- From: Todd Fahrner <fahrner@pobox.com>
- Date: Wed, 25 Feb 1998 18:47:57 -0800
- To: Bill Hill <billhill@microsoft.com>, "'Walter Ian Kaye'" <walter@natural-innovations.com>, www-font@w3.org
Bill Hill wrote (5:10 PM -0800 2/25/98): " Another memory I have of my time at Aldus is of launching Windows 3.1 for " the first time, and calling in my boss to show him how crisp and clear the " type was on screen. That was the point at which I felt Windows overtook the " Macintosh in type quality on screen. I'm sympathetic to a lot of what you're saying, but this is silly. Which fonts were you comparing? Rasterized into how many pixels? Which monitor? How clean was it, and was the CRT properly aligned with magnetic north? What does the OS have to do with this? Could it be that you prefer incorrect gamma? The extreme difficulty of TrueType hinting, combined with the superiority of Mac FONT and NFNT resources over Windows FONs means that many more typefaces are available for the Mac which display with nicely hand-tuned bitmaps, in all major variants (no slanty fake italics are necessary). I agree that bitmaps are a dead end, but until more well-hinted outlines are available, or universal anti-aliasing becomes the norm, Mac type will retain an inherent advantage on screen for a great many typefaces. " At Microsoft, we're going to take advantage of resolution wherever we can in " order to display type at its best for our customers. I find 9pt on the " screen just too small for me to read comfortably, even on Windows. On our " Microsoft Typography Website, we try to use sizes that will work on whatever " system they're viewed. We're aware of the issues. But until the Macintosh " resolution changes, small type is always going to look better on Windows " than the Macintosh at the same size because of the scaling problem you " highlight. This is obfuscatory. Mac displays are not 72 ppi, nor are Windows 96 ppi. At home I run MacOS at a little higher than 120 ppi, with mostly anti-aliased Type 1 on screen. Here at work my Mac runs closer to 108 ppi. 72- and 96-ppi are the resolutions that the respective operating systems assume when attempting to rasterize type "accurately." Since multisynch monitors have become commonplace, both OS's are likely to rasterize type inaccurately for a great majority of users. To their credit, some Wintel systems allow users to tune the scaling factor of points to pixels between boots. Regrettably, Macs lack this facility, which is why I find "14 point" a comfortable on-screen copy size in my tiny-pixel home setup. I'll wager that most Wintel users aren't aware of this scaling function at all, that very few have actually overridden the 96 ppi default, and that fewer still have done this in order to make type display on screen with literal physical accuracy (WYSIWYG). As long as printer res exceeds screen res, people outside of prepress will generally be more interested in magnifying physical units like points than in achieving WYSIWYG. What I'm getting at (in a roundabout way) is that it doesn't really matter how many pixels your OS thinks are in a point. The point is not to use physical units like points if your destination medium is a computer screen. CRTs are a projected medium like film - how many picas tall was Princess Leia's image on screen in the "Help me Obi Wan Kanobi" scene? It's a silly question that the director would never ask. Yet your website, and MS web authoring tools encourage people to use points in their CSS. Never mind that raster artwork generally remains in unscaled pixels, condemning the design to randomly ill proportions between word and image. You use points because - well - because they're easy, and that's what people know from word-processors and other print-oriented tools. And because Internet Explorer 3 can't handle any other unit correctly in CSS. Todd Fahrner mailto:fahrner@pobox.com http://www.verso.com/agitprop/ The printed page transcends space and time. The printed page, the infinitude of books, must be transcended. THE ELECTRO-LIBRARY. - El Lissitzky, 1923
Received on Wednesday, 25 February 1998 21:41:29 UTC