Path: utzoo!utgpu!jarvis.csri.toronto.edu!cs.utexas.edu!samsung!uakari.primate.wisc.edu!ames!ncar!ico!rcd From: rcd@ico.isc.com (Dick Dunn) Newsgroups: comp.lang.postscript Subject: Re: PostScript Language Summary: trying for a reply with some substance this time Message-ID: <1990Mar2.030024.1911@ico.isc.com> Date: 2 Mar 90 03:00:24 GMT References: <9447@imagen.UUCP> Organization: Interactive Systems Corporation, Boulder, CO Lines: 136 In article <9447@imagen.UUCP>, ib@apolling (Ivan N. Bach) writes: > >UltraScript is not an Adobe product; it is a clone PostScript interpreter > >made by Imagen/QMS. ... > I am getting tired of reading about PostScript "clones." The word "clone" > had never been used to describe the implementation of a procedural language > until Adobe started using it to refer to the implementations of the PostScript > language by other companies. Have you ever heard of a COBOL or FORTRAN clone? Without judging whether "clone" is really appropriate, let's consider why it is used here. "Clone" is used to refer to a (presumably) workalike product which is based on an existing product. Thus we speak of an "AT clone" when we're talking about "a personal computer which is intended to be a workalike to an IBM PC/AT"--where the "IBM PC/AT" part is a trademark. We don't speak of FORTRAN or COBOL clones because they are not trademarks. Cutting it a bit finer, we don't speak of Ada clones even though Ada is a trademark, because the trademark isn't based in a product. PostScript is a trademark, and there's a specific set of products behind it, and the use of the name is licensed to the creator of the original product. Certainly there are clones of other software products, and they're called "clones." It shouldn't be taken as pejorative; there certainly exist products which have been bettered by their "clones." > Have you ever heard of a proposal for a new world standard for the storing > and transmitting of image data that wastes up to 90% of the communication > channel's capacity?... No I haven't. Let's dodge the innuendo and get to the point: Is this directed at PostScript? I assume so since it was posted to this newsgroup. I'll assume that for the other innuendoes as well. But PostScript is not a "world standard for the storing and transmitting of image data." It is a page description language. Although it provides one mechanism for accepting transmitted image data, it does not preclude others. In fact, by its very programmable nature, it allows others. It says absolutely nothing about storing image data. For what PostScript is really intended to do, it might waste as much as a factor of two. Where do you get 90%? >...Have you ever heard of a modern procedural language > whose deficiencies are covered up by extending it with specific versions > of comment statements?... Wrong. If the language were extended by these comments, they wouldn't be comments, would they? The comment conventions provide information which is explicitly outside the scope of PostScript. And this IS something which has been done in other procedural languages. It is most commonly called a "pragma," although pragma covers more than that. >...Have you ever heard of a procedural language in > public domain whose most important component has been encrypted?... What part of the language is encrypted? Consider the answer carefully; I said "language," not "implementation." In fact, what is encrypted is the descriptions of the fonts, for the very simple reason that if Adobe hadn't agreed to encrypt them, they wouldn't have gotten the licenses from the typeface "manufacturers". Is this the issue? I admit I may be missing the point. >...Have > you ever heard of a procedural language in public domain whose name you > are not allowed to use when you implement that language? Yes, I have. If I go back as far as I can remember for the beginnings of such issues...I think the name you want is TRAC? Look for references to articles by/about Calvin Mooers, if I'm not mistaken...although that's a very old one, which is why I'm having trouble remembering it. Or look for work by Doug Ross way back in the early '60's. Look carefully at the terms under which "Simula" can be used. For something more modern, how about Eiffel--isn't that owned by Interactive Software (no relation to my employ- er, BTW)? If you're not hung up on procedural languages for the example, look at the various database languages with protected names. There are probably more language names which *are* trademarks than not. The designers of languages often try to protect the name from abuse until the language is established...because software products are still products and still carry issues like "brand identification." Finally, what do you mean by "public domain"? Is the definition of Post- Script in the public domain? I claim ignorance on the finer points there; I'm only trying to point out that "published" and "generally available" are not the same as "public domain." > ...If we leave it to Adobe to design our > page and document description languages, by year 2000 we will still have to > waste our time preprocessing arithmetic expressions from the usual format > supported by the first high-level procedural language into the Reverse > Polish notation in order to make it easier to interpret such expressions. But most PostScript is machine generated anyway...and normal machine trans- lation of arithmetic expressions goes through a transformation which is computationally equivalent to the transformation to postfix. Who's leaving it to Adobe, anyway? If you sit and watch, I'm sure Adobe will keep going, but what's holding you back? > The general-purpose programming part of our page description languages > will be at the level of a programmable calculator... (What's wrong with that, given where programmable calculators will be in 2000?:-) What is the PostScript language missing? >...and we will still have > to worry about doing our own gargabe collection, so that we do not run out > of virtual memory... How does this apply to PostScript? It doesn't. It is neither possible nor necessary to do garbage collection in a PostScript program, as far as I can tell. (If it were, you couldn't print a document of arbitrary length.) >...We will still have printer drivers that convert compact > image descriptions in binary format into very readable long words that are > never seen by a human being... This is confusing at best. If you're sending image descriptions to a PostScript printer, they're not terribly readable. If you don't like the level of encoding, re-encode it and let a PostScript program decode on the other end. That way you can choose whatever encoding fits. If you meant descriptions of text, line drawings, and such, and you're referring to the PostScript operators, the solution to getting rid of long names is trivial--bind short names to them (or more likely, to the specific code sequences you need which contain them). >...We will keep waiting for that readable data > to be sent over a 9600-baud communications link to a printer controller.. Maybe you will. I won't. I don't even now. My PostScript printer gets data a whole bunch faster than 9600 baud, since I hung it off the parallel port. Talk to QMS; get them to show you the PS-810...nice printer. -- Dick Dunn rcd@ico.isc.com uucp: {ncar,nbires}!ico!rcd (303)449-2870 ...Don't lend your hand to raise no flag atop no ship of fools.