00:00:00 --- log: started forth/05.02.28
00:05:22 --- join: segher (~segher@blueice3n1.de.ibm.com) joined #forth
01:13:29 <ASau> swalters: use DEFER
01:13:43 <ASau> Dobry den!
01:16:17 --- quit: Topaz (Read error: 113 (No route to host))
01:45:22 --- join: fridge (~fridge@dsl-220-253-71-200.NSW.netspace.net.au) joined #forth
01:51:21 --- join: samc_ (~sam@203-114-131-141.inspire.net.nz) joined #forth
02:21:51 --- join: Topaz (jonny@wlan-195.ecs.soton.ac.uk) joined #forth
02:21:59 --- quit: Topaz (Client Quit)
02:39:53 --- quit: fridge ("smile if you had a naughty this morning!")
03:07:32 --- quit: SeaForth (Read error: 110 (Connection timed out))
04:08:25 --- quit: segher (Read error: 104 (Connection reset by peer))
04:26:44 --- join: segher (~segher@blueice1n1.de.ibm.com) joined #forth
04:45:39 --- join: tathi (~josh@pcp01375108pcs.milfrd01.pa.comcast.net) joined #forth
05:41:15 --- join: SeaForth (~SeaForth@c-24-1-126-202.client.comcast.net) joined #forth
06:06:41 --- quit: tathi ("leaving")
08:24:25 --- join: TheBlueWizard (TheBlueWiz@modem-139.nyc-tc03a.FCC.NET) joined #forth
08:28:45 --- join: I440r_ (~mark4@216-110-82-203.gen.twtelecom.net) joined #forth
08:33:36 --- quit: TheBlueWizard (Nick collision from services.)
08:33:55 --- join: TheBlueWizard (TheBlueWiz@modem-128.nyc-tc04b.FCC.NET) joined #forth
08:35:02 --- join: Serg[GPRS] (~z@193.201.231.126) joined #forth
08:36:59 <Serg[GPRS]> re
08:37:17 * Serg[GPRS] got FASM and RetroForth
08:38:04 <Serg[GPRS]> can RF compile flat files ? i dislike screens !
08:39:44 <ASau> Dobry vecer!
08:40:13 * Serg[GPRS] listens to Slade 'Cum on feel the noize'
08:40:36 <Serg[GPRS]> dobry ;)))
08:41:28 <ASau> Serg, use "dd conv=block cbs=64".
08:42:50 <Serg[GPRS]> i'd better comment out blocky stuff in src and write a stream compiler if i can
08:43:40 <ASau> I'm going to write streaming access for blocks in Forth.
08:43:51 <ASau> "File system" inside.
08:44:39 <Serg[GPRS]> just forget blocks ! (ex for vintage micro or alike)
08:44:44 <ASau> No.
08:44:55 <ASau> I'd better go with blocks.
08:45:31 <Serg[GPRS]> whom how, but i used to use FAR/Midnight F4 editor ;))
08:45:45 <ASau> I use Emacs.
08:45:59 <ASau> Either for NT and BSD.
08:46:32 <Serg[GPRS]> Emacs is warship-size swiss knife ;))
08:46:56 <ASau> Emacs is grown-up Forth.
08:47:14 <Serg[GPRS]> it is in Lisp btw ;))
08:47:15 <ASau> Everything's in LISP.
08:47:50 <ASau> I've been first. :)
08:47:59 <Serg[GPRS]> brr - deadly load of parens ;((
08:48:16 <ASau> Yeah.
08:48:27 <ASau> Not a one DUP or SWAP.
08:49:31 * Serg[GPRS] used to RPN/stack since MK-64 progged calculator and 'Science and life' magazine of 80-s
08:49:45 <ASau> BTW, where's our Siberian friend?
08:50:08 <ASau> I owned no one.
08:50:14 <ASau> I mean, MK-64.
08:50:36 <Serg[GPRS]> hehe, mee to - but it was my 1st reading on progging
08:51:23 <ASau> I did that on ZX 48K.
08:51:33 <Serg[GPRS]> i bought later both - MK-64 (brand new, w/ waraanty ticket of ex-USSR) and MK-52 (now in repair)
08:51:48 <ASau> KP580xxx
08:52:00 <ASau> uk3, IIRC.
08:52:08 <ASau> (IK3).
08:52:28 <Serg[GPRS]> my 1'st prog ran on BK-00010
08:52:41 <ASau> Does it have our
08:52:43 <ASau> ___
08:52:45 <ASau> /\
08:52:57 <Serg[GPRS]> pirated PDP-11 in keyboard ??
08:53:03 <ASau> Sign?
08:53:19 <Serg[GPRS]> sign what ???
08:53:31 <ASau> USSR quality sign.
08:53:49 <Serg[GPRS]> maybe, not recall
08:54:28 <ASau> BK-0010 was pretty good too.
08:54:54 <ASau> KP1801BM2, IIRC.
08:55:09 <Serg[GPRS]> not sure for comp's, but USSR photo cameras were terrific :((
08:55:30 <ASau> My "Smena-8M" died recently.
08:55:48 <ASau> I doubt any other camera could bear such conditions.
08:56:00 <Serg[GPRS]> a friend of mine undigged new - 'in oil' as said about firearms
08:56:00 <ASau> I mean modern cameras.
08:56:22 <ASau> Now I have "Sokol".
08:56:57 <ASau> It's heavy like a hammer.
08:57:05 <Serg[GPRS]> i got chinese Vivitar SLR, after my fake-Leika (FED-3) finally caught shutter seizure
08:57:15 <samc_> was it a mandate in russia that all vehicles must look ugly, or do russians have strange taste :p
08:57:40 <ASau> Our conditions differ from yours very much.
08:58:05 <ASau> Imported things have usually shorter life than our.
08:58:35 <samc_> yes, but ugliness is no cheaper nor more practical
08:58:51 <Serg[GPRS]> samc_: no it is
08:59:12 <Serg[GPRS]> ugly staright-angle things are easier to make and they are more robust
08:59:42 <samc_> a designer can always make something beautiful within any constraints
09:00:22 <Serg[GPRS]> why bother ? in planned deficit economy w/o competition, even ugly thing will be bought
09:00:29 <Serg[GPRS]> coz no alternative
09:00:41 <ASau> Serg, nothing at all.
09:00:46 <samc_> indeed
09:00:47 <ASau> You've got no clue.
09:01:15 <ASau> All those things were dirt cheap.
09:01:59 <ASau> And anyone could buy them.
09:02:08 * Serg[GPRS] denies any design beyond ergonomics
09:02:51 <samc_> Serg they were well designed ergonomically?
09:02:54 <ASau> Deficit comes from another reason: you can't make many enough things.
09:03:02 <ASau> ...produce...
09:03:14 <ASau> Some LOMO cameras are.
09:03:42 <samc_> I am still amazed that if someone had to put pen to paper anyway...
09:03:43 <ASau> Esp. "Smena-8M" and "Smena-Kompakt."
09:03:51 <Serg[GPRS]> ASau: 'anyone could buy' mostly an ink number in multi-day live queue ;((
09:04:07 <ASau> Serg, one more time:
09:04:20 <ASau> anyone could buy a car.
09:04:59 <ASau> It's because all the factories could not produce cars with appropriate rate, you had to wait.
09:05:02 --- join: Herkamire (~jason@h000094d30ba2.ne.client2.attbi.com) joined #forth
09:05:02 --- mode: ChanServ set +o Herkamire
09:05:08 <samc_> ASau, when are you talking about?
09:05:27 <ASau> 60-80s
09:05:38 <ASau> 1960-80s.
09:06:55 <ASau> Like we say "on fingers":
09:07:16 <ASau> you have deficit in two cases.
09:07:54 <Serg[GPRS]> in one: Money > goods*prices
09:08:14 <ASau> Either, you have too many buyers
09:08:30 <ASau> or too little things.
09:09:17 <ASau> In your words.
09:09:26 <ASau> 1) goods\approx 0
09:09:31 <ASau> 2) prices\approx 0
09:09:35 <Serg[GPRS]> printing extra money not backed up by goods = taking one's work for nothing, = covert slavery (if no option of another employer)
09:10:51 <ASau> Serg, microeconomics hides pretty enough from one's eyes.
09:12:02 <Serg[GPRS]> the 'covert slavery' extra money were all in Savebank, blown on reform - masses blame reformers in it, but the r00t of trubble much earlier
09:12:12 <samc_> Asau, who has no clue?
09:13:57 <ASau> Serg, all those money, you call "extra", were not converted to goods.
09:13:57 <samc_> that wasn't rhetorical
09:14:23 <Serg[GPRS]> yes.
09:14:29 <ASau> All those, who didn't "save" money, lived pretty well.
09:14:55 <ASau> Ivan Bunin, "Derevnja".
09:15:21 <Serg[GPRS]> ppl worked, state gave them 'extra' money not backed by goods, and the only option was to savebank it = they worked for nothing, were slaves
09:15:35 <Serg[GPRS]> to some degree
09:16:08 <ASau> Serg, meat was sold not only in state stores.
09:16:39 <Serg[GPRS]> life is not only meat and bread
09:16:51 <samc_> Asau preferred the old system?
09:17:00 <ASau> Those greedy ones waited in queues.
09:18:22 <ASau> Others bought meat in cooperative stores and paid sewer for clothes.
09:18:24 <Serg[GPRS]> dammit !
09:18:51 <ASau> samc_: I've seen the end of that system, Serg has not.
09:19:10 <Serg[GPRS]> i vote by both hands for surrendering to Germany :)))) or Sweden ;)))
09:19:48 <ASau> Ossi didn't like their "Vereinigung" soon after it's happened.
09:20:40 <samc_> Asau, you didn't answer my question ;)
09:21:06 --- part: segher left #forth
09:21:19 <ASau> I know dark sides of older system, but they're not those Serg speaks about.
09:22:21 <Serg[GPRS]> sure. 'ossi' grew in greenhouse and bah! - competition, no one would like !
09:22:22 <ASau> And, I have to admit, these sides become even darker.
09:22:58 <ASau> Serg, you don't understand what real "competition" means.
09:23:06 <Serg[GPRS]> i won't say 'become' , but i feel some threat
09:23:14 <ASau> Ask KC5TJA if you meet him.
09:24:24 <Serg[GPRS]> heh, i regulary think of jumping right into that 'competition', but i'm too pity coward to dare ;((
09:26:37 <ASau> Nah.
09:26:54 <ASau> You just live in this "greenhouse".
09:28:19 <ASau> You like this propaganda, but do not want to live by the law it pushes.
09:28:42 <Serg[GPRS]> yeah, and i feel this is 'not a real life', here in greenhouse
09:28:59 <Serg[GPRS]> it makes impotent vegetable out of me, dammit
09:29:06 <Frek> umm has atmel discontinued all AVRs but Tiny/Mega?
09:29:25 <Serg[GPRS]> and the longer i live, the harder the addiction and fear of 'real' life
09:29:36 <ASau> Better you spoke to any man, who works for 12 hrs/day.
09:43:36 --- join: Topaz (~top@cerberus.saywell.net) joined #forth
09:46:03 --- quit: Topaz (Client Quit)
09:46:04 --- quit: samc_ ()
09:46:46 --- join: Topaz (~top@cerberus.saywell.net) joined #forth
10:03:42 --- join: samc_ (~sam@203-114-131-141.inspire.net.nz) joined #forth
10:03:59 --- quit: Serg[GPRS] ()
10:04:32 <samc_> Raskin died
10:21:09 --- quit: samc_ ()
10:34:33 --- join: tathi (~josh@pcp01375108pcs.milfrd01.pa.comcast.net) joined #forth
10:45:43 <ASau> God kvaell, tathi!
10:45:54 <ASau> It looks like the world tragedy.
10:46:04 <ASau> Raskin died.
10:46:43 <ASau> Let's do nothing at all except repeating it again and again.
10:46:49 <ASau> Raskin died.
10:56:00 <OrngeTide> ASau, I sorta knew raskin.
10:59:59 <ASau> You know it we'd made such a tragedy of everyone's death we'd better died ourselves.
11:00:39 <ASau> And there'd be nobody on the Earth.
11:07:01 --- join: Serg[GPRS] (~z@193.201.231.126) joined #forth
11:07:03 <tathi> ASau: well, it's just taking a bit for everyone to find out.
11:07:59 <tathi> I just hope he did something sensible with the rights to the LEAP patent...
11:11:21 <I440r_> how did raskin die ?
11:11:33 <tathi> I have no idea.
11:11:45 <I440r_> and whats the leap patent ?
11:11:53 <tathi> first I heard of it was crc pointing out some web site with the announcement, but nothing else.
11:11:58 <tathi> sounds like he may have just died in his sleep.
11:13:27 <tathi> LEAP is how you get around in his "humane environment"
11:13:37 <tathi> hold down one of the leap keys and type the text you want to search for.
11:13:42 <tathi> (basically)
11:14:00 <Serg[GPRS]> hmm... what is the i-face for ?
11:16:10 <ASau> Interface.
11:17:07 <Serg[GPRS]> yes, for what ? common file tasks ? 3D building ? page layout ? searching ?
11:18:24 <tathi> THE -- The Humane Environment
11:20:10 <tathi> just text, so far.
11:20:41 <tathi> he made it sound like he had plans for extending it to images and other stuff, but hadn't worked out the details yet.
11:20:47 <tathi> and...it doesn't have files.
11:29:44 <ASau> True Forth Environment doesn't have files too.
11:30:22 <Serg[GPRS]> crap. real life has files ;))
11:30:36 <Serg[GPRS]> and directories, if not an utter mess
11:31:05 <tathi> well, that was the whole point of his thing, that you don't need them.
11:31:16 <tathi> you just search to get places.
11:31:23 <ASau> Files are only for those who want to dig into details.
11:31:44 <ASau> Real men do everything with a knife.
11:31:53 <tathi> And...AFAIK he used the system on the Canon Cat, so it works really well in practice.
11:32:00 <Serg[GPRS]> even brushing teeth ?
11:32:20 <tathi> Seems to be pretty sound interface design, from what I can tell (though I haven't read all that much of the literature).
11:33:21 <OrngeTide> I440r_, raskin was diagnosed with pancreatic cancer in Jan 2005. that means he only had weeks to live when he found out. it's kinda scary that it was so quick.
11:34:12 <I440r_> :/
11:34:28 <tathi> ah. yeah, that's a bad one.
11:34:33 <OrngeTide> operating systems are moving to search-based interfaces. desktop search is going to be the standard interface in a few years.
11:35:02 <Serg[GPRS]> rapid death is the best one, although ;((
11:35:55 <Serg[GPRS]> weeks of feeling condemned is better than years
11:36:31 <OrngeTide> I suppose. I think I would have an easier time accepting it if I was given 5 years notice.
11:38:06 <Serg[GPRS]> no-o-o-o ! i wanna live in happy ignorance and then just fall w/o any mental suffering
11:38:38 <tathi> and it is possible to beat cancer if you're determined enough, but you need time to get things turned around.
11:39:42 <Serg[GPRS]> aha-a-a ;((
11:40:02 --- quit: Serg[GPRS] ("bails out - sleep angel at six o'clock !")
11:40:32 <OrngeTide> tathi, pancreatic cancer is notorious for being difficult to treat.
11:41:57 <tathi> OrngeTide: yeah, I know.
11:46:52 <Topaz> hm hm
11:47:17 <Topaz> anybody know any good tutorials about cross-compiling in FORTH?
11:49:35 <Topaz> my current system is a tad crazy
11:50:30 <Topaz> my FORTH is hosted on a python-written uC emulator, and some of the words (awkward things like colon) are written in python, with cunning assembly hooks
11:51:35 <OrngeTide> the easiest is to just create a vocabulary on the host. words in that "target vocabulary" write to a target memory block when they are executed. for example IF THEN in your vocab would not branch when you executed them. they would just record their existance in the target memory.
11:52:13 <OrngeTide> is your target going to be using native forth or direct threaded?
11:52:20 <Topaz> subroutine threaded
11:52:24 <Topaz> (ie, compiled)
11:52:52 <OrngeTide> okay. that's almost a native forth. yea. the target technique I mentioned is probably the easist way.
11:53:31 <OrngeTide> I use it for my forth os that is compiled to ARM from gforth. i'm just using gforth as a fancy macro assembler :)
11:54:06 <tathi> Is that for the gumstix boards?
11:54:16 <Topaz> hmm, so the words basically just modify the target dictionary instead of the host one?
11:54:52 <Topaz> (what's the easiest way of managing two dictionaries?)
11:55:25 <OrngeTide> stuff like : can be tricky to implement. but what you are doing is reserving a block of memory in the target from : to ; .. filling it with instructions and : defines a word in the target vocabulary on the host that has the offset of this subroutine in target memory. the word defined will write a call instruction to this address that is stored into the current position of target memory.
11:56:37 <OrngeTide> yea. it's easier to have different vocabs. that way you can flip to the host's to do operations. like write target memory to disk (perhaps you need it as intel hex or motorala srec so you can send it to a jtag program)
11:57:55 <OrngeTide> Topaz, will your target has an interactive forth on it? or is it just an application written in forth? the later is easy. the first requires you to manage two completely different dictionaries.
11:57:59 <Topaz> hmm, i should probably learn FORTH better before i start embarking on this stuff ;)
11:58:11 <Topaz> ah, yeah, it's the latter
11:58:22 <Topaz> ah, yeah, i see now
11:58:42 <Topaz> i'm forgetting that STC FORTHs don't really have a dictionary
11:59:16 <OrngeTide> yea. I tend to think of things like that as a fancy macro assembler :)
11:59:21 <Topaz> FORTH is quite amusing, in that you can write a FORTH without really knowing how to program in it
11:59:37 * Topaz tries to work out where to start
12:00:16 <OrngeTide> oh yea. i totally don't know forth and I have part of a working cross assembler written in forth. :)
12:00:40 <Topaz> i vaguely understand how CREATE DOES> works, just no real clue how to use it ;)
12:00:54 <Topaz> i guess learning-by-example is crucial
12:01:49 <Topaz> hmm, I guess gforth would do as a reasonably boring 'portable' host for my thing
12:02:10 <Topaz> pforth is smaller
12:02:12 <OrngeTide> i like create does> .. some people really despise it.
12:02:32 <Topaz> it's a very clever idea
12:02:46 <OrngeTide> i'm not that picky about the host forth. since it doesn't actually have to do much.
12:02:49 <Topaz> but emphasises the need for being able to clearly understand the difference between compile-time and run-time...
12:02:58 <OrngeTide> IsForth is good in general though.
12:03:12 <OrngeTide> yea. that's something that is hard to wrap your head around. 
12:03:21 <Topaz> what will isForth run on?
12:03:32 <Topaz> (i have crazy visions of trying to run my host on some embedded device at some point)
12:03:32 <OrngeTide> i've found it can help if the host and the target are completely different architectures.
12:03:36 <OrngeTide> IsForth is for Linux.
12:03:44 <Topaz> any architecture?
12:03:57 <OrngeTide> x86 only. it's written mainly in nasm and itself.
12:04:24 <OrngeTide> herkforth runs on powerpc. it's pretty good too.
12:04:26 * tathi did a PPC port...
12:04:30 <tathi> (of IsForth)
12:04:31 <OrngeTide> tathi, neat
12:04:38 <Topaz> though i guess if i don't use too many clever features, it'll be slightly portable
12:04:40 <OrngeTide> tathi, link?
12:04:50 <Topaz> hmm, i'll try pforth to start
12:05:07 <tathi> http://qualdan.com/isforth/
12:05:08 <OrngeTide> IsForth tends to be what I440r thinks forth should be like, rather than ANSI's idea of forth.
12:05:29 <OrngeTide> tathi, i'll have to try it out when i get home tonight.
12:05:58 <tathi> OrngeTide: cool, I'll be interested to see if it runs on other computers than mine :)
12:06:24 <OrngeTide> i wonder how much of a PITA it would be to convert the linux syscalls to darwin syscalls.
12:07:22 <tathi> oh, you're running OS X, not Linux?
12:07:59 <Topaz> hmm, so i need to redefine colon
12:08:33 <tathi> I think I did a couple of things that don't go over at all well with BSD...
12:14:39 --- quit: warpzero ("leaving")
12:26:37 --- join: warpzero (~warpzero@wza.us) joined #forth
12:33:29 <Herkamire> herkforth doesn't work on Mac OS BTW
12:35:45 <I440r_> tathi what did you do that bsd wont like ?? :)
12:38:54 <OrngeTide> tathi, yea. I use osx exclusively on my mac mini. i dual boot my ibook. 
12:39:44 <OrngeTide> Herkamire, yea. i don't think there are any good forths for OSX.
12:39:55 <OrngeTide> oops. lunch time.
12:40:18 <tathi> I440r_: I didn't incorporate the data section yet, for one.
12:42:20 <Topaz> hmm, so for my cross compiler, i redefine a lot of the basic words, and then presumably need to redefine all the compiling words (IF/THEN/ELSE/BEGIN) to use them
12:42:35 <I440r_> what do you mean incorporate ?
12:42:49 <I440r_> you have separate code and data space ?
12:42:52 <I440r_> how did you do that ?
12:43:26 <tathi> I440r_: no, ELF headers, I mean.
12:44:13 <tathi> remember, under BSD a brk only works on a data section?
12:44:19 <I440r_> oh
12:44:20 <I440r_> ok :)
12:44:24 <I440r_> ya
12:46:59 <Herkamire> OrngeTide: I don't realy know what's out there. I haven't used OSX in so long. I've heard good things about MOPS though.
13:24:00 --- quit: I440r_ ("Leaving")
13:45:43 --- quit: tathi ("restarting X")
13:57:20 --- join: tathi (~josh@pcp01375108pcs.milfrd01.pa.comcast.net) joined #forth
14:10:45 <crc> back
14:49:51 --- join: aum (~aum@60-234-138-239.bitstream.orcon.net.nz) joined #forth
14:50:47 * crc completes a major cleanup to the retroforth 8.0 codebase
15:28:55 --- quit: Topaz ("Leaving")
15:51:48 --- join: crc2 (crc@bespin.org) joined #forth
16:14:33 <swalters> crc, are you the author of retroforth?
16:16:44 <arke> swalters: yes he is :)
16:16:47 * arke is at work, grrr
16:17:06 <crc> swalters: yes
16:17:49 <swalters> nifty.  I spent about ten minutes poking at it last night.  Neither the tarball, nor the web-generated binaries worked for me, but I can't tell you any more than that since I only spent 10 minutes looking at it.
16:17:58 <swalters> I like your philosophy, at the very least.
16:18:19 <swalters> (btw, I'm on gentoo, which might be part of the problem)
16:18:46 <crc> are you using an x86 system?
16:18:51 <swalters> yeah
16:19:10 <swalters> Are there currently problems with x86?
16:19:11 <arke> gentoo is quite quirky ;)
16:19:26 <arke> No, x86 is the only thing it runs on :D
16:19:27 <crc> you might try the latest tarballs from the retro.tunes.org site again... 
16:19:36 <crc> I just updated them :)
16:19:51 <crc> RetroForth won't work if you have the SELinux patches installed
16:19:52 <swalters> It's becoming less so, but it can do some pretty strange things by not compiling with standard options.
16:20:02 <swalters> no selinux
16:20:35 <swalters> Like I said, I can't be specific enough to diagnose it.  Let me do some detective-work before we have a long, drawn out convo about it.
16:20:40 <crc> ok
16:28:09 --- quit: aum ()
16:52:29 <arke> what is SELinux?
16:52:59 <crc> security enhanced linux
16:53:08 <crc> developed by the NSA
16:53:45 <crc> doesn't allow writeable code sections :(
16:54:30 <arke> heh
16:54:31 <arke> thats why
16:54:31 <arke> :D
17:31:31 --- quit: tathi ("leaving")
17:36:42 <arke> bbl
17:52:58 <swalters> no writable code sections is only one portion of what the NSA has done to it.  I'll be glad when all x86 chips support a exec flag for the pages.
17:54:12 <swalters> Strike that...  I'll be glad when we move over to chips similar to Moore's designs.
17:54:38 <arke> :D
17:54:56 <crc> :)
17:56:52 <swalters> which reminds me of a question I had.
17:57:44 <swalters> Do any of the forth-chips use one register for top-of-parameter-stack and one for second-of-parameter-stack?  This would make sense since most of the operations act on the top two.
17:58:46 <arke> i think its implied in the implementation :D
17:59:00 <arke> (remember, "drop" "dup" "over" "+" etc. are a single machine instruction)
17:59:17 --- join: Sonarman (~snofs@adsl-64-160-166-168.dsl.snfc21.pacbell.net) joined #forth
18:00:44 <Herkamire> swalters: I'm not sure if they are registers exactly, but tos, and sos certainly are accessible to the instructions
18:01:41 <swalters> Beyond even making these single opcode operations, having a few "registers" to cache the top few items on the stack would probably speed things up greatly.  That way the proc could continue operating while the chip did a fetch from memory of anything it needs.
18:03:05 <swalters> Herkamire, I'm using the defition of register loosely here.  By register, I mean an on-chip store that doesn't have to be fetched from memory.  A series of stores that shifted in unison as the stack was popped and pushed (including for ops like "+") would be a register by this ad-hoc defintion.
18:03:30 <arke> swalters: typically Forth chips have on-chip stacks
18:03:35 <crc> yup
18:03:54 <swalters> In other words, since you need fewer "real registers," the extra could be used as a sort of caching mechanism.
18:05:26 <swalters> Does it provide a stack pointer so that large stacks may be continued in memory?
18:06:47 * crc doesn't believe in large stacks
18:06:53 <Herkamire> what I think of as "forth chips" store the stacks on-chip
18:07:14 <swalters> So, the entire stack is always on chip?
18:07:23 <crc> normally, yes
18:07:33 <Herkamire> how else would you get single cycle instructions?
18:08:19 <swalters> (sorry if these seem like dumb questions, but I'm trying to get my bearings on the subject)
18:08:37 <crc> n/p
18:09:13 <swalters> Unstructions aren't all single-cycle.  You can incriment the tos in one cycle.  Modulus would take memory (if what I recall of asm programming still applies)
18:10:01 <arke> i have to run home now
18:10:33 <swalters> Also, it seems that forth will eat a few stack items, then push a few then eat a few then push a few.
18:11:22 <swalters> On average, the stack won't grow or shrink by much.  This would give the proc some freedom to continue working while stack items are fetched from memory.
18:11:33 <swalters> errr... Modulus takes many cycles.
18:11:52 <crc> on what architecture?
18:12:07 * crc can envision an arch where all opcodes take one cycle
18:12:28 <Herkamire> swalters: you're still thinking register machines
18:12:39 <Herkamire> forth cpus generally do everything in one cycle
18:12:45 <Herkamire> the stack has nothing to do with memory
18:12:49 <Herkamire> there is no stack pointer
18:13:39 <swalters> doesn't such a hard limit on the stack cause problems?
18:13:44 <crc> no
18:14:00 * crc hasn't gone above 10 items deep in nearly a year
18:14:17 <crc> small words, shallow stacks :)
18:14:50 <swalters> that seems very unlikely to me, but I'll attribute that feeling to my inexperience for the moment.
18:15:09 <Herkamire> chuck moore designed a cpu core with circular stacks
18:15:23 <crc> moore's designs use a circular stack of what, 15 or 16 elements?
18:15:24 <swalters> Thanks for entertaining my babblings.  I must run now.
18:15:29 <crc> bye
18:15:32 <Herkamire> later
18:15:37 <TheBlueWizard> bye swalters
18:16:02 <swalters> I believe the article I skimmed said 16 for the circular stack.  I am researching, I'm just still sorting out the new ideas.  :-P
18:21:23 --- quit: KB1FYR (Read error: 104 (Connection reset by peer))
18:22:00 --- join: KB1FYR (~Alex@196-220.suscom-maine.net) joined #forth
18:37:07 <crc2> just a quick message to avoid timing out
18:37:45 --- quit: TheBlueWizard (Nick collision from services.)
18:38:01 --- join: TheBlueWizard (TheBlueWiz@modem-016.nyc-tc03a.FCC.NET) joined #forth
19:10:38 <TheBlueWizard> bye all
19:10:52 --- part: TheBlueWizard left #forth
19:38:06 --- quit: rastm2 ("leaving")
19:48:42 --- join: raystm22 (~color4th@adsl-69-149-47-230.dsl.rcsntx.swbell.net) joined #forth
19:49:46 <swalters> *LOL*  I'm helping my mother with a project for her advertising class.  Case modders is a target audience for the product she dreamed up.
19:50:18 <swalters> She's not getting her paper written because she's too busy reading the case mod sites and ooh'ing and ahh'ing.
19:59:13 <crc> goodnight
19:59:50 <swalters> night man
21:04:26 <swalters> Looking again at Moore's F21 proc, it seems designed for embedded applications.  The return and parameter stacks are 17 and 18 deep, respectively.
21:04:49 <arke> :D
21:05:17 <swalters> In 1997, it was running at 100 to 300+ mips (depending on memory architecture)
21:05:52 <swalters> In relation to the numbers you guys were throwing out earlier, this is very generous in stack size.
21:07:06 <swalters> How would the architecture be affected if you scaled it up to something comparable to an average modern computer?  (that's about 8k to 10k mips)
21:07:47 <swalters> Would huge apps, like web browsers and bloated word proccessors affect how deap of a stack you needed?
21:17:52 <arke> why would you need 8-10k mips if you dont have linux/windows hogging up 90% of the cycles?
21:17:56 <arke> :)
21:22:08 <swalters> What would you suggest?  A web browser on bare metal?
21:22:31 <swalters> :-P
21:23:47 <arke> http://www.extremefunnypictures.com/images/maze.swf
21:24:36 <swalters> A friend of mine wants to register "dothesepantsmakemybuttlookbig.com"
21:24:56 <arke> lol
21:25:05 <arke> i would figure Twix did
21:25:06 <arke> lol
21:36:48 <swalters> blah... it's so hard to stay focused and learn the fundamentals of forth.  There's so many more interesting advanced aspects of it.
22:05:21 --- quit: Sonarman ("leaving")
22:29:51 <arke> swalters: :)
22:37:08 --- quit: crc2 (Ping timeout: 14400 seconds)
23:49:06 <Herkamire> swalters I think that proccessor would be plenty to run _well written_ web browsers and word proccessors etc
23:50:05 <Herkamire> as microsoft and others continually prove, you can make GUIs that run slow on _any_ hardware
23:59:46 <Herkamire> I hear that jpeg decoding is somewhat cpu intensive
23:59:59 <Herkamire> but most of the rest of web-browsing is pushing data around
23:59:59 --- log: ended forth/05.02.28