Re: [css3-lists] list-style-position behavior

On Thu, Feb 17, 2011 at 1:00 PM, Aharon (Vladimir) Lanin
<> wrote:
> I think it would help matters to actually demo how different combinations of
> list-style-position and text-align would behave under Tab's proposal, or at
> least my understanding of it. I am attaching three screenshots - one for
> inside, one for outside, and one for hanging. They are simulated, of course,
> so if I am mistaken, please correct me.

Very nice!  Yes, these diagrams correctly illustrate what happens for
each value.

> Please note that I am assuming that Tab meant that with
> list-style-position:outside, not only would the marker show up on the start
> side of list container's direction (not the item's direction), but the
> marker text would be displayed in the list container's direction (not the
> item's direction). With inside and hanging, the marker is displayed in the
> item's direction. The difference is significant ("1." vs ".1").

Ah, I didn't think of that (letting list-style-position affect the
directionality of the marker).  Good idea.

> Some observations about the results:
> 1. I hope that the intent is to have one default across all browsers,
> despite history. Furthermore, the existing mess affecting bidi documents
> will not be fixed unless that default is (or, nominally, stays) "outside"
> (with the now precisely defined meaning).

Yes, of course.

> 2. I think that the HTML5 default stylesheet should have li
> {text-align:match-parent}. It only makes a difference for list items whose
> direction is opposite to the list, but for them it is what works best (for
> most tastes and cases) with list-style-position:outside. Those who prefer to
> use list-style-position:hanging with bidi lists would probably prefer li
> {text-align:start}, but I think that they are in the minority, and have to
> explicitly set up padding on both sides anyway, so it should not be a big
> deal to also specify li {text-align:start}.
> There is additional pressure to reach a decision about 2, since it is
> something that needs to make its way into the HTML5 spec.

Agreed, though this isn't something I can do in the Lists spec - it
requires 'match-parent' to be added to 'text-align'.


Received on Tuesday, 22 February 2011 20:01:18 UTC