Re: [css3-multicol] unknown available width and shrink-to-fit

Le 05/11/2012 00:40, Simon Sapin a écrit :
>> It seems that the differences you get is due to
>>    - testing the shrink-to-fit algorithm  (which isn't descibed in the multicol spec)
> Yes, this is exactly my point. There is no interop on shrink-to-fit. If
> we define shrink-to-fit elsewhere, then some parts of the algorithm
> become redundant and should be removed.

In other words, I believe than intrinsic sizing (and thus shrink-to-fit) 
is the only case where used-width can ever be unknown.

-- 
Simon Sapin