Re: [CSSWG] Minutes and Resolutions 2010-04-21

On Sun, May 9, 2010 at 7:07 PM, Boris Zbarsky <bzbarsky@mit.edu> wrote:
> On 5/9/10 9:56 PM, Tab Atkins Jr. wrote:
>>>
>>> The "1px to the right" thing doesn't make sense.  You'll end up with gaps
>>> that way, no?  Unless you don't mean CSS px?  Or unless you're defining
>>> "corner" in some way that doesn't quite jump out at me, or both.
>>
>> Sorry, that's meant to specify things to that it won't overlap the
>> previous cell.  The way I was defining "corner" in my head was
>> referring to the top-right pixel of the previous cell, and so the next
>> cell will start 1px over.
>
> Probably better to define in terms of edges (padding edges, etc) then.

An asbpos table-cell's auto position is defined as the first of the
following that exists:

Its vertical position is:
* the top inner edge of the padding box of the table-row it would be
placed in if it weren't absposed
* the top inner edge of the padding box of the following table-row
* the bottom outer edge of the border box of the preceding table-row
* the top inner edge of the padding box of the table

Its horizontal position is:
* the left outer edge of the border box of the following table-cell
* the right outer edge of the border box of the preceding table-cell
(this includes row-spanned cells that intrude into the row the
table-cell would be placed in if it weren't absposed)
* the left inner edge of the padding box of the table-row
* the left inner edge of the padding box of the following table-row
* the left inner edge of the padding box of the preceding table-row
* the left inner edge of the padding box of the table

~TJ

Received on Monday, 17 May 2010 19:05:09 UTC