rat@circle.UUCP (David Douthitt) (09/02/87)
Is there any possibility of a ANSII Forth? Just curious....... -- ::: David Douthitt ::: Madison, Wisc ::: uucp mail: ...!uwvax!geowhiz!uwspan!hobbes!circle!rat fidonet mail: 121/1
shprentz@bdmrrr.bdm.com (Joel Shprentz) (09/07/87)
In article <12.213BFB0D@circle.UUCP>, rat@circle.UUCP (David Douthitt) writes: > Is there any possibility of a ANSII Forth? Here are some notes from the August ANSI Forth meeting that I downloaded from the East Coast Forth Board. -------------------------------- Notes on the 8/87 Meeting of the ANS Forth Technical Committee by Jerry Shifrin, SYSOP East Coast Forth Board, 703-442-8695 These are my notes on the first meeting of the ANS Forth Technical Committee (TC) held August 3-4, 1987 at CBEMA Headquarters in Washington, DC. These are not minutes of the meeting, nor are they part of the ANS Forth Technical Committee's official documentation. They represent my personal impressions only. The only reason I mention this is that during the course of the meeting, it became clear that there are serious liability considerations for anyone publishing anything which someone could interpret as anything approaching a standard. Therefore, I must caution everyone that this document IS NOT a standards document. With that out of the way, I'll try to give you a summary of the events of the meeting. The meeting began at 9 AM Monday and was attended by many of the leaders of the Forth Community, including a number of folks from the Forth Standard Team (FST). I missed a few names, but here's a nearly complete list of attendees, their affiliations, and some of their claims to fame: Don Colburn, Creative Solutions, Inc. (FST, MacForth) Ray Duncan, Laboratory Microsystems, Inc. (PC/Forth, UR/Forth) Charles H. Moore, Computer Cowboys (Forth inventor) Gerald A. Shifrin, MCI Telecommunications (ECFB) Martin Tracy, Forth Inc. (Forth Model Library, Masterforth, ZenForth, FIG, Mastering Forth co-author), alternate for Elizabeth Rather Gary Betts, Saba Technology Ted Dickens, The Dickens Co. Mike Nemeth, Computer Sciences Corp. (MD FIG, Goddard FUG) Elizabeth D. Rather, FORTH, Inc. (polyFORTH, FST) Robert Smith, Maxtor (Secretary of the FST) Bill Ragsdale, Dorado Systems (FIG, figForth) Ronald D. Braithwaite, The Tools Group Guy M. Kelly, (IEEE representative, Chairman of the FST) Lawrence P. Forsley, Laboratory for Laser Energetics (Rochester Forth Conference, FST) David C. Petty, Digitel (FST) James Rash, NASA GSFC Charlie Keane, PPI Chris Colburn, Creative Solutions, Inc., Alternate for Don Colburn Tom Kurihaca (sp?), Dept. of Transportation (X3) John Dohrband (sp?), NASA GSFC, Alternate for James Rash Richard Burton, National Bureau of Standards Fred Virtue (sp?), SPARC Bill Dress (Oak Ridge National Laboratory) and Gary Feierbach (Inner Access) had planned on attending but did not show up. Elizabeth Rather (as the "convenor") was acting chairperson, and Ray Duncan volunteered to be acting secretary for the initial meeting. Elizabeth made some opening remarks on the maturation of Forth and the need for greater acceptance. She described the scope of this effort as mainly oriented towards describing common practice, not attempting to fix Forth nor using the standard as an instrument for change. The agenda was approved without dissent. There followed a discussion on international involvement. The X3 representative suggested that early involvement would lengthen the process. It turns out that there is a requirement for an international liaison, but it wasn't clear at what point that would become important. Chuck Moore felt strongly that international involvement was important. I don't recall that any decision was reached on this. Following this was a discussion on validation suites. The original scope proposal specifically excluded them from this group's activity. The membership voted not to exclude the possibility of such development. Cathy Kachurik of CBEMA presented a tutorial on the X3 structure and process. Charlie Keane proposed adopting the FORTH-83 Standard as a "BASIS" document. As I understood it, the basis document becomes the working document for all activity of the TC. That is, all changes, deletions, additions, etc. are proposed as updates to this document. There was a discussion of whether to restrict this document to Chapter 12 (the Required Word Set). This was defeated 15-1. A motion to adopt chapters 1-16 was carried unanimously. This excluded only the appendices (uncontrolled reference words, experimental proposals, charter, membership, and proposal/comment forms). I put up a motion to include floating point as part of the scope of work. This was defeated by a vote of 4-12. A motion was approved (14-3) to eliminate the time frame constraints on standard language extensions. Previously, this had indicated that language extensions could not be considered until the approval of the ANS Forth standard. Two additions to the Scope of Work document were approved: that the TC will review existing and proposed programming language standards; and to consider the impact of the standard on current and anticipated hardware technology. Another change to this document was to change the number of users required for a Forth system in order to be considered for non-compliance with the Forth-83 standard; this number was reduced from 1000 to 200 users of a particular Forth implementation. The Scope of Work document (X3J14/87-002) was then approved as amended. The TC Subcommittees document (X3J14/87-004) was approved with minor changes. There are four subcommittees: Documentation, Research, Logistics, and Technical. The Plan of Work document (X3J14/87-003) was then approved with a few changes. It was agreed to remove specific topic areas from the meeting plan in order to allow the work to proceed faster if possible. Chuck Moore got a motion passed to require that at least one meeting be held in San Francisco. The membership then agreed to hold the next meeting in San Francisco, but could not find anyone willing to host that meeting. Elizabeth Rather then volunteered to host the meeting in Southern CA and that was agreed on. I put up a motion to co-schedule and co-locate the ANS Forth meetings with the FORML and Rochester conferences. This was defeated 4-10. Next was the Call for Officers. It turns out that the X3 Secretariat appoints the officers from a list of volunteers. The following people volunteered: Chair: Charlie Keane, Bill Dress, Larry Forsley Vice Chair: Bill Dress, Ray Duncan Secretary (appointed by the chair): Martin Tracy International Representative: Larry Forsley Vocabulary Representative (appointed by the chair): Ted Dickens Documentation Editor (appointed by the chair): Ron Braithwaite Applications for these positions may be accepted until August 31st. Each requires a letter of intent and qualification, along with a letter from your employer indicating that they understand the amount of time needed for taking on these assignments. Greg Bailey and Don Colburn then proposed that the technical subcommittee agree to mark up the standards document indicating areas to be deleted, modified, added, and areas of deviation from accepted practice. This was agreed 13-3. The group then agreed (9-4) to hold the next meeting on November 11-12, 1987 in Southern CA at FORTH, Inc. Elizabeth named acting chairs for each of the subcommittees: Documentation - Ted Dickens, Logistics - Gary Betts, Research - Guy Kelly, Technical - Greg Bailey. I passed out documentation and gave a brief description of the ANSForth bulletin board on MCI Mail. There was then a review of all action items and the TC adjourned. This was immediately followed by the convening of the Technical Subcommittee (TSC). Martin Tracy volunteered to serve as acting secretary of the TSC. There was a lengthy discussion on the proper name for this subcommittee (I'm using TSC in these notes, but that may not be accurate) and its voting membership requirements. At issue was whether it was a formal subcommittee which would carry additional documentation requirements. As I recall, no conclusion was reached on this. We did get the impression that in order to be a voting member of the TSC, you had to be a voting member of the TC. The TSC then drew together a list of goals: identify a kernel of highly compatible words, decide a strategy for layering and extensions, amass information on the TC desires and needs, and define a mechanism for handling proposals. I ran out of steam around this point and stopped taking notes, but most of the remaining discussion was on the proposal process, voting membership, and plans for the next meeting. Miscellaneous notes ------------------- X3 has raised its membership fees: $200 for one principal membership (includes one alternate), $150 for observer and each additional alternate. The X3 Secretariat may be reached at: CBEMA, 311 First St., N.W., Suite 500, Washington DC 20001, 202-737-8888. To register for MCI Mail (and gain access to the ANSForth bulletin board), call 800-MCI-2255. In the Washington, DC area, call 833-8484. The registration fee is $18. Commentary ---------- While the preceding describes events to the best of my recollection and note-taking abilities, I thought I'd add a few opinions and observations of my own: First, I think this effort is off to a great start. The membership includes an excellent and reasonably well-balanced group of vendors, users, and other interested folks. I believe most of the early objections to this effort have been resolved by the make-up of the TC. Additionally, there was a clearly cooperative spirit among the attendees. It was very clear to me that the TC was determined to pursue an open organization. Several discussions were concerned with how to publicize our procedures and encourage participation. This, along with CBEMA's requirements for "due process" will, I think, result in an excellent document. Unfortunately, it seems that there is still an IEEE cloud hanging over this effort. I thought that a compromise had been reached, but apparently there are still a few people pursuing the IEEE Forth alternative. We'll have to wait and see what happens. [This cloud has since been lifted; IEEE dropped their Forth standard project.] Chuck Moore, in spite of his avowed opposition to a Forth standard, was extremely cooperative. My impression is that he was mainly concerned with having wide participation and not shutting off the possibility of new Forth development. He also mentioned that he was working on a new Forth compiler. Don Colburn seemed to feel that we could put out a draft document much earlier than planned and was surprised at the idea that there would be any difficulty in reaching a consensus. I felt that Elizabeth did an excellent job of chairing the meeting, but suppose she was wise in not volunteering for the permanent Chair. This way we avoid even the appearance of a FORTH Inc. dominated effort. I have a couple of concerns about the course of this project. Most difficult for me to reconcile is the notion of a standard documenting common usage among the major Forth implementations. In some cases this may cause a reversion of some language features back to the way they were before the 83 standard. For example, Forth Inc. never changed its definition of LEAVE to correspond with the 83 standard; i.e. it does not immediately leave the loop. One could therefore argue that the 83 standard LEAVE is not in common usage. Thus it seems to me that the ANS standard might either leave its effect undefined or else omit it entirely. Worse, I think, would be to revert its meaning back to the 79 standard. The other main concern I have is with the minimalist approach. I guess it's the only sensible way of getting this out in a reasonable amount of time, but I worry that most proposals will simply be put aside with the note that they're outside the documented scope of the ANS Forth effort. On the positive side, I think that this group has enough talent and dedication to complete a superb standard in a reasonable amount of time. I offer my personal thanks to everyone involved for providing two days of stimulating discussion. A special tip of my Forth beanie goes to Elizabeth Rather and Martin Tracy for doing the bulk of the work in pulling this activity together. Other notes of interest ----------------------- Don and Chris Colburn were kind enough to invite everyone over to their house Monday evening for pizza and pool (swimming, that is). It was very pleasant and provided the opportunity for people to get to know each other a bit better. Don took us down for a tour of his workshop; it looked like a Mac farm. Don demonstrated the Mac II running several animated graphics tasks under MacForth in separate windows. Very nice. I was very happy to have Martin Tracy and Guy Kelly stay over my house, but only regret the short time for available for Forth talk. Guy demonstrated his new, implementation-independent Forth editor. It seemed very powerful - it could work with screen files, or native blocks. In addition he provided three ways for moving stuff around - cut and paste, a line stack (push and pop a line at a time), and a "barrel" (push stuff into the barrel, and select from it in any order). I believe Guy will be offering this for sale ($25 if I remember correctly). He includes drivers for several Forths, including F83, MVP, Uniforth, and LMI. Martin, as usual, has numerous irons in the fire including his polyFORTH implementation for Digital Signal Processing (DSP) chips, his upcoming Dr. Dobbs article and column, along with work on Zen Forth, and his continuing involvement with the Forth Model Library. Martin mentioned that Wil Baden (of F83X fame) had started helping him on the Zen project. Martin stopped by the Potomac FIG meeting (on his way to the airport) and talked about several of his projects. -- Joel Shprentz Phone: (703) 848-7305 BDM Corporation Uucp: seismo!bdmrrr!shprentz 7915 Jones Branch Drive shprentz@bdmrrr.bdm.com McLean, Virginia 22102