W3C home > Mailing lists > Public > www-archive@w3.org > July 2008

[Fwd: Re: [css3-gcpm] [css3-page] Named page lists]

From: fantasai <fantasai.lists@inkedblade.net>
Date: Wed, 02 Jul 2008 16:19:06 -0700
Message-ID: <486C0CEA.7080906@inkedblade.net>
To: www-archive@w3.org

Grant, Melinda wrote:

>  fantasai wrote:
>> Grant, Melinda wrote in response to adding named pages without
>> making 'page' a non-inherited property:
>>> Haven't thought about it, but one option would be to add page-list
>>> or some such.
>> And how would that interact with 'page'? (Two properties that try to
>> control the same thing in CSS doesn't really work.) 
> I dunno, last one in or some such.

Last one specified in the cascade? Yeah. That doesn't work.
It's a little complicated to understand why, so we've had
many proposals over the years -- from spec editors -- that
rely on doing that and.. those proposals have had to get
scrapped and rethought because you can't do it. The cascading
order is effectively per-property: once the cascade is over
every property has a value, and you don't know which properties'
values had a higher cascade weight. You can have a value for
each property that says "check the other property", but
ultimately one property has to always override the other.

> What's your proposal?

If we're sure we want named page lists in the future...

We could make 'page' non-inherited, and add an exception in
the spec saying that UAs may treat it as inherited in CSS3
because that's how it was specified in the previous CR.
I don't think it will have much effect on existing content.
Hmm.. we'd need to add another special value, so we have
the initial value mean "look at my parent" and some other
value mean "put me on a normal, unnamed page even though
my ancestor suggests a named page".

I'm not coming up with anything brilliant today. :/

Michael, any thoughts on this? The discussion started with
this: http://lists.w3.org/Archives/Public/www-style/2008Jul/0029.html

Received on Wednesday, 2 July 2008 23:19:46 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 14:43:26 UTC