Posted by Alan J. Flavell on 08/29/05 00:40

On Sun, 28 Aug 2005, jake wrote:

> In message <Pine.WNT.4.63.0508281109170.1992@ZORIN>, Alan J. Flavell
> <flavell@physics.gla.ac.uk> writes
> >
> > And it sure isn't going to work with w3m, emacs-w3, lynx. Nor google nor
> > any other search engines, either.
>
> Maybe it's just as well that it will work for the 90% of the Web population
> using IE ;-)

I don't know where you got your figures from, but the proportion of people
who regard MSIE as a web browser seems to be falling. (I might add that
MSIE deliberately flouts several mandatory requirements of the
interworking specifications and thereby disqualifies itself as a web
browser in the technical sense, but here we're probably discussing user
beliefs rather than theoretical principles.)

> > Which could be terribly misleading, if the web page
> > was relying on it, rather than just using it for optional decoration.
>
> The browser will simply fall back on the default font(s) if it doesn't support
> font embedding. So I'm not quite sure I see a problem here.

The problem is when authors use the technique to present some custom
glyph, rather than merely a cosmetic variation. I'm sorry if I hadn't
made that distinction clear.

> It's just presenting text in a preferred font;

If that's all that it does, and regular fonts are likely to cover the
needed character repertoire too, then you're OK, and my concerns are
ill-founded, indeed. I'm sorry if you felt we were arguing at cross-
purposes.

[Back to original message]