Sub: Minutes of your RT Conference. >From Steno: Jennifer Marlowe Date: 01/11/90 Time: 22:02EST Attendees: [[Bess] PRESS16] [[Dennis] D.RUFFER] [[Doug] D.PHILIPS2] [[Dave] OLORIN] [GARY-S] Minutes: ok, I'll be leader I guess <[Doug] D.PHILIPS2> Go for it. <[Hadil] H.SABBAGH> I am looking for Aladdin-like software for the Mac. <[Doug] D.PHILIPS2> Hows it going with FPC3.5 (if that isn't too painful to ask)? Hadil, Aladdin Inc, is working on a Mac version, but may be a little while before they get it done <[Hadil] H.SABBAGH> BRB is here. <[Doug] D.PHILIPS2> Administrativia: What does /BLANK,/XBLANK really do? <[Hadil] H.SABBAGH> Hi Dennis.This isn't so bad now. <[Dennis] D.RUFFER> uh, I think they add blank line <[Doug] D.PHILIPS2> Hi! <[Dennis] D.RUFFER> Welcome Elizabeth <[Bess] PRESS16> hi <[Hadil] H.SABBAGH> Hi. Elizabeth. You're talking next week, right? <[Bess] PRESS16> right (by the way, real name is Elizabeth Rather) <[Dennis] D.RUFFER> I posted her bio in the Bulletin Board, if your interested <[Hadil] H.SABBAGH> M400 <[Doug] D.PHILIPS2> Elizabeth, when you upload the 20page highlights <[Hadil] H.SABBAGH> M605 <[Doug] D.PHILIPS2> could you also include info on how to get a copy <[Bess] PRESS16> ask dennis <[Hadil] H.SABBAGH> Oops. Sorry about that... <[Doug] D.PHILIPS2> of the BASIS. I ported your message that contained its cost over to comp.lang.forth, but your message didn't say how to order the basis. ga <[Dennis] D.RUFFER> it's ready to go, and I will upload it tonight <[Doug] D.PHILIPS2> Great! <[Bess] PRESS16> send chk for $10 to Forth Vendors Group c/o me at FORTH, Inc... 111 N. sepulveda blvd, manhattan beach CA, 90266 <[Dennis] D.RUFFER> I will also mention the address, <[Doug] D.PHILIPS2> Thanks Dennis. <[Dennis] D.RUFFER> do you need it in a BB message also so it gets ported? <[Doug] D.PHILIPS2> H (Spaz!) No, i can down load it and make it into a message. Everyday when I check the BB for new stuff, I also check the library for new stuff, so if its appropriately labeled, I'll grab it. ga <[Dennis] D.RUFFER> ok, I'll append the address to the summary what we going to talk about tonight, or should I just make something up? is here. <[Bess] PRESS16> what are you going to want to talk about next week? I can study hi gary <[Doug] D.PHILIPS2> Hi Gary Has anyone played with Pygmy 1.2 yet? <[Hadil] H.SABBAGH> Hi. Gary (Might as well...) <[Bess] PRESS16> what is it? <[Dennis] D.RUFFER> Pygmy is a CMForth for the PC written by Frank Sargent <[Hadil] H.SABBAGH> It's Frank Sargent's PD Forth for PCs <[Doug] D.PHILIPS2> Thanks, it was on the tip of my tongue. <[Hadil] H.SABBAGH> Just got MacForth+. Extremely cool. BIG PHONE line probs - sorry folks <[Dennis] D.RUFFER> yes, latest update works on all monitors (haven't heard any complaints :-) <[Doug] D.PHILIPS2> Has anyone modified a PC forth to work when... <[Hadil] H.SABBAGH> Still waiting for the standard... <[Doug] D.PHILIPS2> the PC is being accessed via COM1 (modem)??? <[Hadil] H.SABBAGH> So I can write HadilForth :-) <[Doug] D.PHILIPS2> [This seems like its going to be my 2nd project.] <[Dennis] D.RUFFER> the PCMODEM stuff that I have posted can be set up as a remote terminal if that is what you are talking about <[Doug] D.PHILIPS2> Well, what I want to do is modify pygmy so that it <[Bess] PRESS16> We can use modem, COM1 or other serial as extra users on a PC... <[Hadil] H.SABBAGH> Don't know anything about Pygmy... <[Doug] D.PHILIPS2> uses 'standard input and standard output' <[Bess] PRESS16> We teach classes w/ 16 users on serial muxes <[Doug] D.PHILIPS2> instead of accessing the console/keyboard all the time. <[Hadil] H.SABBAGH> But you'll have to vector KEY and EMIT to use standard io <[Doug] D.PHILIPS2> ga That's what I suspected. <[Dennis] D.RUFFER> no, you can't go into DOS, but all of Forth is available <[Hadil] H.SABBAGH> Blox too, no doubt (Sigh......) <[Doug] D.PHILIPS2> It looks like project 2a is going to be adding streams to pygmy <[Dennis] D.RUFFER> you can set up DOS to do it with the CTTY command in DOS <[Doug] D.PHILIPS2> then project 2b can do revectoring of KEY/EMIT <[Hadil] H.SABBAGH> Streams shouldn't be too hard via INT 21H. NOW the Unix/Forth group is caloring for an emacs type editor <[Dennis] D.RUFFER> for streams, check out Martin Tracy's EVAL code it does it using the existing BLOCK interface <[Doug] D.PHILIPS2> Where do I find this 'EVAL code? <[Hadil] H.SABBAGH> Gary: Yah. Not a bad idea, for Forth. TOO BIG, Hadil <[Dennis] D.RUFFER> in the library, I think it might be called STREAMS.ARC <[Hadil] H.SABBAGH> What do you mean? <[Doug] D.PHILIPS2> I'll poke about there abit, anything else I should look for? Gary: Naw, emacs isn't that big! Or, it doesn't have to be, anyway. <[Dennis] D.RUFFER> depends on what you want Doug :-) The source is 100K ! I've written 6k Forth kernels ! <[Doug] D.PHILIPS2> Well, examples of forth code that interface with DOS files would be useful! <[Hadil] H.SABBAGH> The emacs KERNEL isn't that big... All the source is written in LISP; kernel... <[Dennis] D.RUFFER> search for FILE keyword <[Hadil] H.SABBAGH> Is the LISP interpreter. <[Doug] D.PHILIPS2> And there are only about 10billion variants of emacs to consider Dennis: Ok. <[Bess] PRESS16> The FILES wordset in BASIS should be a good beginning. Chk it out. <[Doug] D.PHILIPS2> Will have to order BASIS first! ;-) <[Bess] PRESS16> :-) <[Dennis] D.RUFFER> we don't have a public domain example of it yet <[Hadil] H.SABBAGH> How many incarnations of BASIS after this? <[Bess] PRESS16> God knows <[Doug] D.PHILIPS2> Just keep revising that document till everyone is too tired to care anymore :-) <[Bess] PRESS16> we hoped we'd be done by now. <[Hadil] H.SABBAGH> Well, I'll wait, I guess. <[Dennis] D.RUFFER> is there a official cut off point? <[Doug] D.PHILIPS2> It's amusing to read back issues of F.D. and see 'done' dates. <[Bess] PRESS16> no. exhaustion by participant's? <[Hadil] H.SABBAGH> Dennis: One cutoff point is when... EVERYBODY's code is broken :-) <[Doug] D.PHILIPS2> Its really tempting to ask stuff that should wait till next week, esp. since I'll have to miss then :-( <[Dennis] D.RUFFER> tonight's notes will get posted also <[Bess] PRESS16> oh, fire away. I need the practice <[Doug] D.PHILIPS2> Well, I'm curious as to how it was decided that it was appropriate to consider an ANSI std. for Forth? <[Dennis] D.RUFFER> are we still there Bess is having technical difficulties brb <[Bess] PRESS16> people havepeople have been talking bout it since 84. in 86 we said why wait <[Hadil] H.SABBAGH> Well, gotta go. G'nite all. has left. <[Bess] PRESS16> it seemed as though things were stabilazing around forth83 but it had problems that needed fixing particularly cell size dependency, etc. ga <[Doug] D.PHILIPS2> AhAh. And how many/who approached ANSI? I'e not gotten far enough back thru the back issues of F.D. to catch the beginning. Is it documented there? ga <[Bess] PRESS16> about 12 people incl. LMI, CSI, IBM, MCI, Chuck, Greg B., Martin T., others... met at FORTH, inc in oct 6 and decided to submit a proposal to ASI. ga <[Doug] D.PHILIPS2> Ok. Is it just coincidence that the 'C' standard was happening about the same time? ga <[Dennis] D.RUFFER> some of it is in 4D, ther comments are here in the standards library, and still more in Rochester proceedings <[Bess] PRESS16> no, C had been meeting 5 yrs already. <[Dennis] D.RUFFER> Is the C standard done yet? <[Bess] PRESS16> yes, just got the final notice today. lord yes <[Dennis] D.RUFFER> 5 years huh <[Doug] D.PHILIPS2> Its almA long and winding road... <[Bess] PRESS16> Took them 7 by the time they were done, and when they started... People joke about Forth-99 , maybe it ain't all that funny <[Doug] D.PHILIPS2> Prophetic perhaps. <[Bess] PRESS16> they said they were just going to republish K&R. <[Doug] D.PHILIPS2> You don't know the POWER of the Dark Side. Do you have any feel for the difference in size between the forth and C committees, in terms of raw numbers of participants? IF they had simply removed K&R ambiguities... <[Bess] PRESS16> RE dark side--FORTRAN 8x is out for public review,... but it didn't work that way <[Bess] PRESS16> and there have been over 1,000 pages of comment, all of which... the poor committee has to respond to. <[Doug] D.PHILIPS2> And have any lessons been learned from the C effort that might save yens some grief? not as loong as ANSI is the affiliation is here. <[Bess] PRESS16> yes, try to keep it simple. So many folks want us to invent... the perfct language. That will take 100 yrs. <[Dennis] D.RUFFER> hi dave <[Dave] OLORIN> Hullo <[Dennis] D.RUFFER> and it probably wouldn't be Forth by then <[Doug] D.PHILIPS2> Yes, X3J11 got burned on the Trigraph stuff. I assume that X3J14 isn't <[Bess] PRESS16> Sorry, have to go. Been fun. CU next week <[Doug] D.PHILIPS2> going to touch something like that at all! THink how long we've been reinventing English :-} <[Bess] PRESS16> has left. X3/J14 has some hot spots that will keep the pot boiling just as vigorously <[Dennis] D.RUFFER> and a bunch of antagonists too <[Doug] D.PHILIPS2> Has a position been taken about levels of standard. (The term I want is eluding me) <[Dennis] D.RUFFER> Yes, there is statments to that effect in the basis <[Dave] OLORIN> What is the status of an ''electronic BASIS''? zip <[Doug] D.PHILIPS2> layers peX3J11 decided to avoid that for various reasons... did X3J14 consider those, or should I wait until I can read the BASIS? <[Dennis] D.RUFFER> under discussion even now <[Doug] D.PHILIPS2> [I was under the impression that ANSI was firm about electronic publication, is that true? ] <[Dennis] D.RUFFER> there is firm and then there is firm, I can't say much more right yet <[Dave] OLORIN> (suddenly a deathly silence falls over the room) <[Dennis] D.RUFFER> I can tell you that editing the basis summary was a real pain <[Doug] D.PHILIPS2> Well, has anyone heard from Wil after last sundays foray? <[Dennis] D.RUFFER> no, did he go away mad? <[Doug] D.PHILIPS2> You mean deciding what to highlight, or the process itself? <[Dennis] D.RUFFER> I mean getting rid of the highlighting <[Doug] D.PHILIPS2> No, just tired. I get the impression he's not a fast typist. <[Dennis] D.RUFFER> they use many special fonts and other word processing tricks that just do not translate to ASCII <[Doug] D.PHILIPS2> You mean you didn't write a little forth programt o do it? :-) :-) <[Dennis] D.RUFFER> what I ended up with was not near as pretty, but it is pure ASCII no streams in PolyForth, remember <[Doug] D.PHILIPS2> Ga! Bit Bangers! I want to write unix-like stream tools! <[Dennis] D.RUFFER> I have word processors that do a much better job than I could ever do with a program <[Doug] D.PHILIPS2> ([I don't want to duplicate the programs, just the style.] oooo- doug , let's tell Peter to post his code ! <[Dave] OLORIN> (be back soon) <[Dave] OLORIN> has left. <[Doug] D.PHILIPS2> Who is Peter? de Silva the wolf <[Doug] D.PHILIPS2> I asked him to post the 'vi' code, but got no answer. Or do you mean something else? that's it ! <[Doug] D.PHILIPS2> ga We'll see. He doesn't like the GEnie influx, so I don't know how well The forth editor with vi syntax <[Doug] D.PHILIPS2> he likes me. <[Dennis] D.RUFFER> the GEnie flux debate should have been appeased by the seperate cat We are starting to get some good two-way traffic on both ends <[Doug] D.PHILIPS2> (NO, GEnie -> comp.lang.forth). <[Dennis] D.RUFFER> are you posting the whole thing Doug? <[Doug] D.PHILIPS2> Finally! I donGod No! I had sort out the useful stuff J Wavrik replied to you and F.Sergeant replied to the SQRT query with a header <[Doug] D.PHILIPS2> and dump the rest on the floor. <[Dennis] D.RUFFER> actually, only the usenet cat really should be ported now that we've done that that was the intention Dennis - Doug is porting to Usenet - the cat 18 doesn't help him <[Doug] D.PHILIPS2> I cross posted the bio on Elizabeth because I thought it was an interesting bit of history. is here. <[Doug] D.PHILIPS2> what about Wavrik? Wavrik responede on Usenet to ForthNet points <[Dennis] D.RUFFER> I thought that was one of the reasons for cat 18, was to cut down on the noise going out to usenet <[Doug] D.PHILIPS2> Rotate 180 degrees. I left e-mail on ECFB for jerry to seperate it there, too - sure hope he does <[Dennis] D.RUFFER> hum, must have been my mistaken impression <[Doug] D.PHILIPS2> Query: Are linked lists something that down much in Forth, or are they a sign that you're not thinking 'in forth' yet? 'that down' -> 'that are done' not very forthish <[Dennis] D.RUFFER> the dictionary is just one big linked list forth is a linked dictionary echo - echo - echo <[Dennis] D.RUFFER> but it only has thread in one direction <[Doug] D.PHILIPS2> But for user data structures? <[Dennis] D.RUFFER> I've used them when they are appropriate <[Doug] D.PHILIPS2> If you wanted to keep around the equivalent of a 'set', is there a trick commonly used in forth? <[Dennis] D.RUFFER> the capability exists in PolyForth's DBMS also a set of what <[Doug] D.PHILIPS2> (A set of integers) <[Dave] OLORIN> I've got them in my bag of tricks. I use them when I need them. In your pile of Forth Dimensions - read the INteger Method <[Dennis] D.RUFFER> put them into a structure (ie CREATE) and let the dictionary keep track of them <[Doug] D.PHILIPS2> Dennis, I'm not quite following you yet. Gary: Any idea which volume, or...even... issue? (volume would help ) <[Dennis] D.RUFFER> CREATE INTEGERS 1 C, 2 C, 3 C, 4 C, etc Doug - I think I see your direction, check out the Integer Method and Quans BRB <[Dennis] D.RUFFER> I don't think I'm following what you want to do yet Doug <[Doug] D.PHILIPS2> BRB? (I dunno know what the means) <[Dennis] D.RUFFER> Be Right Back not me Gary <[Doug] D.PHILIPS2> Dennis: I want to keep several 'sets' of integers, be able to add and delete individual items on the fly. In 'C', I use a linked structure of 'min,max' ranges to keep the set ga <[Dennis] D.RUFFER> I guess I'm missing the purpose do you want to ssociate them together? <[Doug] D.PHILIPS2> AhA! The purpose! <[Dave] OLORIN> Doug: For a set of integers,min,max can become inefficient. What is the range in integer values? Dennis: I understand what he is talking about. <[Doug] D.PHILIPS2> 16 bits <[Dave] OLORIN> Doug: What about a bit-field for the sets? <[Dou] D.PHILIPS2> But negatives not allowed. <[Dave] OLORIN> Ouch. You are using the full range? <[Doug] D.PHILIPS2> No! What I'm using them for is to keep track of <[Dave] OLORIN> What is the range of the integers which you are using? <[Doug] D.PHILIPS2> which netnews articles I've read, out of the entire available set of articles. Since I expire news after about 4 days, <[Dave] OLORIN> Ahhhh. A Forth netnews reader????? <[Doug] D.PHILIPS2> YES! <[Dennis] D.RUFFER> sounds like a bit field to me <[Doug] D.PHILIPS2> I haven't gotten much over 1,000-2,000. <[Dennis] D.RUFFER> what is the range of values? Sorry - I can't find it <[Dave] OLORIN> Ok. Let's assume a maximum value of 4000, use bitfields. <[Doug] D.PHILIPS2> S'Ok. Just 2 more years of purgatory to do. I don't want to assume any maximum. I want to be able <[Dennis] D.RUFFER> yes, simply turn on bits for the ones you have read <[Doug] D.PHILIPS2> to expire news and not have to reset the article counts. <[Dave] OLORIN> What you want to do is create a structure which has 4000 bits of free space in it. (Probably 4096 because it is a power of two). <[Doug] D.PHILIPS2> Which is a *big* pain. <[Dave] OLORIN> Then you initialize all bits to zero, and set those bits which you have read. <[Doug] D.PHILIPS2> I know how to use bit fields, but I don't think that that approach I posted a bit mask almost in the beginnings of the library <[Doug] D.PHILIPS2> is necessarily extensible. <[Dennis] D.RUFFER> it is not, but now I see what you need I just don't know an easy way to do it <[Dave] OLORIN> Now. It isn't extensible. It has a fixed limit. <[Doug] D.PHILIPS2> There is more that I do. I also scan each group for unread articles, and I make up a set for article numbers that all have the same subject (subject threads). I also need to count how many items are in my sets, so I can display how many articles are unread about 'ForthNet: The final chapter' <[Dave] OLORIN> Ok. Multiple sets isn't much of a problem. <[Dennis] D.RUFFER> you do need to create some linked lists <[Doug] D.PHILIPS2> That's what I thought. I don't know that I want to use 'ALLOC' and tie up all that dictionary space, <[Dave] OLORIN> To handle counts, you probably want to just have a count field. (Makes doing a size order(1) rather than order(n) ) <[Dennis] D.RUFFER> VARIABLE TOP CREATE MEMBER TOP @ HERE TOP ! , <[Doug] D.PHILIPS2> And the linked lists I use now can collapsed as the holes get filled in. Free space can get messy. And when I'm done with each newsgroup, I reset all the subject thread pointers and start over for the next group. <[Dennis] D.RUFFER> is the data in a database? <[Doug] D.PHILIPS2> The only threads that don't get freed are those that keep track of the articles actually read. <[Dave] OLORIN> Doug: You need a memory manager other than the "cheap screw" heap manager that Forth uses. There is (I believe) one coming out in thenext issue of _The Computer Journal_. <[Doug] D.PHILIPS2> The data is in text files. I have an index file which gives me 'lseek' offsets to the start of the articles and also contains the seek offset to the subject line and the length of the article in lines. I'm probably going to redo some of that when I move to the <[Dennis] D.RUFFER> You can do what you want by creating the structure on disk, but you need some dbm tools to do it. I don't know of any in the pd <[Doug] D.PHILIPS2> If I can have a seekable stream package (I know, I know, roll your own), I can do that. The other thing I might do is grab another DOS segment to put my lists into. Of course, I'm open to alternative representations that achieve the same effect. ga <[Dennis] D.RUFFER> streams won't buy you much since the database would probably be fixed length indices <[Doug] D.PHILIPS2> Streams by me seeking around in the file that has the text of the news in it. 'by'->'buy ' <[Dennis] D.RUFFER> Yes, streams will help there <[Doug] D.PHILIPS2> So, is this off the wall to be wanting to do in forth? Have I taken an approach to the problem which precludes an easy forth solution? ga <[Dennis] D.RUFFER> I can do what you want in PolyForth, but the tool is not pd <[Dave] OLORIN> Not at all. Forth will work fine. And if you are using F- PC, it already has a full stream library. <[Doug] D.PHILIPS2> I'm not using F-PC at the moment, but I do have it. I'm using pygmy, <[Dave] OLORIN> (And I have placed a minimal stream package for Druma Forth in the library) <[Dennis] D.RUFFER> has anyone done a database in it yet? in F-PC that is <[Dave] OLORIN> Not that I know of. <[Doug] D.PHILIPS2> since I'm a bit overwhelmed by F-PC. Of course, I might try to steal the stream package from F-PC. <[Dave] OLORIN> F-PC is a very nice package. <[Dennis] D.RUFFER> you might also check out ABUNDANCE it is a VERY nice database system in Forth <[Doug] D.PHILIPS2> Abudance is in the library here? <[Dennis] D.RUFFER> yes an older version that is <[Doug] D.PHILIPS2> I don't think I want a full blown database system. I want to keep the news articles in plain text files so that I can edit them with various tools, and let my uucp (UULINK) software deal with them. I just want to create these auxillary indecies into them. <[Dennis] D.RUFFER> you need to talk with Bob Schemmel he has the same desires for a modem package <[Dave] OLORIN> Doug: F-PC will do what you want. And it isn't nearly as frightening as it seems. (BTW: The hypertext help in 3.5 is nice, too) <[Doug] D.PHILIPS2> Hypertext SchmyperText! Hypertext is too often an excuse for not doing the right organization! (Personal opinion, of course :-) <[Dennis] D.RUFFER> Bob Schemmel (R.SCHEMMEL1) has been doing a bunch of thinking about the same kind of things <[Doug] D.PHILIPS2> Actually, Hypertext is good for REFERENCE manuals, but too often a hyperlinked Can we continue this later ? <[Doug] D.PHILIPS2> reference document is considered good enough. <[Dennis] D.RUFFER> well folks, it's getting late and I haven't eaten supper yet <[Dave] OLORIN> I need to head out. Bye. has left. <[Dennis] D.RUFFER> can I go home yet? <[Doug] D.PHILIPS2> Ja. Next week is taken, how about sunday night? let's kill the cat ********** poof has left. <[Dennis] D.RUFFER> just be there and bring it up Doug take the wizards by the horns :-) <[Doug] D.PHILIPS2> Que Pasa? <[Dennis] D.RUFFER> bring up your need next Sunday, and I'm sure it will get some discussion for now, I'm calling it a night <[Doug] D.PHILIPS2> Ciao. has left. === End of Steno notes. ===