Re: [CSS21] Suggestion for Introduction of new display types

On Wed, Aug 1, 2012 at 10:59 PM, Ferenc Vajda <fvajda@doxence.com> wrote:
> 2012.08.01. 18:02 keltezéssel, Tab Atkins Jr. írta:
>> On Wed, Aug 1, 2012 at 8:42 AM, Ferenc Vajda <fvajda@doxence.com> wrote:
>>>
>>> I would like to make a suggestion to introduce a new box formatting
>>> model defining two values for 'display' property:
>>>
>>>     container, container-inline
>>>
>>> Both would be rendered as if they were block/block-inline types, the
>>> only difference would be the behavior of vertical-align. The behavior
>>> for this latter is similar to that of table-cell.
>>>
>>> I know that there are suggestions for a solution, but I would prefer
>>> this solution to e.g. the margin-box model, because this behavior is
>>> more like a "box-display" (e.g. table-cell), than a box-sizing or such.
>>>
>>> A display:container could work on body, as well, resulting in a
>>> dynamically changeable window-size box.
>>>
>>> Since table-cell implements this rendering, sharing with the new types
>>> could not be a hard issue for browser developers.
>>
>> We're actually fixing this use-case (easy centering/alignment of block
>> content) via a different method, by taking the generic alignment
>> methods that Flexbox defines and extending them to Block.  Check out
>> http://dev.w3.org/csswg/css3-align/ for details on our approach.
>>
>> So, rather than using "vertical-align", you'll use "align-content".
>
> Hi,
>
> Thanks, this looks great. This solution covers much more issues. Great work.
>
> However, I would still propose a more rapid solution, because Internet
> Explorer follows new features very slowly (and this one is even more
> complicated). Some browsers will support it in some months, but developers
> won't use it.
>
> Extending current features (without breaking principles) could be a quicker
> solution. And years later "CSS Box Alignment Module Level 3" could serve
> community with much more  possibilities.

We'd need to do roughly the same amount of work with either.  Whether
it's a new property or just a new value on an existing property has no
effect - syntax is trivial. The hard part is defining the behavior
correctly and in sufficient detail, which is pretty much identical in
both proposals.

> I think, aligning vertically is an old and urgent issue...

It can't be both old and urgent.  ^_^

~TJ

Received on Thursday, 2 August 2012 16:28:34 UTC