[whatwg] css3-fonts: New values for generic font families

[Nicholas Shanks]

Hi Håkon, thanks for replying.

> Why not just follow the guidelines in the CSS3 font module?:

Ahh, I didn't see there were instructions on the module itself on  
where to send suggestions, and it doesn't give the main author's name  
(just "the CSS2 authors and Tantek…" et al).
I was on that css list a few years ago. It was full of spam (people  
asking simple "how do I do x with CSS?" questions) and no-one paying  
attention to my suggestions for future implementations so I abandoned  
it.

> While I appreciate the convenience this new functionality may have for
> designers wanting to see text in (say) "blackletter", the
> inconvenience for browser implementors will be disproportionately
> large. Where will they find these fonts?

I don't think there would be any inconvenience.
Either they can allow the user to select a font from all those  
available, or it can filter down using a list of keywords such as the  
ones I suggested. They could also hard-code a few common fonts that  
don't get picked up by the keywords. Perhaps a combination of the  
two, with filtered blackletter fonts at the top of the menu, a  
divider, then all the rest. The same can go for script font,  
filtering the list against keywords such as script, hand, child, pen,  
felt, crayon

> Will they have to ship fonts with browsers?

Not any more than they do currently, which is none. :-)

> The current number of generic font families (5) is already stretching
> it; one might argue that even "fantasy" and "cursive" should be
> dropped as many systems don't offer fonts in these categories.

I don't think there's much use for fantasy, as what exactly  
constitutes a fantasy font was never clearly defined. And if cursive  
is removed, it should be replaced by more descriptive variants such  
as "calligraphic" and "handwritten"

> A better way to support interesting fonts is -- IMHO -- for browsers
> to start supporting TrueType Webfonts.

I have been campaigning for better @font-face support for years too,  
and I will implement it in WebKit once all the more pressing  
typography and CSS bugs that bother me are taken care of (like using  
more than two font weights, proper small-caps and being able to  
switch between lining and text numerals).

The biggest problem I see with using @font-face currently is that not  
many fonts provide the correct Panose 1 numbers, making matching to a  
close substitute very hard. All the concern about violating copyright  
mentioned in the Mozilla bug on this rule are IMO invalid, as there  
are already so many more interesting way to violate copyright than this!
By the way, if you're looking for someone to implement this and other  
CSS stuff in Opera, I am currently seeking employment. I have been  
having difficulty getting anywhere with Apple due to the United  
States' negligible work visa quotas.

- Nicholas.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 2373 bytes
Desc: not available
URL: <http://lists.whatwg.org/pipermail/whatwg-whatwg.org/attachments/20060702/8daf1b09/attachment.bin>