W3C home > Mailing lists > Public > www-style@w3.org > May 2013

Re: [css3-fonts] "font-feature-settings: none" shorthand?

From: Sylvain Galineau <galineau@adobe.com>
Date: Wed, 29 May 2013 13:14:34 -0700
To: fantasai <fantasai.lists@inkedblade.net>, "www-style@w3.org" <www-style@w3.org>
Message-ID: <CDCBAA02.509A%galineau@adobe.com>

On 5/29/13 10:35 AM, "fantasai" <fantasai.lists@inkedblade.net> wrote:

>On 05/29/2013 01:03 PM, Tab Atkins Jr. wrote:
>> On Wed, May 29, 2013 at 9:11 AM, Glenn Adams <glenn@skynav.com> wrote:
>>> It would be useful to have a shorthand on font-feature-settings that
>>>is the
>>> same as enumerating all features and specifying 'off' for each feature.
>>
>> Agreed.  As a general principle, any list-valued property should accept
>>"none".
>
>Um, this makes no sense at all. There's a 'normal' keyword, which does the
>appropriate thing and undoes what font-feature-settings does. There's
>absolutely no reason to have a feature that turns all OpenType features
>unilaterally off. Even if there was a reason, it'd be the
>0.000000000000001%
>use case, at which point, just make a custom font without any features.

I agree with Elika. Fwiw one potential usability pain point I see with this
feature is that when one enables features A, B and C on a parent and
then wants to turn only B off on a descendant they do that by setting A
and C 
on the descendant. 

Of course the feature is only meant to be a rarely used low-level
shorthand to 
those OpenType features that aren't formalized as CSS properties so this
shouldn't be much of an issue in practice.

But until all the other font-variant-*/font-kerning properties are widely
supported font-feature-settings remains the cross-browser door to OT
features 
and this shorthand behavior makes some CSS somewhat harder to maintain;
it's also 
a bit confusing at first. I know I was surprised the first time I used it
to turn 
on kerning on a top-level element and the descendant paragraph where I had
activated 
ligatures was not kerned.

For the record I think the right solution is for implementors to support
font-variant* 
and friends. But until that happens I suspect this may bite the
typographically inclined
far more often than the lack of a global 'OT off' which I'm not even sure
there is a 
use-case for.

Received on Wednesday, 29 May 2013 20:15:10 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 29 May 2013 20:15:10 UTC