Minutes of April 7,1988 'FIGGY BAR' Attendees: [[Gary] GARY-S] [D.BARKER1] [JAX] [R.SULLIVAN1] Minutes: is here. <[Gary] GARY-S> hi guy hi Gary looks like you got it set up right. <[Gary] GARY-S> I got here early - since I... <[Gary] GARY-S> didn't know how many might.. <[Gary] GARY-S> have got my message wiuth the wrong times. <[Gary] GARY-S> ga ok well I really shouldn't stay... got a lot of reviewing to do before I show up... tomorrow. Ya think you can handle it? . <[Gary] GARY-S> Just you and me - we can move to room one or leave it as is - or go and I'll give it a shot. hey you need the experience anyway :-) <[Gary] GARY-S> Jennifer is taking notes - I know to mail them so adios if you wish <[Gary] GARY-S> ga ok, have a good time! <[Gary] GARY-S> thanks - me and Jenny see ya later don't worry someone always shows up! <[Gary] GARY-S> so go earn some money bye bye I'm already earning it! :-) has left. is here. <[Gary] GARY-S> Hi Kevin - we're it <[Kevin] APPERT> hi Gary <[Kevin] APPERT> Hi Jennifer (grin) <[Gary] GARY-S> Dennis was here but split <[Kevin] APPERT> I'll have to take off in a min <[Gary] GARY-S> She's not much of a talker <[Gary] GARY-S> What's on your mind <[Kevin] APPERT> nothing much at the moment <[Kevin] APPERT> what do you do with Forth? Work? <[Gary] GARY-S> Do you use Forth in your work - now ? <[Gary] GARY-S> I hear an echo - go first <[Kevin] APPERT> yup. We build space inst. w/ uProcessors <[Gary] GARY-S> That's right ! <[Gary] GARY-S> I remember now <[Kevin] APPERT> My assembler and development system are in Forth <[Kevin] APPERT> . <[Gary] GARY-S> Did you see the one-timers about the comming conferences <[Gary] GARY-S> ga is here. <[Kevin] APPERT> nope. anything interesting? I don't generaly read that stuff <[Kevin] APPERT> hi len <[leonard] NMORGENSTERN> Hi there <[Gary] GARY-S> You bet - Got Feierbach comming the 20th... <[Gary] GARY-S> and something that should also light your fire.. <[Kevin] APPERT> what's he talking about - super 8? <[Gary] GARY-S> on May 4th - the guys from John Hopkins.... <[Gary] GARY-S> that craeted the 32 bit engine <[Gary] GARY-S> ga <[Kevin] APPERT> that sounds great... I'll mark my calander <[Gary] GARY-S> Hi len <[leonard] NMORGENSTERN> hi <[leonard] NMORGENSTERN> . <[Gary] GARY-S> What do you want to know about the Super-8 <[Gary] GARY-S> ga <[Gary] GARY-S> ga Leonard <[leonard] NMORGENSTERN> On the last Figgy bar ... <[leonard] NMORGENSTERN> Jax suggested a protocol. It is easier to ... <[leonard] NMORGENSTERN> edit than the one we have been using ... <[leonard] NMORGENSTERN> It goes like this. A blank line to get ... <[leonard] NMORGENSTERN> attention, three dots at the end of a line . . . <[leonard] NMORGENSTERN> and a . alone on a line to end. <[leonard] NMORGENSTERN> . <[Gary] GARY-S> I have no problem with either, we will... <[Gary] GARY-S> have to use the formal protocol... <[Gary] GARY-S> on the scheduled conferences... <[Gary] GARY-S> especially if we get other GEnie users. <[Gary] GARY-S> ga <[leonard] NMORGENSTERN> <[Gary] GARY-S> ga Leonard <[leonard] NMORGENSTERN> It occurs to me that a ga alone on a . . <[leonard] NMORGENSTERN> line is as good as dot . <[leonard] NMORGENSTERN> ga <[Gary] GARY-S> agreed. <[Gary] GARY-S> Kevin - were you asking about Super-8 <[Gary] GARY-S> . <[leonard] NMORGENSTERN> <[Gary] GARY-S> ga Leonard <[leonard] NMORGENSTERN> No, a slip of the finger <[leonard] NMORGENSTERN> ga <[Gary] GARY-S> kev <[Kevin] APPERT> I assume that's what... <[Kevin] APPERT> gary F. is going to talk about. <[Kevin] APPERT> . <[Gary] GARY-S> true - with how he is running... <[Gary] GARY-S> F-83 on it. Jax is using this rig himself <[Gary] GARY-S> ga is here. <[Gary] GARY-S> I gather with ll. <[Kevin] APPERT> hi JAX <[leonard] NMORGENSTERN> Hi <[Gary] GARY-S> Jax - how do you like the 8 ? <[Gary] GARY-S> we were just talking about it <[Gary] GARY-S> ga what happened to room one? Love the 8 the Peoples Forth Engine!! :-) . <[Gary] GARY-S> room one is Dennis ' <[Kevin] APPERT> it was only a virtual room. It got swapped out. :-) . <[Gary] GARY-S> ga Jax I have a new religion! The Inmos Transputer! Am writing assembler for same in JForth! . <[Gary] GARY-S> are you serious ? <[Kevin] APPERT> <[leonard] NMORGENSTERN> <[Gary] GARY-S> I heard bad talk about that chip ga Yes. Its a kludge. Len? <[Kevin] APPERT> on the Atari ST, JAX? <[leonard] NMORGENSTERN> What is the inmos transputer? . . . <[Gary] GARY-S> ga Kevin <[Kevin] APPERT> . <[leonard] NMORGENSTERN> . <[Gary] GARY-S> ga Kevin The transputer ... is a parallel engine ... technologically something akin to what the 8080 was in its day ... it is a kludge ... but a marvel nonetheless ... it leads the way to desktop parallel in the mid nineties ... at business-like prices ... it has the potential to demonstrate ... the superiority of TIL's on development time ... in the commercial development environment ... <[Kevin] APPERT> oops, gotta go. 'night all that will spring up around parallel <[Kevin] APPERT> has left. gnite . <[Gary] GARY-S> now - ga Kevin <[Gary] GARY-S> oops - I mean Len <[leonard] NMORGENSTERN> pmi (pardon my ignorance) but what is a TIL? <[leonard] NMORGENSTERN> . TIL .. THreaded Interpretive Language ... i.e. say , er ... FORTH!! . <[Gary] GARY-S> If he won't I will <[Gary] GARY-S> or its sisters <[Gary] GARY-S> Tell me Jax, <[Gary] GARY-S> I heard it was so kludgy as to... <[Gary] GARY-S> be useless... <[Gary] GARY-S> why do you like it. <[Gary] GARY-S> ga Because tho it is a pain in the ass considered as a 32-bit processor ... It has "orthagonalised" the parallel communication protocol ... quite successfully ... the complaints you here have two origins ... One: The difficulty of programmers to adapt to the concepts and algorithms ... and TWO: The troubles like microprocessors had when they were young: New technology . <[Gary] GARY-S> Question - can you get ( easily ) development... <[Gary] GARY-S> goodies for it or is all roll your own ? <[Gary] GARY-S> ga The processor and the language are wrapped in one another ... this is an engine designed by Computer Scientists with a capital "C>S>" and the literature is redolent with high-sounding terms & equations ... but yes, there is a parallel compiler ... it is a theoretical model as radical as ALGOL was in its day ... it is called OCCAM ... but it is a compiled model. You can implement the theoretical aspects ... in FORTH if you like. I like. WORK!!!! . <[Gary] GARY-S> support ? hardware ? Plenty. Call Inmos for lit. THey are generous. THey are trying .. to sell a few! I haven't bought one yet ... But the literature is precise enough ... to allow me to write an assembler ... and start on my Parallel FORTH ... without daily access to the machine! . <[leonard] NMORGENSTERN> Fascinating! <[leonard] NMORGENSTERN> . <[Gary] GARY-S> sounds good ! hack still lives <[Gary] GARY-S> ga leonard <[leonard] NMORGENSTERN> I just wanted to say that it sounds like a real <[leonard] NMORGENSTERN> challenge. How do you mount the chip if <[leonard] NMORGENSTERN> you get one? <[leonard] NMORGENSTERN> . <[Gary] GARY-S> ga jax You Thanx. You put them in a board. That's it ... they don't share the bus!!!! No capacitive penalty for loading ... No bus contention as you add processors ... The concept is the "wire" .... a 10Mbit serial byte transfer connection ... four sets of two wires each coming out of each transputer ... ( vanilla transputer ... you can generate special configs) You just swap bytes. That's the whole trip. Just swapping bytes. SO you swap a hunk of 1024 bytes and what have you swapped between Xputers? <[Gary] GARY-S> sounds tailor made for Forth RIGHT! BLOCKS LIVE !!!! . <[Gary] GARY-S> Puts you back in your stand-alone pulpit :-) . <[Gary] GARY-S> ga jax The architecture is fascination .. but the language is tricky ... since it is a 32 bit architecture ( tho there are 16 & 24 bit xputers) you would expect longword handling in the assembler ... WRONG Constants are loaded *4* bits *at* *a* *time* ... This is why people complain ... but they don't understand ... that that is what makes byte-transfer parallelism practical. The result is that you can pack ... a four bit opcode ... like " ldc " ( load constant) ... with a four-bit data word ... in one one-byte serial message to the neighboring xputer ... at 10 megabaud! SO now you see why people say it's a kludge ... It's waiting for technology & programming to catch up ... with a noble concept. . <[leonard] NMORGENSTERN> <[Gary] GARY-S> unfinished is a word I hear a lot. <[Gary] GARY-S> I was trying... <[Gary] GARY-S> to remember who is also using... <[Gary] GARY-S> OCCAM, bvut I'm drawing a blank. dan miller <[Gary] GARY-S> ga leonard <[leonard] NMORGENSTERN> Is there anything written on it ... <[leonard] NMORGENSTERN> beyond what you get from Inmos? <[leonard] NMORGENSTERN> . <[Gary] GARY-S> ga jax There have been a few magazine articles ... The tech literature is fascinating ... but veddy blittish ( British) :-) They will send you some reprints ... <[Gary] GARY-S> Computer design had a series... if you ask nice! ga gary <[Gary] GARY-S> on it abnd other chips, but about... <[Gary] GARY-S> a year ago. <[Gary] GARY-S> ga jax Len, you should take a look at the literature from INMOS ... being an old bit-banger yourself ... you might appreciate how sofistycated ... the architecture has gotten ... remmeber you told us one night ... about "Nine byte instructions ... with a stop byte" ??[A Well, this is a lot like that... this is the dawn of a new architecture ... that's liable to put old-timers like jax outa business ... if i don't hurry up & learn!! :-) . <[leonard] NMORGENSTERN> <[Gary] GARY-S> ga len ga <[leonard] NMORGENSTERN> I couldn't work on it unless I could get <[leonard] NMORGENSTERN> a board all mounted with power supply etc... <[leonard] NMORGENSTERN> Although I put my old SOL together by myself... <[leonard] NMORGENSTERN> I have found that hardware & I don't mix <[leonard] NMORGENSTERN> . <[Gary] GARY-S> ga jax Excuse me for neglecting to expand ... upon the hardware side of things as amply as on the theory ... of the parallel model. Inmos ... and several third parties ... make quite a variety of boards ... The transputer is a Momma to fix up with a board ... but the interconnection problem is simplicity itself. So what you get yourself ... is a board with a lot of holes for the transputers ... you can't afford yet ... and plug them in one at a time .... since the OCCAM compiler views configuration of xputers ... fundamentally "transparently" ... although it allows you to configure for hardware ... basically, it handles the processors itself. Parallel FORTH would be more flexible ... but somewhat demanding, jaxthinx ... once we get up to speed, however .... i suspect that the advantage FORTH has in development time over compiled langs would be augmented by the square of the number of transputers! . <[leonard] NMORGENSTERN> <[Gary] GARY-S> ga len <[leonard] NMORGENSTERN> pmi, you mount the transputers on a board... <[leonard] NMORGENSTERN> but what do you mount the board on? <[leonard] NMORGENSTERN> . <[Gary] GARY-S> ga jax You can have a host ... or a "master xputer" with slaves. Your choice ... the more sophisticated xputers ... have more sofisticated video services... really, it is a COMPLETE development system ... except ... no FORTH yet! . <[Gary] GARY-S> address for inmos , please <[Gary] GARY-S> ga jax coming up. ga <[Gary] GARY-S> anything leonard ? ga <[leonard] NMORGENSTERN> <[Gary] GARY-S> ga leonard <[leonard] NMORGENSTERN> I'd better watch out, or I ... <[leonard] NMORGENSTERN> might get hooked on this It sounds like .. <[leonard] NMORGENSTERN> the kind of thing I would like to do, but .., <[leonard] NMORGENSTERN> I don't know if my theory is up to it . <[leonard] NMORGENSTERN> . <[Gary] GARY-S> wait till after Gary f,'s conf ... <[Gary] GARY-S> you might have a choice toi make. <[Gary] GARY-S> ga leonard <[leonard] NMORGENSTERN> <[leonard] NMORGENSTERN> I can't make the Wednesday conferences. . . <[leonard] NMORGENSTERN> until summer, so I will download them & look. <[leonard] NMORGENSTERN> . <[Gary] GARY-S> I hate to hear that, but the... <[Gary] GARY-S> guests chose those dates.. <[Gary] GARY-S> and I do want to keep them happy. <[Gary] GARY-S> ga <[leonard] NMORGENSTERN> <[leonard] NMORGENSTERN> Gary is a good writer, and if JAX is there. . . <[leonard] NMORGENSTERN> there should be a lot of good material . . . <[leonard] NMORGENSTERN> in the conference! <[leonard] NMORGENSTERN> . <[Gary] GARY-S> true - I also hope for a full.. <[Gary] GARY-S> turnout to justify more.. <[Gary] GARY-S> conferences. <[Gary] GARY-S> any guests you'd like to see me go for ? <[Gary] GARY-S> ga <[Gary] GARY-S> hello - any suggestions leonard ? <[leonard] NMORGENSTERN> I will keep my ears open. How about Mike Perry? <[leonard] NMORGENSTERN> . <[Gary] GARY-S> he would be good ! fo sure Try INMOS Corporation, PO box 16000 Colorado Springs CO 80935 <[Gary] GARY-S> Good suggestion. <[Gary] GARY-S> Cute - RPN addresses There is also a company ion Ward Road in Wheat Ridge CO ... That is making 4-xputer add-ons ... for the SUN 68020 workstation ... so you might watch the UNIX literature ... also an add-on is coming for the Mac ... and a drop-in for speedy grafix for the Amiga! so stay tuned to the world of parallel ... and read MicroTimes, California's largest monthly free-distribution computer publication, where a somewhat lite-weight article ... on parallel by jax ( jack woehr) is coming up in a month or three. . <[leonard] NMORGENSTERN> <[Gary] GARY-S> address for utimes ? <[Gary] GARY-S> ga leonard <[leonard] NMORGENSTERN> I am getting a MAC <[leonard] NMORGENSTERN> SE soon. After I learn it and learn MacForth. . . <[leonard] NMORGENSTERN> I might want to try the transputer! <[leonard] NMORGENSTERN> . <[Gary] GARY-S> what about MicrTimes address Jax Microx is BAM PUBLICATIONS, Inc, 5951 Canning St., Oakland, CA 94609 12 issues $24 OR FREE ON THE NEWSTAND! . <[leonard] NMORGENSTERN> <[Gary] GARY-S> aha - <[Gary] GARY-S> ga leonard <[leonard] NMORGENSTERN> If I get the issue with Jax's article. . . <[leonard] NMORGENSTERN> I will send it to you, Gary <[leonard] NMORGENSTERN> . Had an articel <[Gary] GARY-S> sounds good - i appreciate <[Gary] GARY-S> ga jax on JForth in the was it Feb? issue ... but it was eclipsed ... by the article about the Brain Surgeons using Amiga + JForth ... to save lives !!! . <[leonard] NMORGENSTERN> I saw those! <[leonard] NMORGENSTERN> . <[Gary] GARY-S> do what !!! Feb issue ... right ... yes ... they have to model three d in real time ... and rotate and slice graphically ... so they hooked jforth to Dpaint ... or was it Aegis Draw ??? ... and they cut people open on the basis of their renderings! Forth has the same appeal to techies ... that BASIC always had ... instant gratification :-) ! . <[Gary] GARY-S> That's amazing as hell ! <[leonard] NMORGENSTERN> <[Gary] GARY-S> ga leonard <[Gary] GARY-S> <[leonard] NMORGENSTERN> Why so? Forth is ideal for real-time . . . <[leonard] NMORGENSTERN> applications <[leonard] NMORGENSTERN> . and jforth <[Gary] GARY-S> cupla reasons... is the fastest I've ever seen . <[Gary] GARY-S> one it was only in a regional pub <[Gary] GARY-S> and two - it was rigged with a paint program. <[Gary] GARY-S> I like it <[Gary] GARY-S> ga jax It was at UCLA he had been using an $80000 DEC ... but found the Amiga did it better ... with JForth! . <[Gary] GARY-S> does micro t always have forth - or regular ? <[Gary] GARY-S> ga Not always ... it is consumer oriented ... but i have a slight in with the editor ... anand besides ... she is practical ... if FORTH is viable .... then she'll write about it ... but FORTH is an op system, not just a language ... and joe buyer really doesn't understand that kind of thing .. thinking that that approach went out ... with Apple BASIC ... . <[Gary] GARY-S> This conference should promote a few inquiries. <[Gary] GARY-S> I hope she is ready <[Gary] GARY-S> ga jax <[leonard] NMORGENSTERN> <[Gary] GARY-S> ga leonard Her name is Mary Eisenhart ... ga len <[leonard] NMORGENSTERN> I have seen regular articles on Forth . . . <[leonard] NMORGENSTERN> in Computer Shopper by Glenn Davis. Do you know him? <[leonard] NMORGENSTERN> . <[Gary] GARY-S> not personally <[Gary] GARY-S> I have had a few exchanges with him.. <[Gary] GARY-S> when he was doing wind tests in the desert <[Gary] GARY-S> why ? <[Gary] GARY-S> ga <[Gary] GARY-S> ga leonard <[leonard] NMORGENSTERN> Where is he? He writes for the beginner <[leonard] NMORGENSTERN> . <[Gary] GARY-S> I really don't know now leonard <[Gary] GARY-S> He was doing wind tests on antennas using forth. is here. <[Gary] GARY-S> That's a long time ago though. <[leonard] NMORGENSTERN> Hi hey db <[Gary] GARY-S> ga jax just saying hi to a new login id! who DBARKER1? ga <[Gary] GARY-S> no prob - hi d.b I'm from spokane, wa Yeah! Apples! <[Gary] GARY-S> into forth ? <[Gary] GARY-S> ga db cold weather! <[leonard] NMORGENSTERN> I was there for the Fair! yes, I use it at work what is work? yours i mean ... i know what work is! :-) I wrk at a place called Itron; we make hand held meter reading competers what chip? compoters that is <[Gary] GARY-S> hold it gang - hit three ... between thoughts <[Gary] GARY-S> ga db RCA 1806 - yes, its archaic what is forth like on an RCA 1806? . <[Gary] GARY-S> ga jax that was it! just that ? . its poly forth, and its mainly <[Gary] GARY-S> db are you aware of conf protocol ? <[Gary] GARY-S> ga db the only forth I've used, accept for a little a home. Protocol? to get in a word ... .... to continue a line <[Gary] GARY-S> yes - type ... at the end of a line of... and . to finish . <[Gary] GARY-S> incomplete thoughts and end with... <[Gary] GARY-S> a single dot or a ga. <[Gary] GARY-S> ga db sorry about that ! . <[Gary] GARY-S> no prob - you're up jax <[Gary] GARY-S> ga jax We make it up as we go! . I like that . <[Gary] GARY-S> how are you loading the forth onto the 1806 ? <[Gary] GARY-S> ga db we compile on an IBM pc... we download the code through another communicaton device. <[Gary] GARY-S> are you meta compiling the app ? . <[Gary] GARY-S> ga db I guess you could say that, if you mean using Create, DOES> . <[Gary] GARY-S> ga jax Recently jax has scoped ... on the Zen of PolyFORTH ... their idea ... is to sell FORTH to groups of , say, engineers ... who are primarily not what the average microcomputer hack ... would call a programmer ... they are engineers ... proficient at solving problems with the tools at hand ... and their attitude is " If the boss bought FORTH ... I damn well better learn to use it!" PolyFORTH ... is somewhat targeted at this crowd ... and FORTH Inc does not try ... to sheep-dip them in FORTH ... just give them enough to get it up & running on the application this is kinda ... what we do at VESTA ... whose coporate account i did not use tonite ... since the boss was supposed to show up here, but didn't! we have the same customers ... people who think FORTH is interesting ... but are much more interested in it as a tool ... to get things DONE EASY & QUICK ... which is why us FORTHiacs still have jobs:-)! . <[Gary] GARY-S> that and huge dams in Arab countries <[leonard] NMORGENSTERN> is here. <[Gary] GARY-S> ga len <[leonard] NMORGENSTERN> Alan Furman should see this. He has, ... <[leonard] NMORGENSTERN> as you know, the idea that the way to promote Forth . . . <[Gary] GARY-S> wait rs <[leonard] NMORGENSTERN> is to get it into the universities. . . <[leonard] NMORGENSTERN> Maybe the enineers are better ! <[leonard] NMORGENSTERN> . <[Gary] GARY-S> hi rs and <[Gary] GARY-S> ga rs <[leonard] NMORGENSTERN> Hi R.Sullivan! HELLO <[Gary] GARY-S> you kept keying rs - any comment ? <[Gary] GARY-S> ga rs <[Gary] GARY-S> ga jax <[Gary] GARY-S> ga jax Len ,,, what allan suggest... is a good idea ... since in my field, at least ... there is a shortage of qualified FORTH programmers ... and most of us DARN TOOTIN didn't learn it in college ... so if Allan pulls ... while you & gary & i push from behind ... we might just get the old nag out of the mud yet! gotta go! <[leonard] NMORGENSTERN> I agree! goodnite all & thanx . <[Gary] GARY-S> We all need to go - any last words Have a good night! . <[Gary] GARY-S> thanks guys this party is over FORTH PROGRAMMERS GET IT UP FASTER !!! <[leonard] NMORGENSTERN> Goodbye <[leonard] NMORGENSTERN> has left. === End of Steno notes. ===