Posted by Ben C on 12/23/06 09:03

On 2006-12-23, -Lost <spam_ninjaREMOVEME@REMOVEMEcomcast.net> wrote:
> "Ben C" <spamspam@spam.eggs> wrote in message
> news:slrneooq0f.m8l.spamspam@bowser.marioworld...
>
>>> The standards state that a floated box *must have* an explicit
>>> width assigned via the "width" property.
>>
>> Which standards and where does it say that?
>>
>>> Your "shrink-to-fit" theory may work in the quirkiest of
>>> situations, but it is not standard.
>>
>> It's not my theory, it's all in CSS 2.1 (which is a "draft
>> recommendation"). Which standard are you referring to?
>
> Oh boy, I hate this part. I hope I have not misunderstood, but:
>
> http://www.w3.org/TR/REC-CSS2/visuren.html#floats
>
> Second paragraph.

You're absolutely right; it turns out this is a difference between CSS
2 and CSS 2.1.

If you look at section 9.5 in CSS 2.1, they cut the bit about explicit
width, and if you compare sections 10.3.5 in both documents, they say:

in CSS 2:

If 'left', 'right', 'width', 'margin-left', or 'margin-right' are
specified as 'auto', their computed value is '0'.

in CSS 2.1:

If ’margin-left’, or ’margin-right’ are computed as ’auto’,
their used value is ’0’.

If ’width’ is computed as ’auto’, the used value is the
"shrink-to-fit" width.

[Back to original message]