[css-nav-1] Modify the definition of the distance function #3755
Conversation
Change the meaning of factor B as the resolution from #3384
Modify the distance function depending on the updated spec (w3c/csswg-drafts#3755)
| * <var>alignBias</var> | ||
|
|
||
| * If the <var>dir</var> is {{SpatialNavigationDirection/left}} or {{SpatialNavigationDirection/right}}, | ||
| height of <var>intersectRect</var> / height of <var>candidate</var> |
There was a problem hiding this comment.
In #3388 you wrote "length of the edge of the search origin". Here you wrote candidate's edge" instead... Mistake?
There was a problem hiding this comment.
The edge of search origin on the comment (#3388) has been replaced with the current expression on the spec which describes an edge of candidate. If you have another opinion, please put it here or make an issue so that we could consider. With minor updates (figures of some weights), we are going to apply the updated expression into Blink.
BTW, the polyfill would operate it with the edge of search origin that seems not to match with the current spec. @jihyerish please check and update on it.
https://github.com/WICG/spatial-navigation/blob/master/polyfill/spatial-navigation-polyfill.js#L1059
There was a problem hiding this comment.
I created crbug.com/958845 for the Blink work.
See this example:
____________________ top edge of container or other focusable. 20 units wide.
YYYYY XXXXXXXXXXXX
Here we want X to win so we do 12/20 .
12 / 20 > 5 / 20 => we give X a larger alignment because we want to decrease the distance to X more than we decrease the distance to Y. (alignment is a negative term in the formula).
There was a problem hiding this comment.
Yes, you're right. My mistake...
Thanks for pointing this out! I'll fix this.
This changes aim to update the spec with the improved distance function (Fixes #3384)