W3C home > Mailing lists > Public > www-style@w3.org > January 2008

Re: Generated and replaced content module (was RE: BR element and generated content)

From: fantasai <fantasai.lists@inkedblade.net>
Date: Thu, 17 Jan 2008 15:23:40 -0500
Message-ID: <478FB94C.9080001@inkedblade.net>
To: White Lynx <whitelynx@operamail.com>
CC: bert@w3.org, ian@hixie.ch, www-style@w3.org

White Lynx wrote:
>>> Would be great to see this text in CSS3 Generated Content. You are 
>>> listed as the only editor - do you intent to be updating it or is the 
>>> spec in need of an owner?
>> If nobody gets to it before I do, I intend to get back to it around 2018 
>> (after HTML5).
> Bert and Ian, I can look at it before 2018. IMHO it is important enough to be maintained.
> My suggestion would be to add Generated and replaced content module status issue to one
> of the phone conferences agenda in order to consider further maintainance of module and
> to add (co)editor.

I think we need to drastically trim down the Generated Content module if
we're going to make serious progress on it. dbaron suggested that the
latest public draft might not be the latest text that hixie had; in that
case we should copy hixie's latest version to dev.w3.org, possibly publish
a WD of it if there are significant changes since the last version that we
want archived, and then republish a WD with just the essential features that
we need *soon* so we can start to stabilize those and send them to CR. E.g.
we need 'content' on all elements (which iirc Opera already implements),
fallbacks, ::marker and the list-item counter. Anything beyond the basics
can be developed in a separate spec.

Hixie, you got a pointer to your latest css3-content spec, or can you upload
a copy to dev.w3.org/csswg/css3-content/?

BTW, here's a copy of a message of mine on this topic. Håkon also had some
comments, I'll ask him if I can forward them here. We need to start collecting
past feedback into an issues list like on http://csswg.inkedblade.net/spec

-------- Original Message --------
Subject: Re: content: require-font()
Resent-Date: Mon, 07 May 2007 21:45:32 +0000
Resent-From: w3c-css-wg@w3.org
Date: Mon, 07 May 2007 17:45:12 -0400
From: fantasai <fantasai@inkedblade.net>
To: w3c-css-wg@w3.org
References: <17983.34752.462167.287328@gargle.gargle.HOWL>

Håkon Wium Lie wrote:
 > At a recent FTF, we discussed downloadable fonts and image
 > replacements:
 >   http://lists.w3.org/Archives/Member/w3c-css-wg/2006AprJun/0111.html
 > There was consensus around a proposal to introduce a 'require-font()'
 > value on the 'content' property. I'd like to see that proposal written
 > up and published. I can volunteer to write into GCPM if there is no
 > new Fonts draft forthcoming.

I wrote up the proposal here:

 > Also, there can be an optional keywoord after require-font():
 >     h1 { content: require-font("foo") contents, url(bar);

That keyword is part of the 'content' property definition, it's
not an artifact of the require-fonts() proposal. See

 > I suggest using "self" instead of "contents" for this. Because:
 >   - "self" is shorter
 >   - "self" has no singular/plural dilemma
 >   - "contents" is very close to the name of a property
 >   - "self" is used in GCPM

We need to discuss the 'content' property. 'contents' isn't the only
part that needs discussion. See e.g. my comments on CSS3 Paged Media,
which references the 'none' value:
   | 'none' is not currently defined as a 'content' value except in CSS3
   | Generated Content. CSS2.1 uses 'normal', and 'normal' is a valid
   | computed value in CSS2.1-- but not in CSS3 Generated Content. Also
   | it seems like the desired behavior is that given by 'inhibit'.
   | Perhaps the CSSWG needs to discuss css3 'content' and nail its values
   | down.

I think what we need is a pared down CSS3 Generated Content module. It
needs to include
   - stuff from CSS2.1
   - ::marker and ::line-marker
   - 'content' property with
        - 'contents' keyword
        - any necessary "I want nothing" keywords
        - comma-separated fallbacks
        - require-font()
and that's it.

Note: require-font() should always succeed in non-visual media.

Received on Thursday, 17 January 2008 20:23:51 UTC

This archive was generated by hypermail 2.3.1 : Monday, 2 May 2016 14:27:33 UTC