Path: utzoo!utgpu!jarvis.csri.toronto.edu!rutgers!usc!apple!sun-barr!newstop!texsun!csccat!larry From: larry@csccat.UUCP (Larry Spence) Newsgroups: comp.lang.postscript Subject: Re: Bitmap of PostScript code.. Message-ID: <3568@csccat.UUCP> Date: 22 Feb 90 21:56:54 GMT References: <1990Feb14.041704.14844@athena.mit.edu> <2761@bacchus.dec.com> <144@heaven.woodside.ca.us> <18001@rpp386.cactus.org> Reply-To: larry@csccat.UUCP (Larry Spence) Organization: Computer Support Corporation. Dallas,Texas Lines: 98 In article <18001@rpp386.cactus.org> woody@rpp386.cactus.org (Woodrow Baker) writes: > >While postscript is indeed a raster type language, and your points on that >are very well taken, and said, it should be realized that the core graphics >primatives are just as applicable to vector devices. Read Warnock & Wyatt's paper (1982 SIGGRAPH?) on the imaging model. They say something to the effect that "the imaging model becomes strained if extended to vector devices." One of the nice things about PS is its elegance, power, and portability. You would lose a lot of that elegance and portability if you started creating "vector subsets" for vinyl cutters, etc. You would end up with a scenario where you would have millions of dialects, just the problem you now have with bitmaps. >The bitmap is the "rosetta stone" of the graphics world. Postscript is >not. Look, PS isn't perfect, but it's one of the few truly useful and powerful standards that exist in graphics. I think you're complaining more about Adobe's implementation than the language itself. Just go buy someone else's implementation, like UltraScript or Go Script or Freedom of Press, they'll let you export bitmaps. If the other guys don't do as good a job, then Adobe deserves whatever spoils they can get -- if the other interpreters are good enough, there will be pressure on Adobe to loosen up. Simple free-market economics. >If I want to move an image between a sun, or a >mac, or a pc, the way to do it, is to convert bitmaps. Why do you say this? PS seems like a good solution to the portability problem. You have a graphic object on your Mac (displayed at 72 dpi) and you want to print it on your 300 dpi LaserJet, and you want to SCALE BITMAPS?!? I thought it was pretty much agreed that device-independence is a Good Thing. And PS is just another device-independent graphics format, if perhaps a very good one. If your app can't export a decent object-oriented format like EPS or CGM or something, that seems like the application's problem. >If I generate a page in Postscript, the bitmap should >be available to other entities. Postscript won't allow that. This appears >to me to be rather myoptic. Buy someone else's interpreter. There are plenty out there now. There may be tradeoffs involved, though. >The bottom line, is that bitmaps are VERY valuable things to have around, >they may be big, but hardly ugly. Scale a 300 dpi bitmap up to 2540 dpi and run it out to your Lino. Ugly. Look at the size of the 2540 dpi bitmap on your hard disk. Ugly. >The world does >NOT revolve around an abstract page description language, and certainly >never will, as it is not suitable for all things. Neither are bitmaps, but >they are a good deal more general purpose than the page description >language. But what about the vinyl cutters, etc., that you mentioned? They won't like those bitmaps one bit (ahem)! The abstract geometric description is a better lowest common denominator than a bitmap, IMHO. If you HAD to, you could write an interpreter for a "vector subset" of PS for a vinyl cutter, while you could not write a bitmap converter for the same device. The world does not have to revolve around PS, but it probably should revolve around SOME object-oriented description of graphics, IMHO. >I have a real problem with the attitude that "nothing else >matters...We set the standard", whether it is IBM (the microchannel) >or Microsoft (OS/2 ) or Alan Bradley (industrial controllers) or....... >you name it. If there's an unsatisfied demand for PS interpreters for vinyl cutters, you could make big bucks by filling it. Do it! You gain a LOT by excluding vector devices from the imaging model. They didn't do it just to piss you off, Woody! It's the usual scenario: you can't make progress if you have to maintain compatibility with EVERY prior technology, or at least not as much progress. PS was a step FORWARD, not BACKWARD. And hey, I work for a company that competes with Adobe in application software to some extent. Why do you assume that Adobe is the problem? Our application didn't want to depend on DPS and its resource and licensing requirements, so we implemented Bezier rendering, clipping, styled lines, etc., ourselves. It's not like if Adobe doesn't let you do something, there's no alternative. The spec is in the public domain, after all. Again, if apps have trouble doing the things that PS does, that means Adobe has earned their bucks fair and square. It would be a different story if PS itself were encrypted, and the spec was private. It's not. Type 1 fonts are the exception, but won't be as soon as Adobe gets around to releasing the spec, as they say they will. If I've gotten a bit strident, Woody, I apologize. It just seems to me like the whole intent of PS is to discourage expedient hacking of the sort you like so much. It's like adding little custom exit ramps to a streamlined (ahem) superhighway. Sure, it may help you with your particular problem, but it's not something that the designers of the superhighway want to encourage. And with good reason. -- Larry Spence larry@csccat ...{texbell,texsun,attctc}!csccat!larry