On Mon, Jan 12, 2015 at 11:51 AM, Anne van Kesteren <ann...@annevk.nl>

> On Mon, Jan 12, 2015 at 5:47 PM, Brian Kardell <bkard...@gmail.com> wrote:
> > Controlling it through CSS definitely seems to be very high-level.  To
> me at
> > least it feels like it requires a lot more answering of "how" since it
> deals
> > with identifying elements by way of rules/selection in order to
> > differentially identify other elements by way of rules/selection.  At the
> > end of the day you have to identify particular elements as different
> somehow
> > and explain how that would work.  It seems better to start there at a
> > reasonably low level and just keep in mind that it might be a future aim
> to
> > move control of this sort of thing fully to CSS.
> I'm not sure I agree. Unless you make all of CSS imperative it seems
> really hard to judge what belongs where.

It's important that at least qsa/find/closest style things defined in
Selectors match the same in script as in CSS matching.  Whatever solution
likely needs to include this consideration.

I've heard some worries about async/sync requirements regarding rendering
here but I'd say it's further than that if it's rule based too from my
perspective - this seems like something we're going to have to deal with
anyway in a larger sense of extensibility.  I wouldn't (personally) let
that dictate that we can't do this in script - there are lots of places
where that seems practical/controllable enough even now and we could make
globally better with some effort.

Basing this off something in CSS matching, as opposed to DOM means that new
roots can come into play (or leave) in a document-wide sense based on the
addition or removal or rules.  This seems confusing and problematic to me
and the combination of these was relevant to my comment about what's
matching what.  It seems to me that identifying a shadow root is concrete
to an instance and once it's there, it's there.  You can consciously choose
to combinator select through it or not, but it's there unless the physical
DOM changes.

> --
> https://annevankesteren.nl/

Brian Kardell :: @briankardell :: hitchjs.com

Reply via email to