Re: [css-logical-properties] the 'inline-{start,end}' values for 'float' and 'clear'

On Thu, Nov 5, 2015 at 5:48 AM, Florian Rivoal <florian@rivoal.net> wrote:

>
> On 05 Nov 2015, at 11:17, Johannes Wilm <johanneswilm@vivliostyle.com>
> wrote:
>
>
>
> On Thu, Nov 5, 2015 at 1:46 AM, Florian Rivoal <florian@rivoal.net> wrote:
> ...
>
>>   "full 2d" floats are a complicated matter, and I am far from sure
>>   a background-position inspired syntax will cut it.
>>
>
> Do ever need full 2D floating?
>
> Does it make sense to only float in the block direction without also
> floating in the inline direction.
>
> If we don't need to float only in the block direction. For a simple
> version of page floats, I think we will need these directions as a minimum:
>
> up/down/right/left (filling 100% in the direction it is not floated in)
>
> right/left (using existing block-level)
>
> top-left-left (priority for left)
> top-top-left (priority for top)
>
> top-right-right
> top-top-right
>
> bottom-left-left
> bottom-bottom-left
>
> bottom-right-right
> bottom-bottom-right
>
> center
>
> top-center-center
>
> top-center-right (if this is not the first top-center float, try to place
> the float to the right of the top center float)
> top-center-left
>
> bottom-center-center
>
> bottom-center-right
> bottom-center-left
>
>
> Right. There are also things like "float to the bottom of the last column
> on a right page",
> "Float to the top of a particular slot in a page template"...
>
> All in all, I am convinced that floats in more than 1D make sense, but I
> am not
> convinced it is a conventional 2D system.
>
> I may be convinced otherwise, and I am happy to have that discussion. But
> if we
> need an answer now about start vs inline-start, inline-start seems more
> future proof.
>


And floats that span columns, or two half-columns. And in print one may
want to float color images all to the same few pages so that only those
will be printed in color whereas everything else is in black and white.
Somewhere we will need to do the cut-off of what is likely to be provided
by browsers, and it will hit also some layout options that are used in
print today, I am afraid.

The 2 x 1D seemed like a simplification that would still cover a large
percentage of common layout options. But we can try to move the cut-off
somewhere else, if implementers think it is realistic they will actually
ship that.

The one thing I would like to avoid is to define a lot of advanced options
that then only will work if there is only one float within an
block/fragment, and break if there are more.


>  - Florian
>