00:00:00 --- log: started retro/06.12.12
01:24:17 --- join: Cheery (n=Cheery@a81-197-54-146.elisa-laajakaista.fi) joined #retro
03:51:00 --- join: neceve (n=claudiu@unaffiliated/neceve) joined #retro
04:13:32 --- quit: nighty (Read error: 113 (No route to host))
04:17:09 --- join: nighty (n=nighty@66-163-28-100.ip.tor.radiant.net) joined #retro
05:15:07 --- quit: neceve (Remote closed the connection)
05:19:32 --- join: _timlarson (n=timlarso@65.116.199.19) joined #retro
05:28:40 --- nick: Raystm2 -> nanstm
06:16:12 --- join: Ray_work (n=Raystm2@199.227.227.26) joined #retro
06:19:17 <Ray_work> Good Morning.
06:29:47 --- quit: _timlarson (Read error: 104 (Connection reset by peer))
06:29:50 --- join: __timlarson (n=timlarso@65.116.199.19) joined #retro
06:33:34 --- join: Raystm2 (n=NanRay@adsl-69-149-37-190.dsl.rcsntx.swbell.net) joined #retro
06:47:47 --- quit: nanstm (Read error: 110 (Connection timed out))
06:48:50 --- nick: Raystm2 -> nanstm
07:20:07 --- quit: nanstm ("Should have paid the bill.")
07:21:10 --- join: Raystm2 (n=NanRay@adsl-69-149-37-190.dsl.rcsntx.swbell.net) joined #retro
07:22:04 --- quit: Raystm2 (Remote closed the connection)
07:30:55 --- join: Raystm2 (n=NanRay@adsl-69-149-37-190.dsl.rcsntx.swbell.net) joined #retro
07:31:35 --- quit: Raystm2 (Remote closed the connection)
07:33:51 --- join: Raystm2 (n=NanRay@adsl-69-149-37-190.dsl.rcsntx.swbell.net) joined #retro
07:33:53 --- join: Raystm2- (n=NanRay@adsl-69-149-37-190.dsl.rcsntx.swbell.net) joined #retro
07:45:28 --- quit: Raystm2 ("Should have paid the bill.")
07:55:42 --- join: tathi (n=josh@pdpc/supporter/bronze/tathi) joined #retro
08:03:30 <thinfu> howdy
08:05:22 <thinfu> hackers of the world unite!
08:12:13 --- join: virl (n=virl@62.178.85.149) joined #retro
08:14:53 <tathi> hi
08:15:04 <thinfu> how's it going tathi 
08:15:10 <thinfu> how long were you on the farm
08:15:21 <tathi> it's going well
08:15:26 <tathi> 3-1/2 months?
08:15:50 <tathi> yeah.  August 11th through the end of November.
08:16:17 <thinfu> so you work on the farm?
08:16:20 <thinfu> milk the cows?
08:16:39 <tathi> er...August 14th, but whatever
08:16:49 <tathi> they don't have gows.
08:16:54 <tathi> but they do have dairy goats
08:16:59 <tathi> and yeah, I did milk them a few times
08:17:25 <thinfu> woo!
08:17:26 <tathi> they tried to mostly have only a couple of people that did the milking though; less stressful for the goats
08:17:30 <thinfu> ah
08:17:40 <thinfu> why were you at this farm?
08:17:42 <thinfu> this a family farm?
08:17:46 <tathi> mostly I worked in the garden and with the chickens
08:17:54 <tathi> I was broke and looking for something to do
08:17:58 <thinfu> ah
08:18:12 <tathi> I was starting to think that I might want to have my own farm, so it seemed like a good way to try it out
08:18:14 <thinfu> you seem to be avoiding tech jobs, is that right?
08:18:18 <thinfu> ah
08:18:18 <thinfu> cool
08:18:22 <tathi> yeah...
08:18:26 <thinfu> do you think you want your own farm after all?
08:18:45 <thinfu> i definitely have an interest in hydroponics & having some livestock
08:18:50 <tathi> yes, definitely.  I didn't like the couple that ran the place much, they were very disorganized.  But the work was great.
08:18:56 <thinfu> ah
08:19:34 <tathi> it's not really a family farm; a guy in NY city with lots of money decided he wanted to finance a farm
08:19:42 <tathi> bring more fresh produce into the city, that sort of thing
08:19:50 <thinfu> i found out about this really neat way to do hydroponics which isn't sterile, you allow bacteria to grow in the water but because you keep the water moving fast & oxygenated (a little waterfall does the trick), only the good bacteria grows
08:19:59 <tathi> huh
08:20:03 <tathi> interesting
08:20:20 <thinfu> and so i have this idea of having like a outdoor swimming pool & indoor hydroponics all using the same water
08:20:25 <thinfu> and drinking from that water too ;)
08:20:33 <tathi> :)
08:20:47 <thinfu> more like an outdoor swimming river i mean
08:20:49 <thinfu> not a pool
08:20:52 <tathi> right
08:22:03 <thinfu> so do you think you want to do a farm?
08:24:04 <tathi> yeah
08:24:32 <tathi> not full-time probably, I'd really like to find a way to make some money doing interesting tech stuff one of these years
08:24:43 <thinfu> yeah
08:24:48 <tathi> but I'd definitely like to grow most of my own food and sell some to other people
08:24:55 <thinfu> well you could always become a ruby on rails programmer or something ;)
08:25:02 <tathi> hmmm
08:25:05 <tathi> yeah, ruby isn't bad
08:25:07 <thinfu> i'm working as a ruby on rails programmer, i'm very lucky ;)
08:25:30 <tathi> I can't stand Java and C++ though :)
08:25:35 <thinfu> yeah me either
08:25:47 <thinfu> well i hate all C-family languages anyways
08:25:59 <thinfu> altho i dunno about python 
08:26:03 <thinfu> haven't touched it yet
08:26:16 <tathi> Python isn't a bad language
08:26:23 <tathi> but it's too slow on my box
08:26:31 <thinfu> eh? how slow is your box?
08:26:42 <tathi> I don't mind C, actually.  It's low-level enough that I can deal with it.
08:26:49 <tathi> 533 MHz G4
08:27:11 <thinfu> thats faster than my 600mhz g3 ;P
08:27:23 <tathi> yeah, and I got two of those processors :P
08:27:27 <thinfu> ah heh
08:27:47 <thinfu> why are you saying python is slow on your box, i don't get it
08:27:49 <tathi> I'm also kind of picky about not waiting for programs to run
08:27:55 <thinfu> ah
08:27:59 <tathi> it has about a 0.2 second startup time
08:28:12 <thinfu> well it takes like 5 seconds for firefox to startup on my 600mhz g3 ibook ..
08:28:25 <tathi> yeah, firefox is a pig
08:28:27 <thinfu> i guess you're running linux then?
08:28:31 <tathi> yes
08:28:34 <thinfu> i'm using max os x 10.4 here
08:28:52 <tathi> and Python is substantially slower than Perl, like 5-10x
08:28:56 <thinfu> mainly i got the ibook to see what the fuss was about with textmate, quicksilver, etc
08:29:06 <thinfu> ruby is slower than python
08:29:06 <tathi> 0.2 seconds isn't actually that bad, but for a command-line tool...
08:29:32 <thinfu> hmm
08:29:40 <thinfu> well i don't think ruby starts up slow or anything
08:29:50 <tathi> no, ruby is a bit better, I think.
08:29:59 <thinfu> but for full fledged programs its gonna be slower than python i think..
08:31:06 <tathi> oh well, as long as you're not doing serious data crunching with it.
08:31:31 <tathi> it just drives me nuts that people write version-control tools in Python.
08:31:35 <thinfu> heh
08:31:41 <tathi> and Gentoo's package manager is in Python also
08:31:46 <tathi> it's painfully slow
08:31:46 <thinfu> yeah
08:31:49 <thinfu> hmm
08:31:54 <thinfu> yeah it is slow
08:32:05 <thinfu> so slow that someone had to make eix
08:32:21 <tathi> but of course most people just say things like "what are you doing with hardware that old?  Buy a new computer already!" :)
08:32:24 <tathi> eix?
08:32:51 <tathi> ah, I see
08:32:52 <thinfu> eix is like emerge -s
08:33:06 <tathi> yeah, I just wrote a quick bash script for that
08:33:13 <thinfu> heh
08:34:44 <thinfu> what text editor do you use?
08:35:06 <tathi> vim
08:35:22 <tathi> funny you should ask; I'm in the process of writing a clone in forth. :)
08:35:42 <thinfu> hmm
08:35:44 <thinfu> i use vim also
08:36:13 <thinfu> been using it for 4 years i guess
08:36:38 <tathi> yeah.
08:36:51 <tathi> I used emacs for a while because I had this perception that vim was hard to use/learn.
08:36:51 <thinfu> mainly i'm like "well people seem to like vim over emacs, so i'll be cool and use vim"  and now i just use it out of habit
08:37:10 <tathi> But once I actually tried vim, I decided I liked it better.
08:37:19 <tathi> dunno why
08:37:20 <thinfu> hmm, i think vim sucks
08:37:30 <tathi> well, it's annoying that it's modal
08:37:30 <thinfu> i use it and i think people that actually like vim are insane :P
08:37:31 <thinfu> heh
08:37:32 <thinfu> yeah
08:37:34 <thinfu> exactly
08:37:42 <tathi> but at least it makes it cheap to recover from mode errors
08:37:43 <thinfu> i still make modal errors every day
08:37:53 <thinfu> pff, i didn't even know about :u until last week
08:37:54 <thinfu> heh
08:38:03 <thinfu> my way of recovering was to quit without writing
08:38:07 <thinfu> and then restart
08:38:15 <tathi> oh man
08:38:26 <tathi> I don't think I'd even consider using an editor that didn't have undo.
08:38:57 <thinfu> i think everyone that uses vim has modal errors everyday, i think they just stop noticing it anymore
08:39:09 <tathi> oh, I still notice them
08:39:46 <thinfu> my idea for keeping vim almost exactly the same while removing the modal errors is to change around the order of the ESC key
08:39:55 <thinfu> instead of ESC:somecommand, do :somecommandESC
08:39:56 <tathi> the order?
08:40:13 <thinfu> where ESC reads back in the textfile to the : point and runs the command and removes it from the text
08:40:22 <tathi> hmm
08:40:32 <tathi> but some normal-mode commands don't have the :
08:40:38 <tathi> most of the ones that I use, in fact
08:40:39 <thinfu> like what?
08:40:58 <tathi> d x w e b /
08:41:01 <thinfu> ah like y d etc
08:41:01 <tathi> etc etc
08:41:09 <thinfu> w e ?
08:41:12 <thinfu> x ?
08:41:13 <thinfu> heh
08:41:18 <thinfu> teach me vim!
08:41:21 <tathi> w moves forward one word
08:41:37 <tathi> e does the same, but moves to the End of the current word instead of the beginning of the next
08:41:43 <tathi> x deletes one character
08:41:57 <tathi> b moves Back a word
08:42:25 <tathi> they all take a count, of course
08:42:25 <thinfu> hmm, as opposed to ctrl-arrowkey, end key, delete key, ctrl-arrowkey
08:43:01 <thinfu> i dunno i'm still used to the windows standard of editing heh, and that keeps it non-modal too
08:43:12 <Ray_work> sorta like the old WORD program on those shortcuts, no?
08:43:21 <Ray_work> hi btw. :)
08:43:28 <thinfu> there's cream, which is vim with the windows standard of editing
08:43:38 <tathi> yeah, I can't stand being forced to waste my time moving my hands over to the arrow keys and the PgUp/PgDown key block
08:43:49 <thinfu> hmm
08:43:54 <thinfu> but i don't have to look
08:43:55 <tathi> I'd love to see something non-modal, but the limitations of terminal emulators make that really hard
08:43:57 <thinfu> its all touch typing
08:44:11 <thinfu> moving my hand to the mouse is worse
08:44:18 <thinfu> and i haven't eliminated that
08:44:18 <tathi> oh sure
08:44:24 <tathi> I don't use the mouse much
08:44:32 <Ray_work> take an old Driveing game footpeddles and make a new kind of mode-er :)
08:44:40 <tathi> except in graphical web browsers, which I also try not to use much :)
08:45:12 <tathi> One of my long-term goals is to create a HumaneLinux distro.
08:45:25 <Ray_work> 0,0 
08:45:40 <Ray_work> crc just got his copy of THE, i see.
08:45:46 * Ray_work needs one now.
08:50:30 <thinfu> tathi: cool
08:50:39 --- join: neceve (n=claudiu@unaffiliated/neceve) joined #retro
08:51:19 <thinfu> tathi: that'd definitely be interesting to see
08:51:26 <thinfu> tathi: maybe you can do that ontop of gentoo some how
08:51:33 <thinfu> like gentoo is supposed to be a "meta" distro
08:52:09 <thinfu> so maybe you could add some scripts to the mix that assemble a humane collection of programs and stuff..
08:52:23 <thinfu> how would you make it humane though?
08:52:54 <tathi> it'd be a huge amount of work
08:53:15 <tathi> I figure you'd start by making a humane editor library
08:53:27 <thinfu> library?
08:53:32 <tathi> and then you'd have to hack all the apps you were including to use it
08:53:37 <tathi> kinda like readline?
08:53:48 <thinfu> hmm, i kept wondering what readline was
08:53:56 <thinfu> what is it?
08:54:19 <tathi> readline is a command-line editing/input/history/completion library
08:54:49 <thinfu> screw libraries, just make an interface thru which a person can use all programs ;)
08:54:54 <thinfu> its less efficient hto
08:55:03 <tathi> well yeah, that's what it would amount to
08:55:11 <thinfu> how about the zoomable thing?
08:55:30 <tathi> I'm not sure I like the zoomable thing
08:55:51 <thinfu> why not?
08:56:06 <thinfu> it doesn't seem like there's been any good implementations of it..
08:56:16 <tathi> it seems like it would be slow
08:56:23 <thinfu> hmm
08:56:35 <thinfu> i had this idea of doing it in CF
08:56:44 <tathi> not necessarily hardware issues, but just the time spent zooming out, zooming back to the new place, etc.
08:56:46 <thinfu> can't remember the exact idea but i had an easy way to do it i think heh
08:56:54 <thinfu> hmm
08:57:22 <tathi> I'd prefer something like hyperlinks, where you can just jump instantly to some position
08:57:25 <tathi> at least for most things
08:57:44 <thinfu> well the whole idea behind the zoomable thing was to take advantage of our ability to remember things spatially
08:57:56 <tathi> yeah
08:58:33 <tathi> I guess I'm just not someone who would use it, so I'm not enthusiastic about it.
08:58:36 <thinfu> in mac osx, you know of expose? 
08:58:43 <tathi> no
08:58:54 <thinfu> think of a little dash thingie over the last e
08:59:09 <thinfu> you hit like f8 and it shows all your programs
08:59:16 <thinfu> shrinks them all to fit the desktop
08:59:25 <thinfu> so you get an over view of what you're working on
08:59:32 <thinfu> or w/e
08:59:35 <thinfu> apparently people use it ;)
08:59:53 <tathi> yeah, the spatial layout thing makes sense
08:59:59 <thinfu> i think maybe you're against the zoomable thing because you don't have that feature..  kind of like how a person might be against laptops until they actually have one
09:00:01 <tathi> I just don't want to wait while I zoom into it.
09:00:07 <tathi> I want to select it and it to be there, instantly
09:00:21 <thinfu> select it with what?
09:00:31 <thinfu> i think the zooming could be quite fast
09:00:34 <tathi> however you select it
09:00:37 <thinfu> it doesn't even have to be real zooming
09:00:44 <tathi> I don't care how fast it is, if I can see it, it's too slow
09:01:01 <thinfu> it doesn't have to involve zooming at all i mean
09:01:07 <tathi> oh, ok.
09:01:18 <thinfu> more like levels
09:01:22 <tathi> yeah, if it's just an instantaneous change of position, I have no problem with it.
09:01:24 <thinfu> of view
09:01:55 <thinfu> well the zooming thing is for spatial memory but i'm sure the zooming could be turned off for those that don't need it
09:02:19 <thinfu> and you'd just basically go thru each level of view
09:02:29 <thinfu> the other thing is you can go through a piece of data
09:02:32 <thinfu> and find more
09:02:40 <thinfu> the zoomability is infinite
09:02:43 <thinfu> there's no end
09:02:44 <thinfu> heh
09:02:44 <tathi> sure
09:02:56 <tathi> at that point it started sounding way more like links to me
09:03:02 <thinfu> yup
09:03:20 <thinfu> i think zooming & links go hand in hand
09:03:25 <tathi> I agree.
09:04:38 <tathi> anyway, the big problem with creating a humane linux distro would be that most programs don't separate functionality from interface well.
09:04:58 <thinfu> hmm, i've never tried out Archy
09:04:59 <thinfu> have you?
09:05:07 <tathi> Only a tiny bit
09:05:15 <tathi> Jason said it was getting fairly good
09:05:45 <tathi> I believe that development stalled for a while when Jef died, but has since started up again.
09:06:03 <thinfu> well i actually don't think i agree with part of the editing feature that was discussed in the book..
09:06:14 <thinfu> where it changes which side the delete key functions on or w/e
09:08:03 <tathi> hmm
09:08:06 <tathi> don't remember that bit
09:08:24 <tathi> I thought it just deleted the current selection
09:08:35 <tathi> which was usually just the character under one part of the cursor
09:08:47 <tathi> I do recall some of the stuff sounding a bit weird though
09:10:41 <thinfu> can't remember, something about some combo of inserting a letter and the delete/backspace key  resulting in the cursor ending up on the wrong side or something,  so that it wasn't perfectly undoable..  i.e. the undo of a delete or backspace is inserting a letter
09:11:30 <tathi> ah.
09:11:39 <tathi> yeah, it seems like that sort of thing is really tricky to get right.
09:12:06 <tathi> one of the things that does bug me about vim is that doing 'i<Esc>' doesn't leave the cursor in the same place.
09:18:55 <thinfu> you know of GTD?
09:20:40 <Ray_work> colorForth seems to get that right, but only in some versions. There are versions where you cannot exceed a certain amount of cut text.
09:20:58 <Ray_work> Or paste more then the buffer holds, as well.
09:21:27 <tathi> google turns up a book titled _Getting Things Done_
09:21:31 <tathi> haven't read it
09:21:33 <thinfu> Ray_work: we weren't talking about anything remotely related to that..
09:21:59 <thinfu> tathi: yeah, its pretty good, check out the wikipedia entry
09:22:53 <Ray_work> you were not talking about inserting and deleting characters?
09:23:31 <thinfu> Ray_work: 'characters' being the key word :P  not cut'n'pasting ;P
09:23:44 <Ray_work> I see, were the delete insert happens on which side of the cursor. 
09:24:20 <Ray_work> in colorforth it amounts to the same thing as cut/delete paste/insert.
09:24:39 <thinfu> by insert i mean retyping the character
09:24:48 <Ray_work> I see.
09:25:35 <Ray_work> apparently Ray_work is trying to get into conversations where he has no business. :)
09:25:48 <Ray_work> Can you say bored. 
09:26:43 <thinfu> O say, can you see, by the dawn's early light..  Ray_work's boredome
09:26:52 <thinfu> dome! dom!
09:29:04 <Ray_work> in the news today: A baby got it's toes chewed off by a puppy as the PARENTS SLEPT NEAR BY! 
09:30:38 <thinfu> eww
09:34:52 <thinfu> ah, it was a pitbull
09:34:55 <thinfu> those suck ;)
09:36:10 <Ray_work> Apparently they suck so much that your toes come off.
09:37:35 --- join: absentia (n=scott@ns6.lobodirect.com) joined #retro
09:40:17 <absentia> We recognize that many countries around the world do not recognize a direct dedication to the public domain in their copyright laws. For this reason, we also grant unconditional, non-revocable permission to anyone who obtains a copy of this software to use, modify, and to redistribute it without limitation.
09:41:06 <Ray_work> absentia: what's that from?
09:41:08 <tathi> where's that from?
09:41:27 <absentia> that is so amazing to see in software.  the authors/developers/contributors should be lauded and commended.
09:41:44 <Ray_work> we would if we knew who they were. :)
09:41:45 <tathi> oh, from retroforth's wiki page on public domain
09:41:54 <Ray_work> AH gotta love crc! :)
09:42:00 <absentia> retroforth wiki -- in my quest for small virtual machines... etc...    -->   http://retroforth.net/wiki/?id=PublicDomain
09:42:04 <Ray_work> well, not physically...
09:42:25 <thinfu> don't lie
09:42:44 <Ray_work> okay physically to! 
09:43:28 <Ray_work> We have someone on that job, T. Childers.
09:43:30 <thinfu> absentia: did you just discover retroforth?
09:44:42 <absentia> yes
09:46:53 <absentia> :-< build script fails.. no "fasm" ...
09:48:11 <thinfu> heh, yeah gotta download fasm
09:48:30 <thinfu> the fasm package usually has a binary tho so you can use that if you're lazy about compiling like i am
09:49:00 <thinfu> true story: i downloaded retroforth & fasm onto my ibook before realizing my ibook isn't x86!
09:49:00 <absentia> looking at freeforth now.
09:49:14 <thinfu> why?
09:49:34 <thinfu> retroforth beats the pants of all forths out there
09:49:53 <absentia> seeing if/how it's different. it's also free (so I'm looking) ...  looks small (10K).  I just need to get my machine to boot up directly into a forth or lisp env :-)
09:49:56 <thinfu> crc is one of the few that code anywhere near Chuck Moore's level
09:50:10 <absentia> well, just looking.  I'm here to learn.
09:50:14 <thinfu> :)
09:50:21 <thinfu> yeah retroforth used to be around 4k-5k
09:50:32 <thinfu> now its around 11k i think
09:50:39 <thinfu> its pretty easy to strip it down though
09:51:03 <thinfu> and there is a native version for booting up into it directly
09:51:09 <absentia> from freeforth:  But I wasn't pleased with RetroForth registers allocation: popping the stack is easy with "lodsd" (EAX=*ESI++), but pushing it isn't as easy and too expensive in code space to be inlined. FreeForth registers allocation, detailed hereunder, saves lots of both code space and processor cycles.
09:51:26 <absentia> ya, I'll hunt down fasm.
09:51:33 <absentia> seems I need it for freeforth too.
09:51:39 <thinfu> hmm, is that a fork of retroforth? there's a number of forks
09:51:56 <absentia> not sure if it's a fork or a complete rewrite.. still reading.
09:52:03 <absentia> http://christophe.lavarenne.free.fr/ff/
09:54:30 <thinfu> absentia: have you tried out colorforth?
09:54:38 <thinfu> colorforth.com
09:55:22 <absentia> no, not yet.
09:55:36 <thinfu> its the "latest" forth that chuck moore (creator of forth) has developed, quite a bit different from classical forths
09:55:56 <thinfu> you boot up into it off a floppy
09:56:17 <thinfu> altho i think there's a usb version developed by others.. 
09:56:26 <thinfu> Ray_work: hows the usb bootable CF coming along?
09:56:49 <absentia> ya, that's 'bout what I want... boot up off floppy (or any other media) ...
09:58:01 <thinfu> absentia: there's lots of forths you should be able to boot off of :)..   forth is very nice, chuck moore even developed forth chips, stack-based & with 27 primitives
09:58:56 <absentia> yup. looking at that now.
09:59:46 <absentia> i'm trying to decide forth vs lisp in a boot.  lisp has 7 primitives ...  so just investigating.
10:02:48 <absentia> forth seems so similar to lisp ... based on stacks... using 32bit cells... etc.
10:03:09 <thinfu> hmm, i believe its been determined that forth can be as small as like 4 primitives or w/e
10:03:57 <absentia> nice.
10:04:13 <absentia> of course, if you are making new hardware :-)
10:04:28 <thinfu> well you can probably do both forth and lisp in like 2 primitives i think
10:05:23 <absentia> for colorforth, I see this:  Except for a small kernal, source is all there is. Object code is recompiled as needed, which takes no discernable time. This saves untold trouble in maintaining and linking object libraries.
10:05:39 <absentia> what is the kernel? pure x86 asm?
10:06:02 <thinfu> yup
10:06:42 <absentia> I don't know x86 asm.  might be time to earn :->
10:07:58 <absentia> do you think forth is just a type of macro assembly?
10:09:37 --- quit: neceve (Remote closed the connection)
10:11:31 <absentia> it's kind of difficult to find colorforth source
10:12:51 <tathi> http://qualdan.com/colorforth/chuck05-jg5.tar.gz
10:13:07 <absentia> thanks
10:13:11 <tathi> that's nasm source
10:13:21 <thinfu> nasm?
10:13:24 <thinfu> so it was converted
10:13:31 <tathi> just think how many different assemblers you'll have installed by the time you're done experimenting with forth! ;)
10:13:40 <tathi> thinfu: no, I disassembled that straight from the binary
10:13:49 <thinfu> absentia: absolutely, i think of forth in many ways as a macro assembler, as a vm, etc ;)
10:13:57 <absentia> suse101-c [318]> file color16
10:13:57 <absentia> color16: x86 boot sector, code offset 0x4e, OEM-ID "cmcf 1.0", root entries 224, sectors 2880 (volumes <=32 MB) , sectors/FAT 9, sectors 2880 (volumes > 32 MB) , dos < 4.0 BootSector (0x0)
10:14:09 <absentia> hmm. better not run that as root on this machine.  :-)
10:14:12 <thinfu> absentia: it definitely is similar to lisp, i find that i just have to think about things in the reverse order for lisp
10:14:12 <tathi> November of last year, Jeff released a new version, but it was binary only
10:14:27 <thinfu> tathi: ah, whats new in the new version?
10:14:50 <tathi> absentia: yeah, you're supposed to copy it onto a floppy and boot it
10:15:00 <absentia> I'm learning scheme now to implement other programming languages, and it is going where I thought it would go -- very complicated.... frames, etc.
10:15:12 <thinfu> some friends complained to me that lisp forced them to think recursively, but i never thought that, just have to do things in reverse compared to forth..
10:15:12 <absentia> forth looks like it cuts through the... much closer to the chip
10:15:25 <thinfu> yeah
10:15:27 <absentia> well, forth is RPN :-)
10:15:43 <thinfu> well the rpn thing is a natural consequence of exposing the stack
10:15:48 <absentia> yup
10:15:48 <tathi> thinfu: I don't know that there's very much new stuff
10:15:51 <thinfu> there's no other way to manipulate the stack ;)
10:15:56 <tathi> it has a 'find' feature
10:16:00 <absentia> actually, it's 100% inline with my AI ideas.
10:16:13 <thinfu> i love my hp48g rpn calculator..  much faster to work with than a ti-83 etc
10:16:23 <thinfu> i don't have to "store" numbers
10:16:28 <thinfu> i just leave it on the stack and come back to it later
10:16:29 <absentia> you basically don't know what you're going to do until you get there (ie: to the word) .. the word is a def of something to do -- and what's the environment? wel, duh, it's on the stack
10:17:05 <absentia> if you have to find through a stack -- your code is probably either severly broken -- or you haven't solved the problem the most efficient way.
10:17:08 <absentia> (or both)
10:17:13 <thinfu> i'd definitely be interested in hearing about your AI ideas
10:17:20 <thinfu> i've got a friend that is into AI too
10:17:26 <thinfu> i'm more into evolutionary computation myself
10:17:28 <thinfu> i think thats the key ;)
10:17:34 <thinfu> evolve solutions to everything
10:17:41 <lukeparrish> they're about the same thing imho
10:17:44 <thinfu> evolve applications, etc
10:18:02 <absentia> I don't want to be off topic here.. but the idea is exciting to me.. I look at chuck's ide driver.. that's so close to what I want to do -- I just want to hook up other "senses" ... and have these senses lead to "defining" words...  in a tight real time loop
10:18:12 <absentia> I don't need windows, linux, any of that to get in my way.  I want raw hardware .
10:18:40 <thinfu> you might need distributed computing though
10:18:51 <absentia> grab a few od the next generation lego mindstorms (there is forth for them, etc)...
10:18:54 <thinfu> to get enough computational power to get the AI happening..
10:19:08 <thinfu> are you aiming for like a singularity-type AI?
10:19:12 <absentia> no
10:19:35 <thinfu> ok
10:19:35 <absentia> my idea is multiple pcs ... executing in parallel.. .so you can either time slice, or if youhave chips .. or go distributed.
10:19:51 <absentia> the idea is that you "teach" the system by example.  to evolve a solution to a problem.
10:20:07 <absentia> the key is, of course,having the system get to the point where it can use enough of its words to evolve its own solutions.
10:20:31 <absentia> the other thing is, I don't think this is cheating, as the system won't know the different between asking a human or another "system" ... it just asks (ie: gets trained).
10:22:47 <absentia> I personally feel that cpus are well past the speed we need to get more "intelligence" out of systems... we're just being crushed under the latest OS fad.
10:23:07 <absentia> to me, an OS is just what is needed to interact with I/O .. anything else is an application -- and for the most part, those just get in my way.
10:23:45 <absentia> that's why booting off a floppy into scheme/lisp/forth is inviting to me.  I just need to be able to save my work, access hardware... 
10:24:08 <thinfu> yeah
10:24:48 <absentia> one issue I have with going too low... is that I want this device to speak to the internet -- and I don't want to have to code up ppp, etc...  :-<
10:24:51 <thinfu> i'd be pretty content with an editor & internet access..  the rest i can bootstrap ;)
10:25:00 <thinfu> right
10:25:00 <absentia> yup
10:25:50 <absentia> so, I can see lisp->asm and now I see freeforth has forth->asm ... I don't knowx86 asm, but if there are only so many primitives, I can learn those and then just convert those to asm for,say, ARM (my zaurus farm) ... 
10:26:21 <absentia> so, in that sense, the forth is 1 step up from asm (ie: macro asm) -- so I like this.
10:26:43 <tathi> yeah, maybe you can
10:27:05 <tathi> some of these minimalist forths are pretty closely tied to the hardware
10:27:15 <absentia> if I could get colorforth to boot from floppy and have code that I get from somewhere that allows me to talk to a modem or wifi card... I'm pretty much let loose :-)
10:28:06 <tathi> I've started to do a retroforth PPC port a couple of times, and given up on it as being more of a nuisance than it's worth.
10:28:26 <absentia> too bad there isn't source to the kernel.
10:28:39 <tathi> source to the kernel of what?
10:28:46 <absentia> colorforth
10:28:56 <absentia> although I do see boot.asm ...  
10:28:58 <thinfu> i think it'd be nice if there was a minimal forth coded for each architecutre, something like 1k or so
10:29:00 <tathi> er...that's what's in that tarball i posted the link to
10:29:12 <tathi> thinfu: yeah! that'd be awesome.
10:29:13 <thinfu> and then run a bigger forth implementation ontop of the minimal forth
10:29:23 <thinfu> something portable across all the minimal forths
10:29:34 <absentia> ya, 1K.... you basically set up the fastest, lowest level other than pure asm....  
10:30:10 <lukeparrish> we've also been giving some thought to a multi-user forth environment where people could interact with AIs or create virtual worlds.
10:30:12 <thinfu> i mean it'd be easy to code 1k easily for each architecture.. setup like 30 primitives or whatever, just enough to bootstrap a bigger forth ..
10:30:18 <tathi> absentia: boot.asm is the boot sector and the floppy-controller code
10:30:23 <tathi> gen.asm is the graphics stuff
10:30:30 <tathi> and kernel.asm is the actual forth core
10:30:57 <thinfu> it kind of annoys me how much of CF is in asm heh
10:31:12 <thinfu> i'm a 1 language guy, i prefer not to know any asm
10:31:28 <tathi> thinfu: actually, I think that's sort of what eforth was trying to be
10:31:32 <tathi> except it was pretty ugly
10:31:54 <thinfu> how small was it?
10:32:00 <tathi> dunno
10:32:02 <absentia> thinfu: exactly.
10:32:09 <tathi> but it was based off of around 30 primitives
10:32:15 <tathi> and then everything else got built on top
10:32:19 <tathi> so that was the only thing you had to port
10:32:25 <absentia> just trying to figure out how to use this color16, color32... kernel16 kernnel32 stuff.
10:32:53 <tathi> color16 and color32 are disk images, they get copied directly onto a floppy with dd or something
10:33:02 <absentia> I have vmware...
10:33:21 <tathi> kernel16 and kernel32 are intermediate build files
10:33:27 <tathi> hmm...vmware
10:33:28 * tathi googles
10:37:41 <tathi> I'm not finding anything useful
10:38:14 <absentia> er, www.vmware.com
10:38:33 <thinfu> absentia: nah, you might not be able to run CF in vmware
10:38:38 <thinfu> apparently CF doesn't run in qemu
10:38:39 <lukeparrish> there's also qemu
10:39:27 <lukeparrish> absentia: have you tried out the native version of retro?
10:39:33 <tathi> I meant I didn't find any information on how to actually USE vmware
10:39:56 <tathi> there's a message on the mailing list saying that the vesa version from Tim Neitz's site works under vmware
10:40:46 <absentia> heheh, vmqware won't let me make a virtual disk less than .1GB
10:40:59 <absentia> I tried to complie retroforth, but I didn't have fasm.
10:41:02 <absentia> still need to get that.
10:42:09 <absentia> honestly, someone needs to make a disk image or .iso of these :-)
10:43:08 <thinfu> absentia: http://flatassembler.net/download.php   its really easy, do it already ;P
10:43:47 <absentia> ok
10:43:49 <absentia> let me go there
10:45:36 <absentia> hmm. don't see any isos for colorforth there.
10:46:25 <thinfu> no, i mean fasm..
10:46:46 <thinfu> there are image files for colorforth
10:47:11 <absentia> ya, still trying to find fasm
10:47:44 <thinfu> the link i just gave is for fasm
10:47:53 <thinfu> anyways, some colorforth images can be found here: http://www.dnd.utwente.nl/~tim/colorforth/
10:48:12 <thinfu> fasm = http://flatassembler.net/download.php
10:49:36 <absentia> ok, now I get it.. ok, going to download
10:57:13 --- join: snoopy_1711 (i=snoopy_1@dslb-084-058-151-129.pools.arcor-ip.net) joined #retro
11:05:28 --- quit: Snoopy42 (Read error: 145 (Connection timed out))
11:05:31 --- nick: snoopy_1711 -> Snoopy42
11:08:19 <absentia> hmmm
11:14:16 <tathi> hmm
11:14:34 <tathi> thinfu: I think it would be very difficult to make even a minimal forth interpreter for PPC in 1K
11:15:54 <Quartus> I'd expect the abstract machine to take somewhere between 500 bytes and 1K, but the interpreter would be on top of that.
11:16:03 <tathi> right
11:16:46 <Quartus> I think that eliminating basic functionality in the name of squeezing the memory utilization is almost always pointless stunt nowadays.
11:17:21 <tathi> sure
11:25:02 <absentia> re
11:25:30 <absentia> ok, manaed to find a colorforth for windows .zip ... ran it...  held downthe "a" key.. crashed the program :->
11:26:12 <absentia> now I'm looking for colorforth for linux
11:26:43 <virl> absentia, not better here..
11:27:20 <absentia> not better?
11:27:34 <virl> yep, also extremly buggy.
11:27:49 <absentia> what? colorforth for linux? colorforth for windows?
11:28:09 <virl> for linux
11:28:17 <absentia> ah
11:28:29 <thinfu> tathi: when i said 1k i meant that as a benchmark for ease of coding
11:28:36 <absentia> :-<  maybe I could get bochs.
11:28:59 <thinfu> tathi: i.e. develop a minimal forth for any given architecture in a day..  
11:29:33 <thinfu> with standard primitives, so that a larger forth could be bootstrapped..
11:30:02 <virl> good colorforth = clean docs + custom building system + well documented source -> a dream of a colorforth
11:30:11 <thinfu> well i mean the larger forth would be the one with the interpreter 
11:33:29 <absentia> I was in windows... fouond a page of colorforth with wifi drivers,ethernet, dhcp... looked amazing.
11:34:08 <absentia> but I don't use windows... I'm in vmware with linux...   I'll start building up my colorforth info... looks like this is going to be better than lisp.
11:37:01 <absentia> anyone know of any quines in forth?
11:38:37 <absentia> hmmm. very little (ie: none?) information about running colorforth under linux.
11:41:20 <tathi> I don't think colorforth does run under linux
11:41:31 <absentia> well, you could run bochs, etc.
11:41:41 <tathi> oh, ok
11:41:48 <thinfu> or vmware
11:42:00 <thinfu> dude, if it doesn't work in vmware, like hell its gonna work in bochs or qemu ;)
11:42:49 <tathi> http://jasonwoof.org/colorforth_under_bochs
11:42:54 <absentia> well, bochs is a purce emulator.  it will work in bochs
11:43:00 <absentia> it's just a matter of getting a freebios, etc.
11:43:57 <absentia> what's the largest/best thing you've coded in forth?
11:45:14 <thinfu> me? irc client & multi-user echo server (had no error checking though so it couldn't handle disconnects)
11:45:21 <thinfu> did it in retroforth using syscalls
11:46:27 <thinfu> thats an excellent question tho
11:46:40 <thinfu> a lot of forthers haven't actually coded much in forth
11:46:51 <thinfu> they spend more time coding forth implementations :/
11:47:40 <absentia> well, I definitely want to get some networking code going.
11:48:01 <absentia> is there a standard for markup colorforth when you don't have colors?
11:48:08 <absentia> (like chinese or arabic, etc)
11:48:25 <absentia> \r macro \w ---
11:48:38 <absentia> seems you can't cut'n'paste colorforth code :-)
11:49:01 <absentia> it also might be a good thig to get a colorforth going with the latest box and get it into the image farm -- same with vmware.
11:49:19 <absentia> thi: would you be willing to share your irc client code?
11:50:06 <thinfu> absentia: http://retroforth.org/projects/darcs/RetroIRC/
11:50:18 <thinfu> absentia: crc hacked it up a lot though 
11:51:09 <thinfu> http://thin.bespin.org/forthcode/chatserv.rf
11:51:25 <absentia> hacked it up good ... or hacked it up bad?
11:51:36 <absentia> retroforth or colorfoth?
11:51:37 <thinfu> goodish ;)
11:51:41 <thinfu> this is retroforth
11:51:46 <thinfu> i haven't done much colorforth coding
11:51:55 <absentia> ok
11:52:10 <thinfu> its out of date too
11:52:12 <absentia> I'll get retro going...  looks like it has sockets and I can cut'n'paste code, and I can run it under linux
11:52:14 <absentia> win win win
11:52:23 <thinfu> i bet it won't work on current version of retro
11:52:33 <thinfu> without a few modifications..
11:52:50 <absentia> see, in scheme... it doesn't have sockets.
11:52:58 <absentia> the first program I write in any language (that I learn) is an irc client.
11:53:09 <thinfu> well i did it all thru syscalls..
11:53:24 <thinfu> but apparently retro can interface iwth the C libraries nicely now
11:53:37 <thinfu> thru the 'ffi' stuff
11:53:43 <thinfu> foreing something interface
11:53:47 <thinfu> foreign*
11:55:24 <tathi> function
11:55:52 <thinfu> crc said something about redoing the irc client using ffi instead of syscalls
11:57:35 <tathi> that would make sense
11:57:54 <tathi> coding the name lookup stuff yourself is a nuisance
11:58:27 <absentia> ya, uffi etc.
11:58:29 <thinfu> hmm, i didn't use name lookup
11:58:47 <tathi> no, I assume currently you have to give an IP address
11:59:58 <thinfu> i think a maybe syscall library would be nice..  provide a more sane & self-documenting interface..
12:00:10 <thinfu> i kinda felt bad about not documenting my syscall usage
12:00:39 <absentia> ya
12:00:46 <absentia> I have no idea ....  :-<
12:00:54 <thinfu> but i didn't want 100 lines of comments just to explain the syscalls
12:01:06 <thinfu> heh
12:01:25 <thinfu> i believe in self-commenting code ;)
12:01:35 <absentia> since you guys are closer.. what would be the big diffs between retro and color ...   ?
12:01:42 * absentia agrees.
12:02:08 <thinfu> retro is more classical forth with some ideas borrowed from color, color is cutting edge, chuck moore made radical changes
12:02:39 <absentia> here an example of some of my self documenting perl code:  ("@_" =~ m@(?:^|\n)\s*\b$pat\b\s+(?:[^\n]*?)\s*([^\n\s]+?)\s*(?:#[^\n]*?)?\s*(?=\n)@ig);
12:02:47 <thinfu> like for example, instead of doing  : foo <code> ; its <red>foo</red> <green>code</green>  and ; can be used multiple times..
12:02:50 <thinfu> lol
12:02:57 <tathi> I haven't looked recently, but they used to be pretty similar, except for the use of color.
12:03:17 <absentia> well, I don't care for color = syntax, that's like python spaces are blocks....   
12:03:39 <absentia> ok, ya, my question (to myself) was if code from one could be used to the other
12:03:47 <tathi> though retroforth went back to a single dictionary with types instead of separate code/macro dictionary
12:03:55 <absentia> the issue I see with color -- is that there needs to be a standard to mark it up for non-color displays.
12:03:56 <thinfu> absentia: code can be ported somewhat easily
12:04:12 <absentia> great.  I figured as much.
12:04:24 <tathi> yeah, I agree
12:04:26 <thinfu> there is a standard markup iirc, i forget what it is though..
12:04:33 <absentia> ah, ok
12:04:41 <tathi> color translates out to punctuation pretty simply
12:04:42 <absentia> can you put comments anywher ein lisp code?
12:04:43 <thinfu> well not standard so much as what people have sort of agreed on in the forums..
12:04:55 <thinfu> in lisp code?
12:05:08 <absentia> (r) macro (w) ( blah blah blah ) (g) dup
12:05:12 <absentia> no, forth
12:05:23 <thinfu> yeah ( ) can go anywhere
12:05:29 <absentia> kind of funny, almost turns forth into "inside out" lisp.
12:05:38 <absentia> ok, so I bet the markup uses comments. :->
12:05:50 <tathi> just tack a ': ' before red words
12:05:58 <tathi> '[ ' is green-yellow transition
12:06:08 <absentia> yup.  I was thinking (:r) blah (:w) (blah blah)
12:06:11 <tathi> '] literal' (?) is yellow-green transition
12:06:29 <tathi> white should be surrounded by parentheses
12:06:35 <tathi> I think that's about it
12:06:51 <absentia> ok, I saw that in the code.. : = define word, etc.    ok.  
12:07:12 <tathi> if you understand both languages, you'll have no trouble translating at all
12:07:14 <absentia> guess I need to get retroforth going... and then I can take my time trying to get colorforth going.  
12:07:19 <tathi> s/languages/dialects/
12:07:32 <absentia> well, they're not that different... other than the color in the display.
12:07:37 <absentia> ya, like lisp/scheme.
12:07:50 <tathi> but they're both fairly different from standard Forth.
12:08:17 <absentia> ah
12:08:51 <absentia> have you seen retroirc?
12:08:58 <thinfu> and by standard forth tathi means fig83 and fig79 forth..  ANS forth is is bastardized and not a real forth ;)
12:09:18 <absentia> :-)
12:09:22 * absentia knew that already.
12:09:22 <tathi> heh. actually i did mean ANS forth
12:09:32 <thinfu> aww :P
12:09:32 <tathi> there's nothing wrong with it.
12:09:36 <thinfu> "
12:09:58 <absentia> I would be interested in seeing any type of gui effort in rf -- or mandlebro (with gfx) in rf
12:10:25 <tathi> crc did a little bit with GTK+ back when he first got FFI going, I think...
12:11:42 <tathi> absentia: I messed with retroirc a little bit several months back -- someone was having trouble getting it to work under BSD.
12:12:15 <absentia> ya, I need to read more... the "vector" stuff seems strange to me.
12:12:21 <absentia> let me try to get rf going under linux now.
12:13:08 <absentia> are you familiar with lambda calculus (the blog lambda-the-ultimate?)
12:14:16 <tathi> I wasn't aware that the two had anything to do with each other
12:14:27 <thinfu> i'm aware of the latter
12:14:35 <thinfu> the former i'm sort of aware of
12:15:13 <thinfu> i don't really like the term 'lambda'..  ruby has Proc.new  ;)
12:15:17 <absentia> I am not sure what it would be called in forth... a nameless-word?  an anonymous-word?
12:15:33 <thinfu> yeah you can do nameless words in forth
12:16:09 <thinfu> :: <code> ;
12:16:41 <thinfu> in retroforth i mean
12:16:47 <absentia> from my naive pov, it seems those languages are hyping the fact that they can put routines onto the stack and then call/reference them.  to me, I guess you could (pollute?) load up the forth word dictionary with names.. .and use that, since in forth, the calls are supposed to be "cheap" ... on the call part -- what I don't know is hw the lookup is done (I assume that's some sort of hash or tree)
12:17:06 <absentia> ya, I saw that in retroirc -- along with loc: 
12:17:17 <absentia> if loc = scope -- then I'm all set.
12:17:23 <tathi> yeah
12:17:34 <tathi> all names inside loc: ;loc are discarded
12:17:46 <absentia> I use to write a lot of shell scripts.. (being a unix sysadmin for 17+ years) .
12:17:48 <tathi> but if you use :: it leaves the address on the stack, so you can give it a name later
12:17:58 <absentia> and I used $_ in my scripts ...  great -- until bash comes along and CLAIMs my var!
12:17:59 <absentia> :->
12:18:17 <absentia> ya, that's like a continuance.
12:21:41 <absentia> amazing how much better "./build linux" works for rf when fasm is installed (and in my path, etc)
12:22:17 <absentia> are you familiar with the great language shootout ?
12:22:32 <virl> in which context?
12:23:04 <absentia> http://shootout.alioth.debian.org/gp4/
12:23:37 <absentia> uses bigforth
12:23:46 <absentia> still amusing looking at it compared to c/c++
12:24:11 <absentia> gcc/g++ better in almost every way
12:25:11 <absentia> moore seemed to say that forth was 10x faster, or 100x or 1000x faster....  yet, in this shootout, the c is the fastest (by far)
12:26:34 <absentia> must be something wrong with either bigforth, or the implementation of the problemsolution -- bigforth even loses out to lisp!
12:29:09 <virl> but it compares nicely, when you look at the fact that forths compile before they can execute
12:29:16 <virl> (at runtime)
12:32:13 <virl> this one speaks a bit a different story, not every programming language is good in every area: http://shootout.alioth.debian.org/gp4/benchmark.php?test=all&lang=bigforth&lang2=sbcl
12:32:35 <absentia> true
12:32:37 <absentia> but I am surprised
12:33:02 <absentia> sure, better for memory use... but not cpu time
12:33:52 <virl> well, cpu time.. that could be bigforths implementation.
12:34:03 <absentia> ha, compare the bigfortyh to erlang
12:34:04 <absentia> that's sad.
12:34:15 <thinfu> absentia: a lot of the 10x etc speed is really about how forth helps the programmer to solve the problem in more elegant ways
12:34:25 <virl> and sbcl, iirc sbcl is damn fast implementation of lisp.
12:34:34 <absentia> yup
12:34:53 <absentia> I have it installed... like I said, I'm looking at lisp vs forth... 
12:34:57 <thinfu> like chuck moore did a sort of gui thing for his chip designer and it was really small, like 5k ? i can't remember
12:35:07 <thinfu> smaller than 5k maybe
12:35:10 <virl> ok, erlang vs. bigforth -> bigforth kicks it's ass
12:35:26 <absentia> but forth seems like it would be closer to the hardware... I've seem some virtual machines implemented in like < 1K c source.  I haven't found that yet for lisp/scheme -- although it should be possible.
12:36:08 <thinfu> bigforth is implemented in C iirc
12:36:13 <absentia> thinfu: I'm being selfish here... I'm sure I'll code some gfx and just make my own, but for the ai .. I want to get some of the latest lego mindstorms....
12:36:42 <thinfu> well one of the advantages C has over forth is gcc..
12:36:49 <thinfu> i.e. an optimizing compiler
12:37:16 <thinfu> altho there are a few forths with optimizing compilation, but nothing serious
12:39:07 <absentia> right, but I don't see c where you can boot up off a floppy and have an interpreted c (or c++) ... where yo can then talk to wifi, net, etc.
12:39:41 <absentia> part of what I like about c, is I think it is also a macro assembler... but for use under another os, etc.. where you have more interrupts, time slicing, guis, etc.... with a forth, I want all the cpu to myself.
12:39:47 <virl> well, howto make forth damn fast?
12:40:17 <absentia> so, in theory, you could just write asm and then forth/c/lisp doesn't matter -- as you know asm.. but I don't like asm as I've used 5 cpus in my short coding lifespan -- and I don't want to be locked into one implementation.
12:40:28 <thinfu> virl: not important
12:40:45 <lukeparrish> run it on a faster system! :P
12:40:58 <absentia> well, forth is supposed to be damn fast...   it's supposed to basucally be compiled and then as low level as possible.
12:41:03 <absentia> luke: muahaha
12:41:23 <absentia> I've got a 3GHz cpu at home.  that should be faster than I ever need to do just about anything that I could think up.
12:41:36 <absentia> instead, with windows ... it takes minutes ... or longer ... to load programs, etc.
12:41:56 <lukeparrish> yeah 3GHz should be bordering on sentience
12:42:45 <absentia> :-)
12:42:54 * lukeparrish uses IceWM on Slackware :)
12:43:29 <absentia> that's wqhy I was looking into implementing my own lisp -- just using 32bit cons cells.. but seems forth uses virtually the same, and it's "pure" ... meaning I have pute 32 bit words, and I don't have the overhead of the cons-cell pair(s)
12:44:53 <absentia> I like xfce... it's nice looking, not too large.. can run it on my pda, x windows, vnc consoles.
12:44:59 <absentia> here is my zaurus:
12:45:30 <absentia> http://zaurus.spy.org/screenshots/scottyelich/pdaXrom/jpg/main1.jpg
12:45:50 <lukeparrish> nice
12:45:51 <absentia> it would be great to boot up that thing into a low level ... and get the entire 400mhz cpu
12:46:15 <absentia> it's arm.  I cancode c++ on it, compile.. and have my same code as anywhere else -- windows, linux, there... with Qt api.. etc...   but the code is huge.
12:46:27 <absentia> moore says forth programs can be 1% the size of c programs.
12:47:10 <absentia> I can type 40wpm on that zaurus... it has 640x480, 16bit color(??) ... wifi, 128mb ram... 6GB local storage....
12:47:37 <absentia> one nice thing is that I can load up emacs... edit c++ code, compile c++ + Qt .. *and* listen to mp3s at the same time.
12:47:39 <tathi> absentia: yes, but only by leaving things out.  moore is more than slightly obsessive about minimalism.
12:48:00 <tathi> granted, there would seem to be plenty of things on present-day systems that could profitably be left out. :)
12:48:21 <absentia> well, I want to leave out the entire OS :-)
12:48:43 <thinfu> chuck moore is big on taking the problem, trivializing it until the programmatic solution is 1 line ;)
12:48:51 <absentia> ha
12:48:56 * absentia wants real-world solutions.
12:49:11 <thinfu> its plently real world
12:49:25 <thinfu> well basically most programmers have a habit of not solving a given problem directly
12:49:38 <thinfu> they end up solving some other problem because its more fun that way
12:49:42 <thinfu> and end up with 100x more crap ;P
12:49:51 <thinfu> and its harder to maintain, etc
12:49:59 <tathi> lol
12:50:02 <absentia> right
12:50:07 <absentia> this is one of the draws of lisp.
12:50:27 <absentia> making domain language specific to the solution at hand
12:50:33 <absentia> but, alas, that's what forth seems to do.
12:50:39 <thinfu> yeah thats exactly what forth does ;)
12:50:49 <thinfu> crc comes from a lisp background
12:51:28 <thinfu> i guess it all depends on how minimal you want to be and how much parantheses annoy you ;)
12:51:51 <thinfu> but yeah you can get a functional interactive environment happening much quicker with forth i think
12:51:55 <thinfu> with an editor etc
12:53:53 <absentia> ya
12:54:01 <absentia> I'm hitting' it hard now.... but it seems that way.
12:55:01 <absentia> sun's new chip will have 32 cores... and each core can run ... 4 threads?
12:55:16 <absentia> you want 128 "forths" running ... or at least 32.
12:55:57 <thinfu> chuck moore had a 25x forthchip designed like 5 years ago that had 25 cores
12:56:17 <thinfu> hasn't been manufactured though :/
12:58:40 <absentia> but I'm missing something when there are claims for 10x 100x or 1000x speeds ('
12:59:08 <absentia> I saw one rediculous page for chess claiming like 108,000,000,000 or something... forth instructions per second.  
12:59:28 <absentia> anyway, if c just goes straight to asm ... why not just learn a subset of c and code that? vs forth?
12:59:49 <thinfu> absentia: http://www.ultratechnology.com/1xforth.htm
13:00:02 <thinfu> that explains where the gains are coming from..
13:00:07 <absentia> ya, I read 1x forth, 25x forth.... lots of those pages.
13:00:13 <absentia> 3 pages of that.
13:00:19 <absentia> refactor refactor refactor.
13:00:21 <absentia> :-<
13:00:48 <thinfu> yeah well he's saying that if you code 10x more than the problem warrants, then bam, its 1000x slower, more complicated, etc
13:10:07 <thinfu> absentia: and thats how he got that ide driver ;)
13:10:19 <absentia> I like that ide driver
13:10:21 <thinfu> absentia: its got 0 cruft, 0 error checking, etc, it does only the bare minimum :P
13:10:27 <absentia> does it work? looks pretty brute-force
13:10:34 <absentia> right
13:10:59 <thinfu> probably works
13:11:36 <thinfu> he hasn't updated his colorforth site in a long time so the colorforth he provides probably works with the ide code..
13:20:43 <absentia> ya
13:21:08 <absentia> just seems like retro is the way to go, I can run it inside of linux, which means I can listen to mp3s while I run it :->  (if there is an arm version)
13:21:38 <absentia> i'm just still concerned 'bout the speed via the shootouot, with c vs forth....  if they're both going down to asm fairly straight -- what advantage does the forth have over the c...    
13:21:56 <absentia> also, since I haven't looked yet -- is there a (floppy/iso) boot to retro forth system ?
13:22:03 <thinfu> hmm, right now its just x86 i believe, but crc's been meaning to port it to arm, etc.. these days he's doing a minimal forth in C, i think maybe he's planning on using that to make retroforth more portable..
13:23:53 <absentia> ya, that's what I'm looking for -- a minimal language in c... or just a vm....  I found one but I can't seem to make it do much and since I don't really know C like an expert, I haven't been able to figure it out yet.. but it is 1K.
13:24:54 <absentia> but a minimal lisp is C was what Iw as looking for -- the closest I found was (1) lisp in lisp with 7 opcodes (graham ref the original) ... (2) lisp in C from 1985 [didn't compile, but I was able to fix that] and (3) lisp in awk...   each of these I could translate, but they're somewhat large...
13:25:29 <absentia> I've found like pico micro scheme -- but the C is still like 4000 lines and it appears to be highly optimized C (which means it's confusing as hell) -- and it still won't do sockets.
13:26:26 <absentia> I don't want to just use C(c++) and then use guile... seems too large.  so, botting into a scheme(-like) / forth language seems best.... what's nice is that is seems that the forth world has more interfaces to devices.
13:26:32 <absentia> I don't see that at all with scheme.
13:26:36 <thinfu> absentia: i think it might be the ./build generic that you dd write to a floppy
13:26:44 <absentia> yup
13:27:19 <thinfu> might be wrong
13:27:40 <absentia> and vesa is not bad... so with a few device drivers -- you're all set.  I'm a little concerned that the kernel is a vm for forth, putting a layer between the asm and the use -- but maybe 
13:27:51 <absentia> that is needed to give interactive support.
13:40:02 --- quit: __timlarson ("Leaving")
13:40:41 <absentia> although, I guess, one nice thing 'bout the kernel... although I haven't verified this.. is that you can't crash the system by bad language context.
13:40:44 <absentia> ie: syntax error
13:41:57 <thinfu> which kernel?
13:42:14 <thinfu> i've crashed colorforth i think
13:42:39 <absentia> ya, colorforth
13:42:48 <absentia> sure, I crashed it too... I just held down the A key.
13:43:02 <thinfu> LOL
13:43:15 <absentia> but if you're accepting input from a user and then eval/compiling it... you have to have a thread/process running.
13:44:03 <absentia> that was my experience with colorforth for windows.. downloaded the .exe self extracting, extracted, ran the .exe -- held down a... got a windows popup... forced the dos code to close... went back to linux 
13:44:18 <thinfu> colorforth didn't have multithreading, but a version of it with that has been hacked up
13:44:39 <absentia> I'm concerned about rf if it only has a single stack ... 
13:44:55 <virl> it has two stacks
13:44:59 <absentia> I am interested in multi-stack -- single/multi-thread ... although I"m sure I could simulate it.
13:45:07 <absentia> er, well, ya, data stack and return stack.
13:45:28 <absentia> I mean multiple data stacks -- multiple return stacks.  almost basically, a multi-process env.
13:45:42 <absentia> the idea for the AI is that a lot of code is just a continuance with a certain environment.
13:50:06 <absentia> n/c
13:54:00 --- quit: thinfu ("leaving")
14:27:32 --- quit: Ray_work ("User pushed the X - because it's Xtra, baby")
14:54:42 --- log: started retro/06.12.12
14:54:42 --- join: clog (n=nef@bespin.org) joined #retro
14:54:42 --- topic: 'RetroForth | Pastebin @ http://retroforth.net/paste | The wiki & paste bin editing key is 'despair''
14:54:42 --- topic: set by crc on [Thu Dec 07 14:14:56 2006]
14:54:42 --- names: list (clog Raystm2 Snoopy42 absentia virl tathi nighty Cheery timlarson crc Quartus_ @ChanServ Quartus lukeparrish Shain)
14:54:56 <Quartus> sure, optimization is typical for native code compilers.
14:55:22 <Cheery> I ended up having VM in my language implementation.
14:55:37 <Quartus> not native-code, then?
14:55:57 <Cheery> well, I'll do the native-code implementation for my VM bytecode. :)
14:56:10 <Cheery> On the VM itself!
14:56:50 <Cheery> I think I benefit from this approach since I can mount the language to be my coding environment soon as possible.
14:57:14 <Cheery> or what do you think? ;)
14:57:36 <Quartus> Depends on your goals.
14:57:52 <Cheery> I thought this is a lot better idea than one I had before of directly trying to compile into native.
14:58:22 <Cheery> my goal is a portable variation of forth environment.
14:58:36 <Quartus> Different from GForth in some way?
14:59:03 <Cheery> yes, syntax is totally different and this handles code in modules.
14:59:11 <Quartus> so not Forth, then?
14:59:30 <Cheery> no forth, more like a sidestep from forth.
14:59:59 <Cheery> ie, using the innovation in forth and introducing my own.
15:00:06 <Quartus> ah. Would be good if you could actually write your language *in* Forth, then it would run on any Standard system and you could take direct advantage of the existing compilers.
15:00:12 <Quartus> And you'd get portability.
15:01:00 <Cheery> well, when I get the things done and things working, this implementation will be so simple that it can completely host itself.
15:01:13 <Cheery> I do not need any middle-level -language then anymore.
15:01:49 <Cheery> or host language.
15:02:03 <Cheery> (which seems, btw. very much better and less insulting -feel)
15:02:05 <absentia> cheery: what is your goal, again... sorry... just got back to the screen
15:02:08 <Quartus> Insulting?
15:02:30 <Cheery> I don't know, totally wrong, bad, nonsense.
15:02:36 <Cheery> nevermind. :)
15:02:41 <Quartus> You may be able to be self-compiling on a given platform, but you'll have to re-write your VM for every new target; that is your 'middle-language', and you can't escape it.
15:03:08 <Quartus> If you use an existing VM, say, a nice native-compiling Standard Forth for the given target, you could code to that and be done with it, focus on what actually differentiates your language from others.
15:04:11 <Cheery> well. If I write the VM in the language it supports and I have native level compiler for the given platform, I can port it.
15:04:17 <Cheery> and all the software written on it.
15:04:27 <Quartus> I'm afraid you lost me there.
15:04:57 <Cheery> hmm.
15:05:09 <Cheery> I'm not sure whether I lost you.
15:05:32 <Quartus> There' s such thing as a VM that will magically re-write itself for a new target. 
15:06:00 <absentia> there is?
15:06:04 <absentia> or there is no such
15:06:05 <Quartus> sorr, no.
15:06:06 <Quartus> no such.
15:06:13 <Quartus> Missed a whole word. :)
15:06:21 <Cheery> I still need to write the compiler for the target platform, I thought there is no problem here. :)
15:06:39 <Cheery> because writing the compiler should be a quite straightforward.
15:06:42 <Quartus> Right, so you have your insulting 'middle-language' no matter what.
15:06:45 <Cheery> If I get this thing right at all.
15:07:07 <Cheery> oh now you use the term different way. :)
15:07:12 <Cheery> well...
15:07:43 <Quartus> As I say, it depends on your goals.  Forth makes an excellent breeding-ground for language design; it's a shame it seldom gets used that way.
15:07:46 <Cheery> I liked the choice to take VM into stronger relationship in my system instead of other choices I were found out.
15:09:48 <Cheery> VM also has so many other useful advantages over the last approach I was trying out that I decided to take this approach.
15:11:06 <Cheery> oh well, I go to sleep now. :)
15:12:27 <Cheery> I think it over night.
15:12:36 <Quartus> Ok, see you!
15:13:23 --- quit: Cheery ("Download Gaim: http://gaim.sourceforge.net/")
15:20:46 --- quit: virl ("Verlassend")
15:24:47 <absentia> scheme also claims that
15:25:18 <absentia> but then lisp and scheme seem very similar at the lowest levels, although I do think forth is more pure because it doesn't have the cons-cell overhead (for pair based datastructures)
15:25:30 <absentia> ah, blah. issed'm by --><-- that much.
15:28:48 <absentia> anyone awake ?
15:28:58 <Quartus> well, if you want pair-based data structures, you need the equivalent of a cons.
15:29:05 <absentia> I don't want it
15:29:32 <Quartus> I'm not referring to your specific desires, I mean that Forth doesn't have some magical way of avoiding that need.
15:30:02 <absentia> perhaps a silly question -- I just got into forth.... if I am doing my own type of data structure, etc... do I have to know where free memory is and handle that myself?  ie:  structure size * count? to implement my own type of "array" for indexng the objects? and then i'd have to do my own type of garbage collection... I assume
15:30:32 <Quartus> I'm not sure what the question is, exactly.  You can allocate and free memory just as you would in, say, C.
15:30:43 <absentia> really? ok.  I'll keep reading.
15:30:45 <Quartus> How you want to manage a specific data structure would depend on what your application requires.
15:31:32 <absentia> do you know both lisp and forth? (well?) ?
15:31:45 <Quartus> I use Forth more than Lisp.
15:31:55 <absentia> why?
15:32:02 <absentia> (if you don't mind me picking your brain ...)
15:32:10 <Quartus> I prefer it.
15:32:22 <absentia> sorry; why?
15:33:22 <Quartus> The abstract machine for Forth maps closely to the actual operation of the machine, and the language is a better fit to the programmer, at least in my case, than is Lisp.
15:33:32 <absentia> yes
15:33:37 <absentia> I am coming to that same conclusion.
15:33:45 <Quartus> Three-sided, really. Better fit to the machine, better fit to the programmer, better fit to the abstractions required by any specific application.
15:34:20 <absentia> how long have you used forth? what's the most interesting thing you've coded in it? any real-world programs ?
15:34:51 <Quartus> Quite a few years now.  Quartus.net is my site, Quartus Forth is a Forth system for the Palm.  I write quite a few things in Forth.
15:35:05 <Quartus> Some of the apps are on that site; others have been for specific employers, some just for me.
15:35:05 <absentia> ya, I have a zaurus ... several, actually.
15:35:11 <absentia> awesome.
15:43:19 --- quit: tathi ("leaving")
16:10:04 --- quit: Raystm2 ("Should have paid the bill.")
16:10:24 --- join: Raystm2 (n=NanRay@adsl-68-93-42-143.dsl.rcsntx.swbell.net) joined #retro
20:28:08 --- join: thinfu (n=wunderwa@bespin.org) joined #retro
20:30:38 <Quartus> I have a Zaurus around the place somewhere too.
21:01:47 --- join: nighty_ (n=nighty@sushi.rural-networks.com) joined #retro
21:08:57 <absentia> c700, 5600 and 760 .
21:09:06 <Quartus> Think mine is a 5500.
21:13:27 <absentia> very old.
21:13:33 <absentia> the z is sooo close.
21:13:41 <absentia> the mame on it only gives 25% performance :-<
21:14:17 <absentia> I "need" 800x600 screen (Qt dialogues have 800x600 as min on some for width?) ... vs 640x480 on current zaurus -- although i'll take the 640x480 over 320x200 or 160x160 any day :-)
21:14:38 <absentia> >128MB ram, as it has now... (because running linux takes  mem! argh)
21:14:56 <absentia> and faster than 400MHz arm cpu.  600mhz seems to be able to do mame with current arch compile.
21:15:44 <Quartus> yes, my gadget is not a new one. I don't use it much.
21:15:52 <absentia> oh, and built in wifi would be nice, since this thing only has one cf slot and I either get 6GB CF or WIFI -- but not both... but, like I said before -- when flying, it's great... 6GB -- with a 2GB SD card for media (mp3) -- you can get a very nice complete linux system in 6GB with room to dev
21:16:32 <absentia> man, my computer screens are a junkyard of languages
21:17:07 <absentia> I've got scheme in an emacs shel, scheme in a unix shell, lisp in a unix shell, forth in a shell, c++ in an editor, php in an editor.
21:17:34 <Quartus> heh
21:17:59 <Quartus> I have a wi-fi thingamabob for the zaurus. Hard on battery life.
21:19:05 <absentia> I have screens for shells for remote access... on 3 domains...  firefox, opera and .. forget the third browser... oh, yea, epiphany... 14 tabs on this opera, alone... SICP, intro to lisp, scheme... forth
21:19:24 <Quartus> well, there's nothing quite like leaping on your horse and riding madly off in all directions :)
21:19:29 <absentia> ya, I have a crapload of cf wifi -- I can get about 1.5hrs with full wifi on batter with the 7600... although I mostly use it plugged in.
21:19:43 <absentia> I just hope nothing crashes -- and I lose all the links :-)
21:20:05 <absentia> hahah, I have 3 computers, each with a different chapter of the SICP video lectures paused.
21:20:30 <Quartus> Too much caffeine in your diet? :)
21:20:45 <absentia> sometimes I wish we just had universal wifi so I could just go to a terminal and pick up where I left off -- i can do that now if I want to have a single home-base and use secure vnc... but that's not gonna give me remote audio, etc.
21:20:58 <absentia> actually, I don't drink caffeine.  makes me sleepy.
21:21:28 <absentia> almost done with thinking in forth.
21:21:28 <Quartus> Myself I try to work on one thing at a time, or none of them get the attention they require.
21:22:01 <absentia> yup.  I am normally that way, but I don't have time lately to concentrate, so i'm doing the speed scan on these things.
21:22:39 <Quartus> you have a deadline to meet as regards learning several programming languages at once?
21:23:12 <absentia> my death?
21:23:22 <Quartus> impending?
21:23:29 <absentia> I've just been lazy 'bout it over the last few years -- only really did one language per year.
21:23:34 <absentia> no, not impending.
21:23:37 <absentia> (that I know of)
21:23:49 <Quartus> I just doubt you'll get much out of speed-learning this sort of thing.
21:24:02 <absentia> I've realized that I don't want to learn any more languages -- I want to make my own and ... if I can get away with it, only use that one
21:24:03 <Quartus> Unless all you want is a 30,000-foot overview.
21:24:35 <absentia> three years ago it was python, two years ago it was c++, last year it was php (yuck) ... this year was lisp/scheme -- but I think I'm going to skip that and go to forth after I am done with sicp
21:25:06 <absentia> well, to me, all the languages seem the same -- the only difference is key words and syntax ... syntax just gets in the way... hence lisp (and now forth)
21:25:35 <absentia> CL has like 900 symbols -- that would take forever to learn.  c++ is huge -- if you use all parts of it -- it's also a moving target.. leads to many ways to blow off your entire le, etc.
21:25:41 <Quartus> Then you're doing yourself an injustice, particularly as regards a language like Forth.  It is not a mere keyword/syntax variant.
21:26:05 <absentia> well, it seems like it's closer tot he metal than lisp -- although lisp can be close.
21:26:28 <Quartus> But it's not just an Algol variant with differing syntax and keywords.
21:26:28 <absentia> forth is the closest to the AI that I was thinking about -- even closer than lisp.
21:26:44 <absentia> the auto calling of functions -- data already on the stack, etc... (well, that's not as close, but whatever)
21:27:40 <absentia> one of the issues I see with lisp/scheme/forth is all the variations due to dialects.
21:27:59 <Quartus> There's a Forth standard.
21:31:30 <absentia> what do you use forth for ?
21:31:50 <Quartus> Writing programs.
21:32:07 <absentia> heheh
21:32:13 <absentia> what... type... of programs ?
21:32:24 <Quartus> Any type.  It's a general-purpose programming language.
21:33:25 --- quit: nighty_ (Remote closed the connection)
21:34:27 <absentia> :-<
23:47:29 --- quit: Quartus_ ("used jmIrc")
23:59:59 --- log: ended retro/06.12.12