Re: CSS3 Web Fonts issue with ?block on downl oad?

On May 7, 2009, at 4:42 PM, Philip TAYLOR wrote:

> David Hyatt wrote:
>> I'm well aware that visual jumping could occur on poorly designed  
>> Web sites if the feature were to become wildly popular.
> I don't understand why you believe that this would occur (only or
> primarily) in "poorly designed Web sites" : I would have thought
> that exactly the opposite would be the case.  A well-designed web
> site (IMHO) adjusts its flow to accommodate widely varying font
> sizes so as to allow each viewer of the site to adjust the font
> size to best meet his or her needs; such a site would (I suggest)
> also adjust well to the replacement of a default font by a web
> font once the latter has loaded, but the reflow would be clearly
> visible.  A poorly designed web site, on the other hand, that
> uses fixed width containers, may well avoid the reflow problem,
> but at the expense of cropping and/or enforced horizontal scrolling
> when the font dimensions cause the container to overflow.

Maybe so.  This is all just speculation until people start using the  
feature enough to have real examples.  Really what I should have said  
was that the visual jumping in many cases would be very mild (no worse  
than what you might see during typical page loads from other elements  
forcing reflows) not that it would be completely non-existent.   I'm  
certainly not defending visual jumping.  It should be avoided whenever  
possible. :)

At any rate the original question seemed to be more about whether you  
show the glyphs of the fallback font or not before you load the real  
font.  I still think it's better visually to not show the wrong  
glyphs.  You're jumping visually in the exact same way in both cases,  
but in the case where you show the glyphs you also have to replace  
them with the right ones.  That is going to be a much more dramatic  
and ugly change, especially if a poor fallback font is chosen (or none  
is specified at all).


Received on Thursday, 7 May 2009 22:01:54 UTC