Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!sdd.hp.com!decwrl!ucbvax!CALSTATE.BITNET!JAJZ801 From: JAJZ801@CALSTATE.BITNET ("Jeff Sicherman,CSU Long Beach") Newsgroups: comp.lang.forth Subject: Off BASIS Message-ID: <9008271927.AA29643@ucbvax.Berkeley.EDU> Date: 27 Aug 90 08:38:44 GMT Sender: daemon@ucbvax.BERKELEY.EDU Reply-To: "Jeff Sicherman,CSU Long Beach" Organization: The Internet Lines: 77 It would seem, from the recent postings about the distribution policy for BASIS13 that the committee is trying to inhibit, not prevent, the wide dispersal of the proposed standard. It would probably be 'illegal' to prevent it, given their mandate, but making it slower, harder, and at some expense to get when an efficient medium is available is sort of a logistic stonewalling. Restricting it to MS Word format makes it minimally portable. Convertible standard formats, especially Postscript (tm) which I suggested before are available and presumably verifiable. I'm a relatively new Forth programmer so I don't have much investment either intellectual or financial in any particular set of choices or the effects of the standard, but I'm not encouraged much for the future of Forth by the standards process and progress, at least as I've viewed it through the discussions on the net and the relative silence of the committee itself (see above). This is based largely on supposition, but I can't escape the by the opinion that the committee's intent, by and large, was to protect their own interests and perspectives. The expense and time involved in the process itself restricts and discourages wide participation so that it will involve largely those with both heavy resources and investments (plus some fanatics). There's nothing inherently wrong with that: these people have the right and responsibility to protect those interests. But given the diversity of opinion in the Forth community (if the net is any indicator), there's a strong risk of narrowness of view. I'm not suggesting that things should have been put to popular vote on the net or any other medium, put I would have been encouraged to see solicitation from the committee for discussion fo various topics before and/or while they were the subject of strong debate within the process instead of dispersing decisions. There's an air of arrogance about that, a sort of attitude that 'We are the Forth community'. Given the issues raised over the result, I have to assume that the standard will be the object of much protest and discussion after it is published for comment by ANSI. If the issues are not resolvable, I fear it will never really be accepted (implemented) even though it may be adopted. And I don't know how they can be resolved given the failures and weaknesses of the process that has lead to it. I don't doubt that the standardization of other languages were exercises in politics also, but it seems from this distance to have dominated the Forth one. I feel this may be inherent to the nature of the language. It is too close to the hardware in practice, structure, and use for a sufficient level of abstraction to have been adopted early on and it appears to be much too late now to define a significant, robust virtual machine that will be adhered to. Too many Forth experts with too much provinciality and influence exist for the process to succeed. It's too much a hacker's (in a non-pejorative sense) language with a paucity of academic rigor (and consequently interest, you will note) in it's definition and implementation. This is not meant to belittle the quality of the products or their power - they seem quite good - it just means that everybody is likely to go their own way. It may be a misguided effort to try to standardize Forth in the first place; it's akin to trying to invoke get a single machine architecture, given the low level Forth operates at. That hasn't happened for much the same reasons: a multiplicty of ideas, interests, and investment in established implementations. Forth is more like a common machine model (e.g. the Von Nuemann) that everyone has built variations of. Maybe this is as it should be, not to cheapen the attempt too much, I just think it may (have been) doomed. I'm also not convinced the portability is all that necessary, given the nature of Forth applications and the likely potential for popularity. Based upon what I've seen and heard, I can't see large projects built with it. That doesn't mean there haven't been important and significant ones built - I mean REALLY big stuff that represents the big bucks and contracts in the industry. It's just too idiosyncratic to manage (literally) and too strange a language. I just can't see a big pool of programmers out there that will become proficient and comfortable with it. Let's face it: it's unnatural from an English language point-of-view and that's what's been burned into our neural nets in this country. (Does anybody know if the Japanese adapt to it any better ?). I know it's a trite joke, but it IS rather write-only and that's the principle inhibition to both learning the language and to code sharing and reading that are inherent to large projects these days. As a consequence, I think it is destined to be a niche language project- and programmer-wise. Maybe that's as it should be. Sorry, I don't think Forth would be much more popular no matter how standard and portable. Jeff Sicherman jajz801@calstate.bitnet