W3C home > Mailing lists > Public > www-style@w3.org > July 1997

Re: the *precise* definition of 1em

From: Douglas Rand <drand@sgi.com>
Date: Tue, 22 Jul 1997 10:18:56 -0400
Message-ID: <33D4C150.FBF67EF5@sgi.com>
To: "Joel N. Weber II" <devnull@gnu.ai.mit.edu>
CC: davidp@earthlink.net, www-style@w3.org
Joel N. Weber II wrote:
>...
> The CSS1 spec allows up to a 20% variation from what the user
> specified.
> However, it is possible for a UA to make this configurable.
> My understanding is that a UA can conform to the spec as long as it
> will never use a font which is more than a 20% difference from the
> user-specified size.  So if a UA usings 10% or 5% or 0%, that's fine.
>....
> In short, it's hard for me to tell you that you're going to reliably
> get any particular set of fonts on X11.

You're right that X11 is difficult,  but nearly every X11 server,  X
terminals not withstanding,  comes with a fairly standard set of fonts
including the basic Helvetica, Times, Courier, Lucida, Palatino, etc.
So I think you can do well most of the time.  Even the terminals could
use a font server to supply these.

What I do on SGI is to get all the information for a particular font
family,  then when it comes time to create a specific size font I first
try to find the font closest to the pixel size.  If that isn't within
1.5 pixels (the calculated size is a float,  not rounded) then I try for
a scalable font.  At least on our platform, the scaled fonts from the
Type1 renderer almost always look worse (for small sizes) than properly
optimized bitmaps, so it pays to try to use the bitmap font.

I wasn't aware that CSS specified the amount of allowable error.  I may
need to go back and change the pass/fail test for bitmap size matching.

Doug
-- 
Doug Rand				drand@sgi.com
Silicon Graphics/SSO			http://reality.sgi.com/drand
Disclaimer: These are my views,  SGI's views are in 3D
Received on Tuesday, 22 July 1997 10:19:36 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 27 April 2009 13:53:50 GMT