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

Re: Suspect CSS2.1 test case

From: Ben Cotterell <ben.cotterell@antplc.com>
Date: Fri, 12 Sep 2008 07:47:49 +0100
To: "L. David Baron" <dbaron@dbaron.org>
CC: www-style@w3.org
Message-ID: <0630aa10e4376d210411f759f0141e8e35fd7568@localhost>

On Thu, Sep 11, 2008 at 04:35:02PM -0700, L. David Baron wrote:
> 
> On Thursday 2008-09-11 22:18 +0100, Ben Cotterell wrote:
> > Still it is very unusual. Nowhere else do you have to take into account
> > what something's height is going to be when working out its available
> > width.
> 
> Well, it's more that if it doesn't work out, you retry at the
> narrower width.

Yes, exactly.

> (You can't know how tall a line is going to be until you know what's
> on it, and you can't know what's on it until you know how wide it is.

Indeed.

> However, retrying at a narrower width can't mean you end up with more
> on the line, so the worst case is that you retry once and end up
> sticking with a result that's shorter than necessary.)

It's a similar problem to another bug Firefox used to have (that
actually our browser also has)-- you're supposed to try to align a float
to the top of the line box of the line it's on. But if that makes the
line break before the character position where the float was, you try
again with the float on the next line. The way to fix it is the same:
you need to be able to "retry" lines.
Received on Friday, 12 September 2008 06:51:39 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 27 April 2009 13:55:12 GMT