Re: [css3-fonts] revised definition of font-family

Tab Atkins Jr. wrote:

> > During the F2F, Bert stated that he thought this was a change from
> > CSS 2.1, that unquoted font family names like 'foo inherit' should
> > not be rejected as invalid.  I don't really feel strongly either
> > way but I'm wondering if you see a strong reason to make the use
> > of any keyword within a multi-word font family name invalid.
> 
> It *is* a change, but fantasai and I believe that it only
> unintentionally allowed them before.
> 
> The reason to disallow it is to have a consistent story for where
> you can use 'inherit' and 'initial'.  "Only as the sole value of a
> property" is easier to understand and teach than "only as the sole
> value of a property, or a *piece* of a font-family name, unless it
> conflicts with the former".

If there aren't other situations where sequences of identifiers occur,
then I don't think there's really any great reduction in complexity
with this change, the language describing font families still needs to
describe how to merge together sequences of space-separated
identifiers and how to match these against font family names.  By
simply saying that an unquoted font family name cannot be the same as
a reserved keyword, we can define the behavior unambiguously without
changing 2.1 behavior.

I don't think this should be considered a 2.1 issue and we should
avoid adding unnecessary errata unless there's a *very* good reason to
do so.  Keeping Bert happy is also a good thing. ;)

Regards,

John Daggett

Received on Wednesday, 16 May 2012 00:42:15 UTC