Path: utzoo!utgpu!water!watmath!clyde!att!osu-cis!tut.cis.ohio-state.edu!bloom-beacon!apple!casseres From: casseres@Apple.COM (David Casseres) Newsgroups: comp.sys.next Subject: Re: WYSIWYG Keywords: WYSIWYG Message-ID: <144@internal.Apple.COM> Date: 25 Oct 88 19:10:52 GMT References: <6937@orstcs.CS.ORST.EDU> <12908@oberon.USC.EDU> Organization: Apple Computer Inc, Cupertino, CA Lines: 66 In article <12908@oberon.USC.EDU> venkat@brand.usc.edu (V. Venkat) writes: >With a 17" diameter screen i.e. apprx 12" X 12" screen and a 1024X1024 pixels >-- the screen resolution will be about 86 pixels per inch. >Since the printing is essentially WYSIWYG a laser printer with 86 dpi >will suffice. What is the use of a 400 dpi laser printer unless the screen >represents only a part of the printed page and you have some scrolling >mechanism? Maybe I haven't understood the fundamentals here. Can someone >illuminate me further on this matter. Much thanks. Aha, we are into the wonderful wierd world of "just what DOES WYSIWYG really mean, anyway?" First, when Alan Kay coined the phrase it was not about visual fidelity between a computer screen and printed paper at all; it was about having a good visual representation (on screen) of your work, i.e. the contents of your head. At least that's how I understand it. But of course Kay's meaning has now disappeared in a fog of marketing hype. In its purest form the hype goes "what you see on the screen is EXACTLY what will be printed." Of course this is baloney. If it were true it would mean you just get a printed screen dump, at screen resolution, but nobody has ever been satisfied with that. A slightly more evolved form of the hype is "what gets printed is exactly what you saw on screen, only you couldn't see ALL of it on screen without scrolling, and the printed image is nicer because the printer has higher resolution." A very simple but bogus statement is struggling to turn into a rather complicated, somewhat less bogus statement. Myself, I like to avoid the term WYSIWYG when talking about printing, and use the term "visual fidelity." To me this means that the imaging and printing software will go through whatever contortions are necessary to render an ideal page image on the screen and on the printer, finding the best balance of A) quality of the printed image, B) quality of the screen image, and C) geometric and graphical correspondence between the two images. This is an extremely complicated and difficult problem to solve, EVEN when you get to use PostScript for both the screen and the printer. For example, users want to have 8-point text. Hell, they even want 6- point, and the printer prints 6- point beautifully. If you use the same letter designs on the screen (graphical fidelity), the tiny letters are illegible. If you change the letter designs for small sizes so as to make them legible, they don't lay out the same way any more, i.e., you find you have not only given up graphical fidelity but geometric fidelity as well. The string of letters may overlap something on paper that it didn't overlap on the screen, or a margin that was straight on the screen is ragged on paper, etc. For another example, PostScript is "resolution independent" and everything you draw can be arbitrarily scaled. That's a nice way to take care of differing resolutions on the screen and printer. But people want to in- corporate scanned images into their documents, and scanned images are bitmaps, and bitmaps are NOT resolution independent. When you scale them, ugly artifacts appear. Because of problems like these, achieving visual fidelity is as much an art as a science; a lot of judgement has to be used, and ultimately the solutions that emerge are the result of history and the interactions of system designers, application designers, and end users. The solutions on the Mac have always had a strong tendency to become printer-dependent, as third party developers sought to squeeze every bit of advantage out of each kind of printer; this has added to the general complexity of the problem. I'll be watching NeXT's development from a relatively clean slate with great fascination. David Casseres