- From: Octavio Alvarez Piza <alvarezp@alvarezp.ods.org>
- Date: Tue, 25 Apr 2006 08:51:49 -0700
- To: Philip TAYLOR <P.Taylor@Rhul.Ac.Uk>
- Cc: www-style@w3.org, Bert Bos <bert@w3.org>, Dave Raggett <dsr@w3.org>
On Tue, 25 Apr 2006 15:52:39 +0100
Philip TAYLOR <P.Taylor@Rhul.Ac.Uk> wrote:
>
>
>
> Dave Raggett wrote:
>
> [...]
>
> > font-family: url(TSCu_Comic.ttf), "TSCu_Comic", sans-serif;
>
> would this not be more likely to be
>
> > font-family: "TSCu_Comic", url(TSCu_Comic.ttf), sans-serif;
>
> with the semantics "First look for a resident copy of "TSCu_Comic",
> then try to download one if there is no resident copy, and finally
> fall back on the browser sans font if the download fails or is not
> supported ? I cannot see why one might want your ordering in
> 99.9% of cases.
Being CSS information about a web page, it should *never* tell a
browser how to behave: that is up to the standard. That should be
simply
font-family: "TSCu_Comic", url(TSCu_Comic.ttf), sans-serif;
and have a global definition of the fonts, say:
@font_url('TSCu_Comic', 'b', url(TSCu_Comic-b.ttf)); /* Where to find bold */
@font_url('TSCu_Comic', 'n', url(TSCu_Comic-n.ttf)); /* Where to find normal */
@font_url('TSCu_Comic', 'o', url(TSCu_Comic-o.ttf)); /* Where to find obliques */
(I prefer font servers, though)
Received on Tuesday, 25 April 2006 15:54:36 UTC