Re: [css3-fonts] font descriptor default values

Michael Day wrote:
> Hi Thomas,
>
>> Because of this, my conclusion has always been that CSS was
>> fundamentally broken as far as font selection is concerned.
>
> Isn't this an issue that can be solved by font authors and platform 
> developers?
Browser vendors may take matter in their own hands, and wrap their own 
code for "font matching". This is a complex issue though. One cannot 
rely on the platform doing the "right thing" since although it may be 
correct for the platform, it may by definition not allow for a cross 
platform solution. I don't expect most UA to bother though, given that 
people don't necessarily expect text to look exactly the same across 
platforms and browsers.
>
> If someone makes a new font called "Vulgaris Neue" with several 
> weights and in condensed and expanded versions they should have no 
> problem keeping all these faces within the same family if they so choose.
For sure. There may be enough information in the fonts to identify fonts 
by the family name, which a particular platform may not support. This 
very much depend on the font format and the platform.
>
> On the other hand, if someone puts every face in a different family, 
> it's not clear why CSS should be responsible for doing the heavy 
> lifting to try and straighten everything out.
I would agree as well that we are talking about an inherit difference 
between platforms that is causing problems, and not necessarily an issue 
with CSS. There are lots of other issues with CSS from a typographical 
point of view though, where things are just plain wrong (e.g. line 
height and vertical alignment of text) or just not clearly defined 
(underline position and thickness, etc). Will be bugging this list in 
the future about that ;-) CSS rocks, but it needs more tuning to get the 
typography right.

I guess this boils down to expectation as well. Given that most people 
don't use fonts in web pages beyond sans-serif, there are lots of issue 
that never surface (like Thomas mentioned before). People don't use 
stuff that don't work, or at least accept that they don't really have 
full control of the text in web pages.
>

>
> Cheers,
>
> Michael
>

Received on Wednesday, 4 March 2009 08:06:17 UTC