- From: Todd Fahrner <fahrner@pobox.com>
- Date: Mon, 23 Jun 1997 10:04:54 -0700
- To: Chris Lilley <Chris.Lilley@sophia.inria.fr>
- Cc: www-style@w3.org
y'all wrote: > On Jun 14, 7:03pm, while I was out of email contact sorry, Todd Fahrner >wrote: > > > Typefaces and type sizes are now specified independently in CSS. More > > precisely, there is a mechanism to suggest alternate typefaces, but > > not to bind alternate type sizes and line-heights to the alternate > > typefaces. This is not especially useful, as typefaces have highly > > irregular metrics and suitability for screen display at a given size. > > So the best thing, then, would be to > > - describe their metrics (in an opsys-independent way) > - avoid pixel units, or implement the pixel definition of CSS1 > - make less assumptions about pixel density. You have just described Adobe Acrobat. I agree that it's a very sophisticated solution, but there's a lot of delicacy and overhead I thought CSS would avoid. PDF could probably make an outstanding output format for a DSSSL engine. > > Consider this case: > > > > font: 9pt/15pt Verdana, "Gill Sans," sans-serif; > > > > This will result in a very legible Verdana (unless it's a Mac), > > This looks like a crude hack designed to compensate for the different > descriptions of ascent and descent used by Win32 and MacOS platforms, > and perhaps also for the unfortunate built-in leading found in older > versions of some CSS-enabled browsers. I don't follow. I don't think the specification above has anything to do with the quality or completeness of the information available to the UA, but with the design of the typeface. It would be nice, however, to calculate suitable leading for any face based on some function of line-length, type size, and the ratio of ex- to em-height of the face. Somehow I'm pessimistic of UAs surfacing these figures for calculation when OS's are a few steps behind, and meanwhile MS and NS can't agree on the meaning of "margin." David Opstad, where are you now? > > This facility will become more important once embedded type comes > > into wider use - authors will sometimes want to suggest (and provide) > > highly idiosyncratic typefaces, whose size and leading are not > > suitable for second- or third-choice alternates. > > An arguable point, it depends on whether you see second and third > choices as being close alternatives or as completely separate designed. > The latter are perhaps better handled by using separately titled > stylesheets. Ah, so in addition to "print," "screen," "aural," and "projector" as media selectors for WD-Print, you could have @media: "Can't Stand the Sight of Monotype Perpetua Titling Caps" eh? But seriously, you should add "Handheld device" to the list, to give HDML advocates something to think about. The "deck of cards" metaphor is a variation on "stack of paper" if you ask me - lose "print" and go with "paged." __________________ Todd Fahrner mailto:fahrner@pobox.com http://www.verso.com/
Received on Monday, 23 June 1997 13:24:39 UTC