W3C home > Mailing lists > Public > www-style@w3.org > October 2010

Re: @font-face and slow downloading

From: Håkon Wium Lie <howcome@opera.com>
Date: Wed, 20 Oct 2010 16:46:23 +0200
Message-ID: <19647.191.92176.585268@gargle.gargle.HOWL>
To: Brad Kemper <brad.kemper@gmail.com>
Cc: Håkon Wium Lie <howcome@opera.com>, John Daggett <jdaggett@mozilla.com>, Beth Dakin <bdakin@apple.com>, www-style@w3.org
Also sprach Brad Kemper:

 > > There are cases where the webfont will not be downloaded at all. For
 > > example, on a mobile phone with narrow bandwidth, or if every kB costs
 > > you money, you may not want to use webfonts. Just like images can be
 > > turned off in certain browsers.
 > > 
 > > Therefore, I don't think the spec should try to define delays or other
 > > behavior which is dependent on the user's environment. 
 > Your conclusion seems like a non sequitur to me. If webfont
 > downloading is turned off, then there is no issue at all with
 > delays.

It's not necessarily a known, binary choice. There may be restrictions
like: don't spend more than 50kB on any web page. Which means that
webfonts will be downloaded sometimes. 

The more general point I'm trying to make is that the web is used in
many types of devices in many different circumstances. 2 seconds may
be a long time to wait sometimes for some users, other times it will
be very short. In Opera, we offer this as a user preference in

I believe the user is in a better position to set a value here than a
spec will be.

 > The issue is only relevant when the font is going to
 > download. I think it is reasonable for the UA to estimate how long
 > the download will take after, say 1-2 seconds, and if it is going
 > to be much longer than that then paint with a fallback font while
 > it waits.

In many cases, that seems reasonable. But it's still not something the
spec should have an opinion on.


              Håkon Wium Lie                          CTO °þe®ª
howcome@opera.com                  http://people.opera.com/howcome
Received on Wednesday, 20 October 2010 14:47:06 UTC

This archive was generated by hypermail 2.3.1 : Monday, 2 May 2016 14:38:39 UTC