Re: [css3-flexbox] alignment test

On 12/08/2011 12:51 PM, Tab Atkins Jr. wrote:
> On Thu, Dec 8, 2011 at 12:15 PM, fantasai<fantasai.lists@inkedblade.net>  wrote:
>> On 12/07/2011 02:07 PM, Tab Atkins Jr. wrote:
>>> On Wed, Dec 7, 2011 at 1:54 PM, fantasai<fantasai.lists@inkedblade.net>
>>>   wrote:
>>>> Firefox's flexbox behavior requires adding<spacer>    elements in order to
>>>> have some things align to the left and others align to the right.
>>>>
>>>> I would like that not be the case.
>>>
>>> I believe it's cleaner and easier to understand if we wait for the
>>> ability to create arbitrary pseudo-elements, which we expect to do
>>> anyway to help out with Regions.  Then you can use pseudo-elements as
>>> spacers, and have more control over alignment as well.
>>
>> I don't think creating elements or pseudo-elements is a good way of
>> controlling spacing. That's what margins are for. We've come a long
>> way since using spacer GIFs, please let's not go back to that.
>
> We've got three options:
>
> (1) use pseudo-elements here, because they play nicely with the layout
> algorithm and easily expose *all* the knobs you might want
> (2) use margins, which either means we have a relatively crappy
> interaction with the rest of flexbox (all margins start from 0, have
> the same flex, and are at a particular unchangeable location in the
> flex hierarchy with elements and packing), or we complicate margins to
> give them full access to flex abilities
> (3) introduce another property that lets you insert "spacers", which
> act like margins but have full access to the flex stuff.
>
> I don't like (2).  I'm okay with (1), and we're going to end up
> designing the pieces it needs anyway, so it becomes essentially free.
> I'm also okay with (3), but I don't think it's needed quite yet.  I'd
> design it if the WG thinks it's necessary, though.

What controls do you need, besides what is offered with flex()?

~fantasai

Received on Thursday, 8 December 2011 21:58:33 UTC