Re: [css3-backgrounds] Should a non-zero border-radius create a new stacking context ?

On Nov 22, 2010, at 12:57 PM, Sylvain Galineau <> wrote:

> If the side effects of creating a new stacking context is so likely to break pages then opacity and transforms are equally problematic for authors. 

It may be that the change would not break pages, but I don't know that from this argument alone. Opacity < 1 has ALWAYS created a new stacking context, hasn't it? So there was nothing to break before this became true for opacity. Bur border-radius has been in use for many years, and now with no prefix. So I'd like to know if existing content would be broken, less than I want to know about whether or not authors found workarounds for this possible limitation with opacity. 

I think we generally try to limit jumpy changes that depend on the precision of rounding to zero, and mode switches that occur by using .0001px instead of zero. Do you feel this would be an unnoticeable difference in this case, and that it is fairly safe to chance it?

Received on Monday, 22 November 2010 22:00:49 UTC