Re: Inline formatting model document

--- Tantek Çelik <tantek@cs.stanford.edu> wrote:
> >From: Ian Hickson <py8ieh@bath.ac.uk>
> >Date: Thu, Jan 27, 2000, 1:40 PM
> >
> >
> > On Wed, 26 Jan 2000, Matthew Brealey wrote:
> >>> I know that neither Opera not Mozilla currently underline images as
> >>> they should.
> >> No no no.
> >
> > Yes yes yes. ;-)
> >
> > If 'text-decoration' is applied to a SPAN element which contains,
> > amongst other things, some text, then all the contents of the SPAN
> > should be underlined, since the underlining spans all descendants.
> >
> > Thus:
> >
> >    <span style="text-decoration: underline">
> >       <em>some text</em>
> >       <img src="..." alt="an image">
> >       <em>some text</em>
> >    </span>
> >
> > ...would result in:
> >
> >    some text [X] some text
> >    -----------------------
> >
> > ...and not:
> >
> >    some text [X] some text
> >    ----------   ----------
> >
> > Note that the EM and IMG elements do not have 'text-decoration' set.
> > The underlining does not apply to the IMG element at all, since it is
> > not _set_ on the IMG element.
> 
> Yes, Ian is correct.
> 
> This is how text-decoration works in CSS-1 and CSS-2.  In fact, this
> case is
> tested in the CSS-1 test suite.
Indeed it is, but I though we were on CSS-2 now.

The wording of CSS-1 and CSS-2 is different, so I don't see how much use
something that was designed for an old spec is - the CSS-1 test suite
tests for 'spanning', which is referred to in CSS-1 but not CSS-2.

In fact:
<q>
If the element has no content or no text content (e.g., the IMG element in
HTML), user agents must ignore this property. 
</q>

It seems to me that this creates a pretty strong presumption against
underlining images - it does not say 'it must ignore this property except
when it is underlined due to an ancestor element'.

I don't see why the interpretation as spanning (rather than , as the
*current* spec says, 'affecting') should be made. It gives ugly results.
> >On Wed, 26 Jan 2000, Matthew Brealey wrote:
> >> I know that neither Opera not Mozilla currently underline images as
> >> they should.
> 
> When making statements like that, at a minimum out of respect for the
> implementers, you really should check the CSS1 test suite first to see
> if it
> agrees (or perhaps even helps demonstrate) your assessment.  If you
> disagree
> with the CSS1 test suite (which has been worked on long and hard by many
> many
> CSS experts), bring that up and also send mail to css-test@w3.org
Seems you've gotten a little confused.

I did not write that - Ian Hickson did (in
http://lists.w3.org/Archives/Public/www-style/2000Jan/0174.html)

So perhaps before you make statements like that you should check whether
the person whom you accuse of making such statements has actually said
what you think they did.


=====
----------------------------------------------------------
From Matthew Brealey (http://members.tripod.co.uk/lawnet (for law)or http://members.tripod.co.uk/lawnet/WEBFRAME.HTM (for CSS))
__________________________________________________
Do You Yahoo!?
Talk to your friends online with Yahoo! Messenger.
http://im.yahoo.com

Received on Friday, 28 January 2000 06:46:30 UTC