <Herkamire> if I knew how I might be able to.
08:56:26 <Herkamire> hmm... I'm sure as root it would let you
08:56:45 <cleverdra> darn, no Forth in here today?
08:57:02 <Herkamire> I probably won't bother. I'll write directly to the framebuffer on my own OS
08:58:56 <Herkamire> heh :) this server started blocking my bot while I was testing it (presumably because I was connecting and disconnecting a lot :))
08:59:33 <davidw> with zog you could write wherever you want
08:59:41 <Herkamire> I'm suprised it still is
08:59:48 <Herkamire> what's zog?
09:00:23 <davidw> dedasys.com/freesoftware
09:00:28 <davidw> go to projects
09:13:01 <Herkamire> cool
09:13:23 <Herkamire> I'd try it if I had an X86 box
09:13:40 <davidw> ecos runs on a lot of architectures
09:13:52 <davidw> don't know much about booting it on ppc though
09:14:03 <davidw> I'm not about to do a lot of rebooting to try it on mine
09:15:23 <Herkamire> ecos looks cool
09:15:43 <cleverdra> so, uh, what do OUTB and INB do?
09:16:23 <davidw> out byte and in byte
09:16:38 <cleverdra> out to where, and in from where?
09:16:45 <davidw> depends where you want
09:16:46 <davidw> ports
09:17:06 <cleverdra> ah, I see.
09:17:20 <cleverdra> I'm familiar with these under the names pc! pc@
09:18:00 * cleverdra has a simple speaker driver for that.
09:18:12 <davidw> cool
09:18:19 <davidw> I wonder if it works with this...
09:18:26 <davidw> want to send it my way?
09:18:31 <cleverdra> ought to.  One sec.
09:19:25 <davidw> dude, some .mil guy looking at my site got there looking at 'Italy transportation system' on google
09:19:29 <davidw> what are they up to now?
09:21:07 <cleverdra> Mind if I paste?  It's short.
09:21:27 <davidw> please email it so I don't lose it, if you don't mind
09:21:30 <davidw> davidw@dedasys.com
09:22:04 * davidw fires up the x86 box to have a try...
09:22:42 <cleverdra> OK, it ought to be coming your way.
09:25:57 <cleverdra> letsee,
09:26:52 <cleverdra> 'cnt' means 'counter', since what this code does is set the count of one of your three counters and attachs that counter to the speaker (so to speak).
09:27:25 <cleverdra> 'cntwarn' is so named because you have to 'warn' the counter before you set the count.  ha ha.
09:31:28 <davidw> cleverdra: oops, girlfriend arrived... I will try it later... thanks though
09:31:39 <davidw> cleverdra: do I have your permission to put it on the ZOG site?
09:32:14 <cleverdra> sure, go ahead.
09:33:39 <davidw> thanks
09:36:53 <cleverdra> ZOG would certainly be useful for PACs (Publicly-Available-Computers, as in an electronics store).
10:09:16 --- join: kc5tja (~kc5tja@stampede.org) joined #forth
10:54:35 --- quit: davidw (Read error: 113 (No route to host))
10:55:10 --- quit: cleverdra (Connection timed out)
11:21:21 --- join: I440r (~mark4@1Cust56.tnt2.bloomington.in.da.uu.net) joined #forth
11:22:35 <rob_ert> Hey :)
11:25:52 <I440r> hi
11:26:02 <I440r> been coding that text user interface
11:26:05 <I440r> its kinda complex and i dont like the code
11:26:14 <I440r> i might have to rm it later and start over
11:26:51 <rob_ert> :/
11:27:13 <I440r> i do that sometimes
11:27:18 * rob_ert too.
11:27:20 <I440r> not often but its a good habit to get into
11:27:29 <rob_ert> Too many times :P
11:27:41 <I440r> only when i think the code i wrote sucks usually heh
11:28:11 <kc5tja> I do it very often.  In extreme programming, they're called spike solutions, and are heavily encouraged for projects you don't have a solid basis to work from.
11:30:28 <kc5tja> In fact, I'll be coding up a TUI myself pretty soon.  I'm in need of a file management tool that requires some non-trivial screen updates, I think.  Still working on the program's requirements.
11:33:34 <Herkamire> :) good code is rewritten code
11:34:02 <I440r> kc5 thers alot about extreme programming that i have been doing from day 1. not because i read some formal "code like this" document but because it was the way i felt it SHOULD be done
11:34:20 <I440r> i.e. 2 people working on the same project
11:34:32 <I440r> gives you a chance to brainstorm things when needed
11:34:51 <I440r> but i still say commenting is required :P~
11:35:12 <I440r> i consider the time take to comment to be "extra thinking" time
11:35:28 <I440r> if i cant find a decent way to explain the code to you thats when i usually delete it all and start over
11:36:42 <Herkamire> my line-drawing algorithms on rwrite #3 or 4
11:36:52 <Herkamire> s/rwrite/re-write/
11:37:22 <I440r> herk want to look at my old 16 bit vga line drawing routines?
11:37:27 <I440r> they are in x86 asm
11:37:35 <Herkamire> no thanks
11:37:44 <Herkamire> I don't know x86 asm
11:37:51 <I440r> i recon my vga mode 12 routines are about the fastest there are for that mode heh
11:38:33 <Herkamire> I'm pretty happy with my algorithm. I'll port it to PPC asm later.
11:38:50 <Herkamire> probably :)
11:39:13 <Herkamire> first I gotta get a anaglyph wireframe cube spinning on my screen :) 
11:41:16 <kc5tja> I440r: Commenting time is fruitless typing time.  If the reader of the code is familiar with your coding conventions (usually documented elsewhere, especially if you're part of a programming team), then the source alone should be enough to tell the reader what's going on.
11:41:40 <kc5tja> I can't think when I'm typing.  I either think, or I type.
11:42:29 <kc5tja> Also, I've been doing *ALL* of XP for years, except unit testing and pair-programming, since I could remember.
11:42:50 <kc5tja> I refer to it because it's formally documented and verified in the market place, while "we" are not.
11:43:34 <kc5tja> So I would appreciate it if no chest-beating occured every time I mention XP.  There's a very real reason why I do, and it's not just for looks.
11:45:28 <I440r> kc5 err i wasnt chest beating. i just disagree with you on the "no comments" part
11:45:45 <kc5tja> You can disagree with me all you want.  XP is proven.
11:45:54 <I440r> kc5 so is OOP
11:46:02 <kc5tja> What's your point?
11:46:41 <I440r> ive yet to see OOP methodologies produce anything worth a damn
11:46:44 <I440r> ever
11:46:47 <I440r> not a fucking thing
11:47:03 <kc5tja> Boy have you a lot to learn...
11:47:03 <I440r> sure. it works
11:47:13 <I440r> but its 27847823585 gigs bigger than it needs to be
11:47:19 <I440r> specially in that lil old 8051
11:47:21 * kc5tja laughs
11:47:22 <I440r> ugh
11:47:32 <I440r> you gona teach me daddy ?
11:47:40 <n_> jeez
11:47:43 <kc5tja> Are you aware that I write object oriented code in everything I do, including Forth?  Are you aware that the code is incredibly small?
11:47:47 <n_> flamewar going on here?
11:47:59 <kc5tja> I440r seems to have something against OOP.
11:48:00 <n_> not the old OO vs. non-OO again. ;(
11:48:05 --- join: XeF4 (hupcry@12-245-116-85.client.attbi.com) joined #forth
11:48:08 <Herkamire> no shit
11:48:08 <kc5tja> I know -- it gets old really fast.
11:49:38 <XeF4> it just occured to me that custom VLSI is totally unnecessary for the yesterday-mentioned pseudoproject
11:49:48 <n_> but really, i think I440r has had too many bad experiences with C++ and the linke
11:49:50 <n_> like rather
11:50:00 <n_> OO is really a way of thinking of things
11:50:10 <kc5tja> XeF4: How so?
11:50:27 <I440r> there are parts of OOP that are fine
11:50:37 <XeF4> and that it would be better to just put 2 rings of 25Xes around some RAM and hang a DAC off one of the outer leads
11:50:44 <I440r> but people use OOP to the same extend they used GOTO in basic
11:50:47 <Herkamire> I have something against OOP too... but I know it's almost entirely because of C++, so try to keep my mouth shut until such time as I see something cool in OOP
11:50:59 <XeF4> kc5: because the same can be accomplished with vanilla X25s, if only they would exist
11:51:02 <I440r> Herkamire: dont hold your breath
11:51:05 <kc5tja> Herkamire: Smalltalk.
11:51:37 <Herkamire> kc5tja: should I do smalltalk or python?
11:52:13 <kc5tja> Herkamire: Python is very Smalltalk-like in overall capability.  Python if you're just learning; Smalltalk if you want something more advanced.
11:52:38 <Herkamire> kc5tja: I tried squeak, and I was pretty impressed, but I could only stand the interface for about a week and a half.
11:52:46 <n_> heh
11:53:18 <kc5tja> Herkamire: Like I said, it's pretty advanced.  It's not for beginners, no matter what Alan Kay says.  But once you get used to its conventions, it's very consistent (versus X11, which is inundating AND inconsistent)
11:53:51 <I440r> err kc you lost me
11:53:56 <I440r> squeak is a x11 replacement ?
11:54:16 <michaelw> Herkamire: try Objective-C :)
11:54:16 <Herkamire> kc5tja: it was very cool. I was able to fix the scrolled textbox controll that is used _everywhere_ after a couple days
11:54:35 <n_> squeak is its own Operating Environment
11:54:42 <n_> it can run almost anywhere, on top of other OSes and such
11:54:51 <kc5tja> I440r: No.  But Smalltalk (and by extension Squeak) is a self-contained operating environment, like Forth originally was.  Under the hood, it even runs on a stack virtual architecture.
11:55:20 <Herkamire> Objective-C is cool. I haven't done anything serious with it. I made a couple dinky test programs for OS X
11:55:25 <kc5tja> I440r: Smalltalk invented the GUI, and is still decades more advanced than anything Windows or Mac has ever demonstrated.
11:55:44 <I440r> smalltalk is what XEROX invented then ?
11:55:48 <kc5tja> Yes.
11:56:18 <Herkamire> squeak made the gui...?
11:56:27 <n_> squeak is a relateively new opensource implementation of smalltalk
11:56:35 * kc5tja sighs
11:56:36 <kc5tja> No.
11:56:39 <kc5tja> Listen to what I'm typing.
11:56:43 <n_> lol
11:56:45 <kc5tja> Read what I'm saying.
11:56:52 <kc5tja> Smalltalk invented the GUI.
11:56:58 <kc5tja> Squeak is a version of Smalltalk.
11:57:14 <Herkamire> oh.  :) I head what you said, I just thought you were wrong :) :)
11:57:21 <michaelw> kc5tja: tell 'em to 'dict smalltalk' and keep your breath ;)
11:58:48 <Herkamire> I don't have a use for OOP right now.
12:00:10 <kc5tja> Not everybody does.  99% of my Python scripts are definitely not object oriented.
12:00:22 <kc5tja> But when I need object orientation, I really need it.
12:00:42 <kc5tja> Now, because of the inadequacies of C, nearly ALL of my C programs are object oriented.
12:00:53 <kc5tja> (Note that I'm writing in C, not C++ or Objective-C).
12:01:13 <Herkamire> what's an example of something you've needed OOP for recently?
12:01:31 <kc5tja> A test generator for our latest series of chips.
12:02:08 <kc5tja> I treat each network packet as an object, readers and writers to and from storage files are each objects, I emply the decorator design pattern with formatter and scanner objects, etc.
12:02:43 <n_> latest series of chips?
12:02:47 <kc5tja> There is not a single object method which is not called at least twice.
12:02:49 <n_> btw, how's work on Dolphin?
12:02:55 <kc5tja> n_: I work for a semiconductor manufacturer.
12:03:31 <kc5tja> Dolphin isn't making much progress right now.  When I Have time to work on it, I work on it.
12:04:28 <n_> it is still written in C, right?
12:04:30 <kc5tja> I need to finish FS/Forth for it first.
12:04:39 <kc5tja> 0.5 will be written in Forth.
12:05:27 <n_> ah
12:05:39 <n_> doesn't the lack of security in forth disturb you?
12:05:41 <n_> safety rather
12:07:03 <I440r> n_: no it is one of its biggest assets, the damned compiler shouldnt be protecting the user from him/her self
12:07:19 <kc5tja> C is no safer.
12:07:35 <I440r> actually c is less save
12:07:36 <I440r> safe
12:07:37 <kc5tja> And no, it doesn't disturb me either.  
12:07:44 <I440r> forth isnt as subject to buffer overflows
12:07:51 <I440r> it doesnt use ascii z strings
12:07:56 <I440r> it uses COUNTED strings
12:07:59 <n_> but wouldn't you rather use a safe language, or modify forth to be safer?
12:08:10 <n_> (i'm not advocating C or C++)
12:08:24 <kc5tja> Forth's interactivity and ability to switch between compilation and interpretation modes at will eliminates the need for type safety.
12:09:19 <kc5tja> Use of unit tests, run while interpreting, will typically detect a very large number of errors (it will not detect algorithmic errors, but then, neither will C).
12:09:29 <n_> right, hah
12:09:53 <kc5tja> Furthermore, Forth software has a substantially reduced need for local variables and data structures because most everything is kept on the stack.
12:10:08 <kc5tja> Forth is also mathematically provable to be a correct language.  C is not, for it has many side effects.
12:10:25 <I440r> kc5 you lost me on that one
12:10:32 <kc5tja> Forth is a concatenative language.
12:10:34 <I440r> what is mathematically correct ?
12:10:41 <I440r> i.e. what makes forht correct and c not
12:10:50 <n_> you can generate a proof of its safety
12:10:54 <kc5tja> Mathematically correct is when you can prove the correctness of a program or algorithm purely by mathematical induction.
12:13:02 <I440r> kc5 thats proving the program, not the language
12:13:05 <I440r> or am i missing something
12:13:22 <kc5tja> You missed something huge.
12:13:28 <I440r> lol
12:13:43 <kc5tja> We're talking about why I chose Forth over C for implementing the next version of Dolphin.
12:13:53 <I440r> how is the language forth mathematically proven and yet c is not
12:13:54 <n_> not necessarily C
12:14:10 <I440r> is it because nobody bothered to prove c or because c has been disproved ?
12:14:12 <n_> i was hoping you could use something better than forth, or modify forth to be better
12:14:32 <Herkamire> n_: better in what way?
12:14:47 <kc5tja> n_: Why?  Forth has all the features I **WANT** it to have.
12:14:50 <I440r> kc5 you made a statement that forth is mathematically proven but c is not. THAT is what i wanted to understand
12:14:54 <n_> more safety
12:14:56 <kc5tja> Type safety is quite often an illusion.
12:15:23 <kc5tja> I440r: Forth, as a language, is proven by virtue of the fact that it's concatenative.
12:15:48 <kc5tja> You can take a program, split it almost anywhere you want, and you end up with two sub-programs.
12:16:02 <Herkamire> n_: you mean type safety? or like there's no way to crash the interpreter/compiler no matter what garbage code you give it?
12:16:02 <kc5tja> Placed one after the other, you have the complete program again.
12:16:26 <I440r> kc5 you can do that with c too can you not ?
12:16:32 <kc5tja> Nope.
12:16:34 <I440r> isnt c absolutely 100% free form ?
12:16:44 <I440r> can you not reduce any c program to a single line of source ?
12:16:50 <kc5tja> I440r: That's an ambiguous question.
12:16:56 <I440r> kc5 i know
12:17:11 <kc5tja> But that's not what I'm talking about.
12:17:15 <I440r> im not arguing wit you this time, im just trying to understand what you mean by mathematicaly proven heh
12:17:29 * rob_ert says BASIC is the one and only language, and then leaves this discussion alone.
12:17:38 * kc5tja ignores rob_ert
12:17:43 <I440r> lol
12:18:10 <kc5tja> I440r: Consider converting temperature from F to K.  You'd write it something like this : F2K   2C 273+ ;
12:18:16 <I440r> assembler is the only true language there is. any other language has to be reduced to asm at SOME point (except machine forth which IS assembler :)
12:18:28 <kc5tja> Well, what is 2C?  That's: : 2C 32 - 100 180 */ ;
12:18:45 <kc5tja> But given those two definitions, we can eliminate 2C completely:
12:18:55 <kc5tja> : F2K   32 - 100 180 */ 273 + ;
12:18:58 <kc5tja> BUT!!!
12:19:04 <kc5tja> We can now re-factor the code out like this:
12:19:11 <kc5tja> : abc   32 - 100 ;
12:19:19 <kc5tja> : def   180 */ 273 ;
12:19:19 --- join: Speuler (~l@195.30.184.4) joined #forth
12:19:25 <kc5tja> : ghi   + ;
12:19:26 <rob_ert> Hi, Speuler.
12:19:28 <Speuler> g'day
12:19:30 <kc5tja> : F2K abc def ghi ;
12:19:42 <I440r> kc5 ok i understand
12:19:47 <kc5tja> and we'd have ***EXACTLY*** the same program with EXACTLY the same semantics with EXACTLY the same execution performance.
12:19:56 <kc5tja> That's what's meant by concatenativity.
12:20:02 <kc5tja> C doesn't have this property.
12:20:07 --- join: davidw (~davidw@adsl-32-74.38-151.net24.it) joined #forth
12:20:08 <I440r> what your saying is that you can factor every single component part of a word out
12:20:28 <kc5tja> C is what's called an "applicative" language, because you apply functions to data.  In a concatenative program, however, data comes to the function, not the other way around.
12:20:38 <kc5tja> BTW: Lisp is also an applicative language.
12:22:18 <kc5tja> Lisp is also a mathematically proven language (in fact, it started out as a mathematical proof!), which is why I still insist that Lisp is still the ultimate in functional programming.
12:22:49 <I440r> kc5 is it the ability to factor properly in forth what makes it mathematically correct? or are there also other things forth has that c doesnt ?
12:22:49 <kc5tja> Languages like ML and Haskell are mere more complicated attempts to cash in on what Lisp has been offering since 1959.
12:22:55 <I440r> ok... i know the answer to that one heheh
12:23:13 <michaelw> kc5tja: lisp doesn't offer static typing
12:23:27 <michaelw> kc5tja: neither is it purely functional
12:23:29 <kc5tja> The answer is actually both.  Concatenativity is one way to achieve its correctness; good factorization is another.
12:24:01 <kc5tja> michaelw: Lisp does offer static typing, but you need to explicitly "hint" it to the compiler.
12:25:06 <kc5tja> At any rate, this is all a side discussion.  The reason I chose Forth for Dolphin is because I can logically reason about the code, while with C I cannot.
12:25:14 <kc5tja> (or other so-called type-safe languages)
12:25:19 <michaelw> kc5tja: well, let me rephrase: strict static typing, i.e. all types are determined at compile-time.
12:25:48 <kc5tja> michaelw: In the end, it really doesn't matter.  And also "real world" versions of ML and Haskell also are not purely functional.
12:26:46 <michaelw> kc5tja: ML is for sure not. Haskell isn't either, I know, but you have to apply some dirty tricks
12:27:54 <kc5tja> What it boils down to is this: the ultimate programming language is both functional and imperative, object oriented and procedural, structured and non-structured, all at the same time.  It would allow you to use the features of the language that best apply to the particular problem being solved, and would permit free mixing of each technology.  Forth and Lisp are all of these.
12:29:00 <n_> riight
12:29:07 <n_> Java is your answer :))
12:29:11 <michaelw> i agree, at least in the lisp case.  don't know forth enough to judge
12:29:13 <michaelw> NOOO
12:29:27 <michaelw> Java is crap. EOT. :-p
12:29:52 <n_> please, Sun designed Java to fulfill all those requirements.
12:30:03 <michaelw> muahaha
12:30:28 <kc5tja> n_: Java is not imperative.  EVERYTHING in Java sits in a class, and you must instantiate objects of those classes to get anything done.
12:30:32 <kc5tja> n_: Stop trolling.
12:30:46 <kc5tja> n_: I'm not a complete idiot when it comes to programming languages.
12:31:27 <n_> Bill Joy would disagree with you
12:31:35 <kc5tja> Bill Joy can kiss my ass.
12:31:43 <michaelw> mine next :)
12:31:45 <kc5tja> He's the the end-all and be-all of programming languages any more than Chuck Moore is.
12:32:04 <kc5tja> I base my observations of Forth and Lisp on what I've learned through practical experience.
12:32:08 <kc5tja> Forth kicks ass.
12:32:12 <kc5tja> Java sucks ass.
12:32:17 <kc5tja> Lisp kicks ass.
12:32:20 <kc5tja> Scheme sucks ass.
12:32:24 <kc5tja> There's a patterm forming here.
12:32:34 <n_> *gasp*
12:32:50 <n_> Scheme sucks??
12:33:18 <kc5tja> You can't do anything in Scheme without writing huge quantities of definitions, which are more complicated and involved than their equivalents in Lisp.
12:33:52 <kc5tja> Nearly *ALL* of those definitions invoke the auspices of macro processor, which itself is known to be sorely deficient compared to Lisp's.
12:34:05 <michaelw> well, i read some stuff on ultratechnology.com.  it would make some software engineer's heads explode :)
12:34:59 <michaelw> like the stuff about documentation or reuse...
12:34:59 <kc5tja> Contrast this with Forth, where you write, maybe, a couple of definitions, very FEW of which would be immediate words (Forth's answer to macros), and all of which fits on a single piece of paper.
12:35:42 <michaelw> kc5tja: well, does forth scale to big projects?
12:35:48 <kc5tja> Lisp also needs relatively large definitions, but it's all at a higher level of abstraction, so you need fewer of them (fewer even than Forth), so Forth vs. Lisp is a wash -- they're equally powerful.
12:35:57 <kc5tja> michaelw: Yes.
12:36:30 <kc5tja> But in different ways.
12:36:41 <michaelw> like?
12:36:56 <n_> forth scales about as well as assembly, hah
12:37:06 <kc5tja> Well, you have garbage collection in Lisp, but not in Forth.  So a large Forth program will be structured vaguely like a C program in that sense.
12:37:14 <kc5tja> n_: You're trolling again.
12:38:03 <XeF4> but one can add GC to Forth or remove it from Lisp with about equal ease
12:38:12 <kc5tja> The user interface of a Forth program will typically be very close to a normal Forth interpreter.  For example, in VIBE (my block editor, admittedly not a large-scale project), it maps every key-stroke to a Forth word for execution.
12:38:47 <kc5tja> OKAD-II by Chuck Moore is a large-scale "program".  In reality, it's many smaller programs used together.
12:39:11 <n_> interesting
12:39:13 <kc5tja> This represents another method of Forth scaling -- a large collection of small, highly tuned programs working together, instead of a single large program that does everything.
12:39:18 <n_> i suppose i was wrong
12:39:22 <n_> i stand corrected.
12:40:06 <kc5tja> And in both Lisp and Forth cases, the language's runtime and interpretter (and in Forth's case, the compiler) is often exposed to the user directly -- they can develop their own extensions to the software after-market.
12:40:24 <kc5tja> This is not to say that I would choose Forth for all applications every time.
12:41:00 <kc5tja> If I knew ahead of time that my software was going to need dynamic memory management and use it heavily, to the point where it dominates most of the code, I'd sure as heck use Lisp.
12:41:08 <I440r> kc5 me either, thers alot i would never do in forth
12:41:41 <kc5tja> OTOH, if I knew my code was going to be brutally simple, without the need for complex memory management, I'd probably choose Forth (as I have with Dolphin).
12:41:51 <kc5tja> In between those two extremes, it's pretty much a wash.
12:42:21 <kc5tja> A HUGE number of the nation's telephone switches all run on Lisp, by the way.
12:42:34 <n_> really?
12:42:38 <n_> how do you know this:
12:42:39 <n_> ?
12:42:50 <kc5tja> Despite C being designed by AT&T ("Bell Labs" back then), it has a venerable minority in this application.
12:43:13 <kc5tja> Worked 10 years in the ISP and network engineering industry.
12:44:07 <kc5tja> When you hear, "We're sorry; the number did not go through.  Please try your call again later," on the phone, that's Lisp talking to you.  :)
12:47:55 <n_> heh
12:47:56 <n_> really?
12:47:59 <n_> that's pretty neat
12:48:06 <kc5tja> Yup.
12:48:20 <n_> i didn't know LISP had such a sexy voice :)
12:48:33 <kc5tja> Well, it's playing audio clips, of course.  :)
12:49:04 <kc5tja> Forth's biggest field of application right now, I believe, is in satellites and in the OpenFirmware environment.
12:49:08 <n_> what kind of hardware does the switching system run on ?
12:49:36 <kc5tja> They're usually proprietary (because they can get more profit that way), but they're not really more complicated than a typical PC.
12:52:23 <Herkamire> one reason I really like forth, is that when I want to do something (like say draw colored lines on the screen) I spend my hours finding out how that works.
12:52:37 <Herkamire> and how to write a good algorithm
12:52:48 <Speuler> Herkamire: good point
12:53:00 <I440r> bongo !!! when did you show up :P
12:53:09 <michaelw> kc5tja: well, ever heard of Erlang? :)
12:53:09 <Herkamire> whereas in C I spent probably three times as long trying to get GTK to draw in red.
12:53:18 <Speuler> t'is also my opinion that forth helps learning about all that programming stuff
12:53:23 <Speuler> hi I440r
12:53:31 <Speuler> about when i said  "g'day"
12:53:35 <Herkamire> I consider the knowledge gained in the forth version much more useful.
12:53:36 <kc5tja> Sometimes in doing that, you can find optimizations that allow you to take shortcuts too, which makes your code that much faster and easier to maintain too.
12:53:45 <I440r> speuler lol i was being my usual observant self :)
12:53:49 <Speuler> 9 screens ago
12:54:06 <Speuler> (20 lines/screen)
12:54:11 <kc5tja> michaelw: Yes, but I don't know much about it, so I tend to stay away of getting involved with Erlang discussions.
12:54:51 <Speuler> Herkamire: rather then using those closed-source 3rd-party libs,
12:55:24 <Speuler> you are prompted into workig it out yourself.
12:55:36 <Herkamire> kc5tja: definately :) BTW with my anaglyphs, I realized that in my 15 bit color mode the red bits are all in a different byte than the blue bits :) so blending the colors is transparent and just as fast (or faster) than overwriting :)
12:55:38 <Speuler> that's an excellent learning opportunity
12:56:20 <Herkamire> it depends on your interests :) I want to be able to write my own operating system.
12:56:25 * kc5tja agrees with Speuler and Herkamire 
12:56:44 <kc5tja> However, when time to market is the limiting factor, you often are forced to use closed-source solutions.  :(
12:56:47 <Herkamire> some, are more interested in their code working on other people's computers :)
12:57:03 <kc5tja> Forth is more portable than poeple realize, I've found out.
12:57:21 <Herkamire> kc5tja: I'm sure.
12:57:21 <Speuler> more portable than people are
12:57:53 <Herkamire> I like the model of making something easily portable, rather than making it all-compatible (doing all sorts of compiletime checks etc)
12:58:15 <I440r> and in those places where its NOT as portable its easier to convert :)
12:58:44 <I440r> herk if you mean #ifdefine this-architecture or that-architecture i disagree
12:59:05 <Herkamire> forth!
12:59:06 <I440r> all those conditional compilations visually clutter your sources all to hell:)
12:59:38 <Herkamire> I'm saying I _don't_ like those
13:00:27 <Herkamire> I prefer to isolate the architecture dependant bits, document what they do, and what is most likely to need changing.
13:01:54 --- join: Fractal (rzdizri@h24-77-171-228.ok.shawcable.net) joined #forth
13:02:17 <I440r> herk good. like in isforth seperating out linux.1 and fbsd.1 
13:03:06 <Fractal> I440r : You wanna OpenBSD shell so you can devel isforth there? It shouldn't be to different from FreeBSD.
13:03:29 <kc5tja> Exactly my experience too.
13:04:06 <I440r> fractal openbsd uses pass by register for syscalls i believe, i want a fbsd, openbsd, netbsd etc etc port of isforth
13:04:19 <Herkamire> my main problem with all the #defines is that you open up the source code and you can't even tell what's being compiled. And I asume it makes it harder to test
13:04:37 <I440r> fractal does openbsd have man 2 like linux does?
13:04:39 <Fractal> I440r : Cool. I'll give you a shell then. Just don't r00t me.
13:04:47 <Fractal> Um. Let me check.
13:04:52 <XeF4> I440r: may I ask why you are porting Isforth before it metacompiles?
13:04:59 <I440r> herk thats how i feel about it too :)
13:05:13 <Fractal> Yes, it does.
13:05:20 <I440r> fractal i was rooted once. i lost 10 years of assembler source files (later recovered most of it)
13:05:20 <I440r> so i would never root
13:05:32 <kc5tja> Macros in C are fairly difficult to test.  With a consistent unit-testing system in place though (such as my very own CUT, *wink* *wink* *nudge* *nudge*) it isn't so difficult.
13:06:32 <I440r> fractal cool! so man 2 mprotect gives the man page for the syscall ?
13:06:34 <Herkamire> I'm done coding in C I think.
13:06:39 <I440r> freebsd doesnt
13:06:39 <I440r> grr
13:06:42 <Herkamire> I'd do it if someone were paying me :)
13:06:45 <Speuler> I440r: 10 years w/o backup ?
13:07:13 <I440r> bongo my father had an old copy on his machine
13:07:15 <Herkamire> I mostly just fix or tweak other people's C programs.
13:07:19 <Fractal> I440r : telnet or ssh to irc.hcsw.org  login i440r, I'll msg you the pass.
13:07:29 <I440r> and i never back shit up, i cant keep track of the backups :)
13:07:34 <kc5tja> Herkamire: I am being paid to code in C.
13:07:36 <Fractal> Just don't r00t me. :)
13:07:40 <kc5tja> Herkamire: So I do it.  :)
13:08:09 <I440r> ok i got it
13:08:15 <Fractal> Cool.
13:08:36 <kc5tja> I440r: My backups usually consist of paper hardcopy of my source code, stored in file folders.  Though I've got a CD-burner now, so I'll probaby start to uburn CDs now.
13:08:55 <n_> paper hardcopy?
13:08:56 <n_> heh
13:09:03 <tathi> kc5tja: you have a unit-testing system for C?
13:09:09 <Herkamire> I440r: backup your work! put the date in the filename if you want.
13:09:32 <I440r> herk lol heh
13:09:57 <n_> I440r: just use the Sourceforge compile farm
13:10:03 <n_> to test your bsd/linux/osx/solaris apps
13:10:31 <kc5tja> tathi: Yes.  It currently requires Python to run, as it's used to emit a .c driver program for it.  But the next version will be written completely in C itself.
13:10:46 <I440r> n_ sourceforge is obfuscated
13:10:55 <I440r> i deleted my sf account because im a coder NOT a manager
13:11:03 <I440r> i dont want all that bulshit managerial babysitting you have to do with sf
13:11:09 <n_> uh
13:11:11 <kc5tja> tathi: And it'll be its own example, in that CUT 3.0 will use CUT 2.0 as its unit testing tool.
13:11:12 <n_> whatever
13:11:20 <n_> you've got some issues, man
13:12:28 <tathi> kc5tja: publicly available somewhere?
13:13:02 <michaelw> http://sourceforge.net/projects/cut/
13:13:19 <kc5tja> tathi: Yes, but not very well publicized or documented.  I think documentation is what it needs sorely right now.  But I've written it to help me out at work, and I've not had a lick of free time to document it adequately yet.
13:13:32 <kc5tja> michaelw: That's it.  :)
13:14:17 <tathi> thanks, just found it anyway :)
13:18:31 <Fractal> Ok, I'm compiling nasm now...
13:22:21 <Fractal> .profile and .login might do stuff with bash, I'm not sure.
13:23:06 <Fractal> Heh. I dunno... Don't even know what it does...
13:23:12 <Fractal> So I don't miss it much. :)
13:24:26 <I440r> .profile and .bash_profile are what bash uses :)
13:24:49 <Fractal> Oh. You got a .profile, no?
13:25:02 <Fractal> Er... Damn BX, I dunno why I started talking in here...
13:25:12 <I440r> yes
13:25:27 <I440r> heheh
13:25:34 <kc5tja> .bash_profile is executed for interactive shells.
13:25:35 <I440r> bitchx hates everyone :P
13:25:48 <Fractal> Ah. That'd make sense...
13:25:49 <kc5tja> .bashrc is executed for non-interactive shells.
13:26:17 <kc5tja> It took me many years to find this out.
13:26:17 <I440r> yes
13:26:19 <kc5tja> :)
13:26:28 <I440r> .bash_profile or .profile is executed for login shells
13:26:46 <XeF4> .bash_profile is executed for login shells.  .bashrc also for non-login interactive shells
13:27:02 <I440r> usually .profile is what root has. .bash_profile is what users have
13:27:08 <kc5tja> And here's something you might not be immediately aware of: the screen tool DOES NOT launch an interactive BASH -- it launches a non-interactive BASH.  Strange, but the execution of configuration files do not lie.  :)
13:28:02 <XeF4> really: so that is not readline I'm editing my bash commandline with under screen?
13:29:28 <kc5tja> no, it is.
13:29:47 <kc5tja> But screen works by launching a Bash session under a pipe, instead of directly to a console.
13:30:36 <kc5tja> Pipes do not have teh same interface as consoles do (for example, they don't support out-of-band signals like cooked CTRL-C), etc.
13:31:46 <kc5tja> Recently someone asked if they loved or hated Unix, and I was indifferent.  This is one of the things that I DO NOT like about Unix -- sockets, pipes, and interactive console file handles all have different interface requirements.
13:37:04 <I440r> kinda throws a spanner into the "everything is a file" concept
13:40:06 <kc5tja> Well, yes and no.  All the things which can be implemented as a filehandle do successfully implement read()/write()/close().
13:40:12 <kc5tja> This is logical.
13:40:24 <kc5tja> Even Dolphin follows a somewhat similar system.
13:40:26 <kc5tja> BUT...
13:40:40 <kc5tja> Interfaces which derive from other interfaces must be consistent.
13:41:13 <kc5tja> And this is where Unix's technique fails.  There's no interface hierarchy.
13:56:24 --- join: Chris-- (chris@dialin-212-144-128-145.arcor-ip.net) joined #forth
13:56:43 --- part: Chris-- left #forth
14:09:32 <Speuler> g'd night.
14:09:52 <Speuler> will hop over to the hippies :)
14:10:15 --- quit: Speuler ("caphuso: first time ive seen that :)")
14:13:33 --- quit: XeF4 ("pois")
14:19:32 <tathi> kc5: /quit laters...
14:19:36 <tathi> grrr
14:19:42 --- quit: tathi ("leaving")
14:20:21 --- quit: Herkamire ("leaving")
14:22:54 <kc5tja> Heheh :)
14:23:05 <kc5tja> Anyway, I'm temporarily QRT myself; food time.
14:23:07 --- nick: kc5tja -> kc-food
14:23:09 <I440r> brb
14:41:25 --- quit: I440r (No route to host)
14:42:57 --- join: CrowKiller (Vapo_Rulez@cnq5-233.cablevision.qc.ca) joined #forth
14:45:07 --- nick: kc-food -> kc5tja
14:53:10 --- join: I440r (~mark4@67.241.42.125) joined #forth
14:53:17 --- quit: I440r (Read error: 104 (Connection reset by peer))
14:53:18 --- join: I440r_ (~mark4@1Cust125.tnt2.bloomington.in.da.uu.net) joined #forth
15:13:52 --- join: Atlas (fare@samaris.tunes.org) joined #forth
15:14:45 <I440r_> hi fare :)
15:15:09 <Atlas> :)
15:17:16 <I440r_> trying to get isforth to work on fractals openbsd box
15:17:32 <Atlas> fractals?
15:17:41 <Fractal> y0
15:17:53 <Atlas> what are difficulties in porting isforth?
15:18:22 <I440r_> not much.
15:18:37 <I440r_> just need to get the syscall parameter passing fixed for each OS and the syscal numbering
15:18:37 <rob_ert> Hrm.
15:18:46 <rob_ert> When will IsForth do floating-point? :)
15:18:58 <I440r_> syscall number X is syscall number Y etc
15:19:12 <I440r_> robert as soon as it has an assembler that allows floating point instructions
15:19:23 <I440r_> i doubt very seriously if ill implement a floating point stack
15:19:52 <rob_ert> Hmm..
15:19:58 <Atlas> shouldn't you be linking to the libc?
15:20:02 * Atlas ducks
15:20:09 --- nick: Atlas -> Fare
15:28:05 <I440r_> err isforth isnt linking to ANYTHING heh
15:28:11 <I440r_> SPECIALLY not libc
15:28:56 --- join: Herkamire (~jason@ip68-9-58-81.ri.ri.cox.net) joined #forth
15:29:05 <rob_ert> Hi Herk.
15:29:28 <Herkamire> hy :)
15:31:46 <n_> I440r_, are you the guy that once said he would only code in C if he was being paid $40/hour? :)
15:31:57 <rob_ert> Haha
15:31:59 <rob_ert> Yes, he is.
15:32:05 <I440r_> n_ im NOT coding in c
15:32:12 <Herkamire> I would too! pick me pick me :)
15:32:53 * kc5tja is being paid for his C programming services, though not at that rate.  It's highway robbery, I tell you... ;D
15:33:21 <Fare> what's highway robbery?
15:33:34 <Fare> "well, of course I program in C!"
15:33:42 <n_> I440r_: so Forth is the only language you'll ever code in again? :/
15:33:44 <kc5tja> That I work with C for such low pay....  :)
15:33:56 <kc5tja> Of course I'm being facetious.
15:34:01 <I440r_> unless paid enough its forth/asm for me all the way
15:34:15 <Fare> forth(" DUP . ")
15:34:15 <n_> wow, what a purist
15:34:29 <I440r_> fare its not highway robbery if your a contractor
15:34:30 <I440r_> erm
15:34:32 <I440r_> no it is 
15:34:41 <I440r_> they are robbing me for paying me that small an ammmount
15:34:51 <Fare> indeed
15:35:55 --- quit: dsmith ("later..")
15:35:56 <I440r_> ok if i do 
15:36:02 <I440r_>  foo (x,y,z);
15:36:09 <I440r_> which parameter is pushed first ?
15:36:12 <I440r_> x ?
15:36:13 <I440r_> or z
15:36:35 <rob_ert> z :D
15:36:39 <rob_ert> At least with gcc-.
15:36:46 <rob_ert> Not sure if it's standardized.
15:38:06 <Fractal> What ISN'T standardized is in what order x,y,z will be evaluated before passing.
15:38:21 <Fractal> Dunno about the passing order.
15:38:54 <I440r_> i know if you do foo(bar(), bam());
15:38:57 <rob_ert> I think I saw TC for DOS using the opposite order.
15:39:01 <rob_ert> Might be wrong, though.
15:39:02 <I440r_> bam is called first, then bar then foo
15:39:12 <Fare> I440r: that depends on the architecture
15:39:23 <I440r_> fare x86 heh
15:39:29 <Fare> on i486, first parameter is first popped
15:40:18 <kc5tja> Many RISC boxes do not have hardware stacks, so parameters can be evaluated in any order.  The specific layout in memory is determined by the operating system's ABI specification.
15:40:39 <Fare> Fractal: I think in Common LISP, it is standardized
15:44:18 <n_> er
15:44:27 <n_> what risc architectures do not have support for hardware stacks??
15:45:32 <kc5tja> Pretty much all of them.  PowerPC, MIPS, any IA-64 processor, SPARC (register stack isn't what I'm talking about here), etc.
15:46:25 <Fractal> Fare : Yeah, it may be...
15:46:26 <Fractal> I dunno.
15:46:29 <CrowKiller> the hardware stack is much more suited to pass return values instead of data
15:48:00 <CrowKiller> I tried to design a one stack forth one time and it was too complicated to get any performance gain, but its pure speculation only, never tested it
15:48:03 <kc5tja> CrowKiller: I think it's suitable for both.  I never did like the use of the stack to pass parameters, and CPU registers for results.
15:51:04 <CrowKiller> I was saying that because I was thinking of the chip im writing code on now,; the AT90S1200 only got a three level deep hardware stack that can only be used by rcall/ret
15:53:42 <CrowKiller> it bring me to a question
15:53:57 <CrowKiller> im trying to send data at 115200 using a software routine down the wire
15:54:08 <CrowKiller> but im running at 4MHz
15:54:22 <CrowKiller> so 4MHZ/115200 = 34.72222 etc
15:54:33 <CrowKiller> the error rate is too big
15:54:51 <CrowKiller> so the question is do someone here did a variable bit length uart or something like that
15:55:22 <CrowKiller> like doing a loop 34 cycles long, then 33, then 34 then 34 then 33 etc etc etc
15:55:32 <kc5tja> I'm not sure I understand the question.
15:55:37 <kc5tja> Can you give more specific example?
15:55:49 <kc5tja> So far it sounds like you're looking for pulse width modulation of data.
15:55:52 <CrowKiller> I want to code a software uart on a 4MHZ system
15:55:57 <CrowKiller> i dont have any other crystal
15:56:12 <CrowKiller> i cant get 115200 bps
15:56:21 <CrowKiller> the fixed bit rate of the gba in multiboot mode
15:56:45 <kc5tja> OK...
15:57:24 <CrowKiller> i will find by myslef if its possible, I was asking just in case i could save some time
15:57:25 <kc5tja> So you want to amortize the 0.72222 cycle error across multiple loop implementations?
15:57:33 <CrowKiller> exactly
15:58:20 <kc5tja> I'm not familiar with any software implementation that does that.  If you keep the amortizations close to each other, it should be possible.  
16:02:18 <CrowKiller> ill do it using a cad package
16:02:57 <CrowKiller> to do graphical simulation
16:03:01 <kc5tja> Let me know how it works out.  I may have some use for the technique for some of my ham radio experiments.
16:03:34 <CrowKiller> every hobbist might want to use this, so they can always use 4MHZ crystal to get practically any baud rate
16:03:59 <CrowKiller> MHz
16:05:29 <CrowKiller> do someone here worked with Rhinoceros before?
16:05:37 <CrowKiller> the CAD software?
16:05:42 <kc5tja> Nope; never heard of it.
16:06:00 <CrowKiller> www.mcneel.com
16:06:03 <CrowKiller> impressing stuff
16:06:10 <CrowKiller> and it cost a lot less than autocad
16:06:47 <CrowKiller> The version I use is quite old, but I cracked it myself, so i dont want to upgrade ;ppp
16:08:57 <CrowKiller> v1.1, the version 2.0 is out and also their Flammingo renderer, steady improvement curve along the developpement of this product, really great value
16:09:37 <kc5tja> Cool.  I'll have to check it out when I have more time.
16:09:44 <kc5tja> Windows only, or is it available for Linux as well?
16:10:04 <CrowKiller> windoze only i think
16:10:12 <kc5tja> Ahh
16:10:30 * kc5tja ponders a mechanical CAD written in Forth.  I wonder what it'd look like.
16:10:47 <CrowKiller> *Rhino will NOT be ported to any other operating system, but Rhino does run on
16:10:47 <CrowKiller>                                    Apple Macs with Virtual PC
16:11:49 <CrowKiller> they used rhino to model props for the SpiderMan movie
16:13:44 --- quit: Fare (Read error: 113 (No route to host))
16:18:40 <kc5tja> Oh oh....had one chip failure.
16:18:54 <kc5tja> This can't be good.  Quite likely a bug in the software...
16:19:16 <kc5tja> And money says it'll be unbelievably difficult to track down too... :(
16:25:47 <kc5tja> Only one failure in 50 tests...definitely suggesting a software bug.
16:25:59 <kc5tja> This makes me feel somewhat better... :)
16:28:04 <CrowKiller> hehehe ;p
16:28:51 <kc5tja> The life of a semiconductor verification technician... *sigh*
16:29:01 <kc5tja> Use broken tools to make non-broken chips.
16:30:03 <I440r_> kc5 when i worked for joy mining machenery they wanted me to write the code to display a message on the lcd and restore what was there previously. being write only i had to double buffer it.
16:30:11 <I440r_> every time the code ran the whole system died
16:30:19 <I440r_> it seems that since the day the system went into production there was a bug in the code that cause the entire system to just HALT
16:30:47 <I440r_> nobody knew what caused it. it was totally random. the code had been in use for 10 years and nobody could find it because they couldnt MAKE it happen
16:30:53 <I440r_> heh
16:31:28 <I440r_> ive no idea why my code MADE it happen but it did. everyone of corse was blaming MY code - but i traced it to a DMA read.
16:31:45 <I440r_> the same register you write your config for a dma read had a bit in it that said "HALT processor"
16:31:45 <I440r_> lol
16:31:52 <CrowKiller> lol
16:31:54 <I440r_> gotta love bugs like that :)
16:32:21 <kc5tja> Yeah, intermittent faults are nasty.  I can sympathise with that.  At least our test environment keeps failed tests so we can repeat them, but that is no guarantee that they'll reproduce the bug.
16:32:38 <I440r_> everyone was pointing their finger at MY CODE and laffing at me
16:32:38 <I440r_> lol
16:32:59 <I440r_> ya
16:33:27 <kc5tja> See, this is why I like my code to be mathematically provably correct.  Then I can point my finger instead, and say, "Look, my code cannot be the problem.  Look elsewhere."
16:33:45 <I440r_> i didnt fix the bug, im not even sure my boss did either. at least i found out what was causing the system to hang. 
16:34:06 <I440r_> still needed to find out who was causing that bit to be set when it shouldt have been
16:34:16 <kc5tja> At times like that, you need an in-circuit emulator, or at least, something comparable.
16:34:52 <I440r_> i once had a professor of mathematics tearing his hair out for a week trying to disprove my trisection of an angle :)
16:35:12 <I440r_> hehe
16:36:34 <I440r_> he did it in the end. so did i but not mathematically
16:37:25 <I440r_> its easy to divide a LINE into 3 equal sections but impossible to divide an arc into 3 equal sections
16:38:01 <CrowKiller> i call it pi/3
16:38:10 <CrowKiller> or pi/3 scaled
16:38:15 <CrowKiller> no?
16:38:50 <CrowKiller> you mean easy geometrically?
16:39:00 <kc5tja> Using something like a compass.
16:39:18 <I440r_> im talking about the construction of a trisection of an angle
16:39:27 <I440r_> thats using a compas and a straight edge only
16:39:29 * kc5tja did figure out how to trisect a pure circle, but not an arbitrary arc.
16:39:39 <I440r_> its been mathematically proven to be impossible
16:40:01 <I440r_> a complete circle is easy. even if you dont know its centre.
16:40:19 <kc5tja> Not really, you have to find its center first, and that's the hard part.
16:40:33 <I440r_> draw a line through the circle anywhere. the normal to that line will go through the center
16:40:39 <I440r_> nope its easy
16:40:44 <I440r_> draw a second line thru the cirle
16:40:51 <I440r_> draw the normal to that line too
16:40:58 <I440r_> the 2 lines will intersect at the center
16:41:08 <I440r_> you now have the circles radius
16:41:56 <kc5tja> I was about to reply to the first line, but your follow-ups came quickly enough that I didn't have to.  I was about to object to your use of the term "center".  But really, I don't see a better way to express it yet.
16:42:16 <kc5tja> But yeah, that makes sense.
16:42:20 * kc5tja will have to experiment with that
16:42:27 <CrowKiller> yeah me too
16:42:38 <CrowKiller> ill try ir roght away
16:42:42 <I440r_> heh
16:43:07 <CrowKiller> let say i have a circle in front of me, i have that straight edge and my compass
16:43:07 <I440r_> what was wrong with my use of the word center ?
16:43:18 <CrowKiller> i draw a line intersectin g the circle
16:43:21 <I440r_> draw a line through the circle. close to the edge
16:43:34 <I440r_> it will cut the circle at 2 points
16:43:48 <CrowKiller> using the compass i find the middle of the line
16:44:03 <I440r_> put compas point on those 2 locations and draw 2 arcs either side of that line
16:44:12 <kc5tja> Right.  But center here has two meanings.  You have the geometric center of the circle, and you have the interior of the circle.  The chord doesn't necessarily go through the center, though it does cut through the interior.
16:44:13 <CrowKiller> and i draw a linbe between the two intersections of the 4 arcs
16:44:28 <I440r_> draw from the arc interesctions to the arc intersections and THAT line wil pass through the centre of the circle
16:44:38 <I440r_> yes
16:44:47 <I440r_> do that twice and you have found the centre of the circle
16:44:48 <kc5tja> I was confused by what you meant by "center," that's all.
16:44:55 <CrowKiller> whoa lol, its my breakthrough geometric discovery of the day
16:44:58 <I440r_> and therfore its radius
16:45:21 <I440r_> heh
16:45:29 <Herkamire> geometry is cool!
16:45:33 <I440r_> i cant do mathematical geometry but if i can SEE it i can compute it heh
16:46:03 <I440r_> like (x+y)^2 = x^2 + y^2 + 2xy
16:46:16 <I440r_> draw a line of size x. extend it by size y and create a square
16:46:25 <I440r_> you have one square of size x^2
16:46:25 <I440r_> another of size y^2
16:46:34 <I440r_> and 2 oblongs of size x*y
16:46:45 <CrowKiller> i saw this last year in math
16:46:53 <CrowKiller> about factorizing expressions
16:47:26 <I440r_> i cant factorize them unless i can visualize what they are describing
16:47:53 <I440r_> im very good at visualizing things, specially in 3 d
16:48:08 <I440r_> i used to be able to do a rubics cube in about 4 looks :)
16:48:15 <I440r_> look at cube, look away. do alot of moves
16:48:24 <I440r_> repeat 3 or 4 times
16:48:24 <I440r_> and be solved
16:48:27 <Herkamire> in 7th grade our teacher had us draw a right triangle, then a square using the hypoteneus as one side.
16:48:42 <kc5tja> I am puzzle impaired; I can't solve physical puzzles like that at all.
16:49:15 <CrowKiller> rubik cube means eternal enslavement to me lol
16:49:54 <Herkamire> then we drew lines through the square (i forget were) and cut it up and fit the pieces. then we made two more squares, each having one of the other sides of the tryangle as sides. then we fit the pieces of the big square into the little ones :)
16:50:06 <I440r_> kc5 the rubics cube is actually quite easy if you memorize the sequence of moves it takes to move a specific piece from a specific location to another without moving any other pieces
16:50:38 <kc5tja> I have to do that first, and that's the problem.
16:50:50 <CrowKiller> herkamire: i see what your talking about, like a square with anotehr square inside but rotated
16:50:54 <I440r_> actualy you can swap 2 edge pieces, rotate 3 edge pices or rotate 3 corner pieces
16:50:54 <I440r_> thats all you need to know how to do.
16:50:55 <CrowKiller> we did the same ;p
16:50:57 <I440r_> well. that and how to re-orientate a piece or 2 :)
16:51:07 <I440r_> you can have 2 edge pieces in the right location but the wrong way up
16:51:07 <I440r_> etc
16:52:12 * kc5tja also used to have Rubik's Pyramid too -- that was a joke.  I played with it for about three hours, and never touched it again. It made a great paperweight.
16:52:31 <kc5tja> And I thought the cube was bad enough.  :(
16:52:40 <I440r_> well i gotta get offline to finish coding this TUI of mine
16:52:52 <I440r_> the pyramid wasnt invented by rubic tho
16:53:03 <kc5tja> I know, but it was marketed with his name for awhile.
16:53:16 <davidw> the 80ies....
16:53:17 <I440r_> the pyramid was easy too heheh
16:53:17 <I440r_> i always wanted a 4*4 rubics cube
16:53:24 <I440r_> they did exist too
16:53:24 <I440r_> but i never got one
16:53:42 <kc5tja> I've seen one; they (used to) have them at our local game shop.
16:53:52 <kc5tja> Not sure if they still do.
16:54:00 <I440r_> bbl. time to pretend i know how to code :P
16:54:04 <kc5tja> Hehe :)
16:54:06 --- quit: I440r_ (Read error: 104 (Connection reset by peer))
16:59:44 --- join: tathi (~josh@ip68-9-68-213.ri.ri.cox.net) joined #forth
17:00:24 <CrowKiller> hi
17:01:26 <kc5tja> re tathi
17:01:41 <tathi> hey
17:02:09 <tathi> I like your CUT, btw
17:03:01 <kc5tja> Thanks.  :)
17:03:14 <kc5tja> It's quite a difference from, say, C++Unit or some other SUnit-derived package.
17:03:58 <tathi> yeah
17:04:56 <tathi> enough to make life a lot easier though :)
17:09:01 <kc5tja> :)
17:09:41 <CrowKiller> CUT?
17:09:52 <kc5tja> C Unit Test tool.
17:10:50 <kc5tja> Uses compile-time code generation to compensate for C's lack of introspection.
17:10:59 <davidw> interesting
17:14:15 <CrowKiller> Atmel application notes are some of the finest around
17:17:18 <CrowKiller> i looked for info about software uarts
17:17:45 <CrowKiller> and I saw that their routines for a polled 115200 bps uart only has an error percentage of 0.8
17:19:04 <kc5tja> I wonder what the allowable error rate is for those kinds of speeds.
17:19:52 <CrowKiller> 1%
17:36:23 --- quit: davidw (Read error: 113 (No route to host))
18:03:39 <CrowKiller> im back
18:03:47 <CrowKiller> after doing the drawing
18:04:00 <CrowKiller> and exploiting my brain's visual computing advantage
18:04:14 <CrowKiller> i can see that at every 1.5 mhz
18:04:27 <CrowKiller> the little 115200bps step align almost perfectly
18:04:48 <CrowKiller> to a 0, 16% percent of error!
18:06:53 <CrowKiller> ill try to beat atmel on a uart implementation scheme im sure i can get something better
18:20:42 <CrowKiller> hmmm how to emulate a 1.5MHz using a real 4MHz clock...?! i refined the definition of the problem i had before, ill try to come up with some code soon
18:21:34 <kc5tja> Not sure.  You may still need to amortize the error across multiple I/O transactions.
18:21:58 <CrowKiller> the transaction is 18 bit long
18:22:09 <CrowKiller> 1 start bit 16 data bits 1 stop bit
18:22:45 <CrowKiller> the timming trimming start at the start bit and end at the stop bit
18:23:40 <CrowKiller> the transaction is simply send a 16 bit word to gba then receive one, could be called symmetric
18:25:53 <kc5tja> I don't think you're understanding what I'm saying.
18:25:57 <kc5tja> Yes, you can do that.
18:26:02 <kc5tja> That's a protocol issue.
18:26:30 <kc5tja> Forget it.  I can't explain it.
18:29:29 <CrowKiller> i redefined the word "transaction"  2 times in my previous statements, im entirely misunderstandable, what a shame ;p
18:29:45 <kc5tja> No no, I really can't explain it.
18:29:49 <kc5tja> I just lack the vocabulary.
18:30:06 <kc5tja> I swear, the older I get, the worse my language becomes.
18:30:17 <kc5tja> And I'm a native speaker.  I wonder if I have Alzeimers or something.  :(
18:31:33 <CrowKiller> me too its been doing that recently
18:31:50 <CrowKiller> some new age folks attribute it to "heightened energy levels"
18:32:39 <CrowKiller> i say some but in fact
18:32:43 <CrowKiller> they are right
18:32:54 <CrowKiller> the energy level increase is in fact information
18:33:00 <CrowKiller> there's more information everywhere
18:33:18 <CrowKiller> anyway maybe our brains can't keep up ;p
18:33:37 <kc5tja> It really can't.
18:33:42 <CrowKiller> they attribute it generally to the global consciousness
18:33:48 <kc5tja> If it could, we wouldn't have "specializations" in the marketplace.
18:34:33 <CrowKiller> like every individual form a whole "consciousness" or amount of knowledge
18:34:44 <CrowKiller> anyway its weird
18:35:02 <CrowKiller> the weirdest thing i've read is about monkeys on isle somewhere
18:35:07 <CrowKiller> i think it in Japan
18:35:20 <CrowKiller> there was many isolated isles
18:35:25 <CrowKiller> with monkeys on each
18:35:38 <CrowKiller> they were studied by biologists independently
18:35:43 <CrowKiller> on one isle
18:35:54 <CrowKiller> monkey developped the behavior of washing their food
18:36:07 <CrowKiller> this habbit slowly gained acceptance
18:36:16 <CrowKiller> and suddendly paff
18:36:20 <CrowKiller> from one day to another
18:36:39 <CrowKiller> when the "critical mass" was attained,every monky on every other isle were washing their food
18:37:12 <CrowKiller> it puzzled the biologists a lot
18:37:32 <CrowKiller> its been a long time since ive read about this, anyway ;p
18:38:34 <kc5tja> Cool, I guess.
18:40:40 <kc5tja> Man, I want to make my own compressed air engine at some point.  
18:41:00 <CrowKiller> a french man did it I think
18:41:08 <kc5tja> The Tesla turbines work great, but I think I want to move on to Wankel now.
18:41:22 <CrowKiller> i dont recall exactly but i think someone got one working
18:41:38 <kc5tja> I know what you're talking about -- http://www.mdiaircar.com I think
18:42:42 <kc5tja> Nope, that's not it.  Googling.
18:42:51 <CrowKiller> me too ;p
18:43:13 <kc5tja> Ahh, http://theaircar.com
18:47:24 <CrowKiller> first time i see a motor like that
18:49:27 <kc5tja> Yeah, it's not too impressive though -- it's all just common sense.  However, I'm disappointed at the use of reciprocating pistons though.  
18:50:36 <kc5tja> A Wankel would have saved the author the effort of developing the connection rod system that he uses (ingenious, but is it really necessary?).
18:50:49 <CrowKiller> hmmm i know another method
18:51:00 <CrowKiller> the two pistons are opposed
18:51:10 <kc5tja> You still have a connection rod.
18:51:12 <CrowKiller> like this <P r P>
18:51:18 <CrowKiller> yes
18:51:21 <CrowKiller> but the r
18:51:23 <CrowKiller> the rotor
18:51:31 <CrowKiller> is only a quarter of a rotor
18:51:39 <CrowKiller> so when the first piston expands
18:51:43 <CrowKiller> the other is compressed
18:52:04 <kc5tja> Right.  (The thing it's driving is called a crank, BTW, not a rotor.  :)  At least in America)
18:52:08 <CrowKiller> and teeths get the rotor rotating
18:52:17 <kc5tja> A rotary engine doesn't have a crank, but an eccentric instead.
18:52:18 <CrowKiller> ha ok
18:52:25 <CrowKiller> http://web.ukonline.co.uk/Members/jr.marsh/wankel.html
18:52:49 <CrowKiller> i saw those picture, i recall a similar motor from a guy in quebec
18:52:56 <CrowKiller> using 4 "zones"
18:52:58 <CrowKiller> instead of 3
18:52:59 <kc5tja> The quasi-turbine.
18:53:40 <kc5tja> I don't see that engine going anywhere because it's more mechanically challenging to build.  Note: by "anywhere" I meant large commercial acceptance.
18:53:50 <kc5tja> It'll definitely find a niche market though.  It's a neat concept.
18:54:41 <CrowKiller> we had a good electrical motor divison at HydroQuebec R&D lab, but the prohect has been saboted and interrupted
18:54:50 <CrowKiller> the lectrical motor was in the wheel
18:54:59 <CrowKiller> and could recharge the battery while braking
18:55:03 <CrowKiller> something like that
18:55:12 <kc5tja> Ahh, not the first time that's been thought of.  And that's where it really belongs, because that's where you get the most torque!!
18:56:20 <CrowKiller> i cant argue on this i just finished physics this year using f=ma and other basic acceleration speed and distance formulas lol
18:57:39 <kc5tja> Well, if you extend Newton's law into the realm of angular mechanics, you get t = F*r, where r is the radius of the moment-arm (wheel in this case).  So the further out the electromagnets are, the more torque they produce.
18:58:22 <CrowKiller> yeah we studied simple machines too
18:58:31 <CrowKiller> with lever formulas and all
18:58:42 <CrowKiller> im doing my final exams this week
18:58:47 <CrowKiller> after that college!
18:58:49 <CrowKiller> =)
18:59:08 <CrowKiller> in quebec we need to go to college before going at the university
18:59:43 <CrowKiller> i had my chemistry exam today
18:59:45 <kc5tja> Here too, though we use different words.
18:59:53 <CrowKiller> kk!
19:00:03 <kc5tja> College and university pretty much have the same meaning.
19:00:16 <kc5tja> But "junior" college and "graduate" college are used to distinguish between the two.
19:00:47 <CrowKiller> and i hit my head for not learning more about oxydoreduction
19:00:56 <CrowKiller> this damn thing was driving me crazy lol
19:01:16 <kc5tja> Hehehe -- yet another term that's different in America.  We use "redox" instead.
19:01:26 <kc5tja> Redox is maddening from what I recalled in Chemistry.
19:01:36 <kc5tja> It's all accounting, and I make a horrible accountant.  :)
19:01:52 <CrowKiller> hahah i should have recalled that one, our redox table of reduction potentials is in englush
19:01:57 <CrowKiller> English
19:02:23 <CrowKiller> come on the guy was talking about negative anode and all i said whoaaa gimme a break lol
19:02:41 <CrowKiller> the anode is negative but it output positive current
19:03:11 <kc5tja> Well, they use conventional current for the anode/kathode charge designations.
19:03:29 <kc5tja> It was put into place long before we knew that electrons flowed from negative to positive charges.
19:03:40 <kc5tja> So things get confusing.
19:04:13 <kc5tja> The field of electronics, for example, uses *BOTH* forms of current flow; for analysis, conventional current is often used, but for circuit design and testing, electron flow current is used.
19:04:26 <kc5tja> Fortunately, you can convert from one to the other by reversing all the signs.
19:05:28 <kc5tja> Oh, that's bad.  The MDI air car site lists a 20:1 air compression ratio in their cylinders, but says that air at those pressures is 400C.
19:06:07 <kc5tja> Unfortunately, PV=nRT works only with absolute temperatures.  At 20:1 compression, the air temperature would be 6000K, or 5726C.  :)
19:06:54 <kc5tja> I somehow doubt they're getting 20:1 compression ratio, unless they're just giving the wrong air temperature.
19:21:01 --- quit: CrowKiller (Read error: 110 (Connection timed out))
19:22:50 <tathi> speaking of rotary engines, you know about that Moller guy who makes rotary engines and is trying to make personal VTOL aircraft?
19:22:55 <tathi> http://www.moller.com/
19:23:27 <kc5tja> Yeah, but I don't see his stuff going anywhere.  He's about five years overdue on his SkyCar concept, let alone full-scale production.  :)
19:24:15 <tathi> yah
19:44:48 --- quit: Herkamire ("leaving")
20:07:29 --- quit: tathi ("leaving")
20:35:12 --- join: skylan (sjh@Riverview18.tbaytel.net) joined #forth
21:45:33 --- quit: kc5tja ("THX QSO ES 73 DE KC5TJA/6 CL ES QRT AR SK")
22:08:21 --- join: Serg_penguin (~snaga_NOI@nat-ch1.nat.comex.ru) joined #forth
22:09:16 <Serg_penguin> hi ;(
22:09:24 <Serg_penguin> seems no one active...
22:28:44 <skylan> Indeed.
22:33:20 --- join: Soap` (~flop@202-0-42-22.cable.paradise.net.nz) joined #forth
22:42:52 --- quit: Serg_penguin (Read error: 104 (Connection reset by peer))
23:49:13 --- join: davidw (~davidw@adsl-32-74.38-151.net24.it) joined #forth
23:51:36 --- join: Serg_penguin (~snaga_NOI@nat-ch1.nat.comex.ru) joined #forth
23:51:40 <Serg_penguin> hi
23:59:59 --- log: ended forth/02.06.18