With respect to Dimitri's *unpacking* notion, we think it can work.

In most cases today the redistribution tree is not large; however, if there
were a lot of final destination points in that tree, unpacking would be
cumbersome because you need to define a slot at every level for every final
destination point from that level's perspective. We'd like the working
group to explore some short hand syntax for this scenario.

The one hypothetical use case we came up with that would be problematic for
*unpacking* is dynamic creation of insertion points. This would require
dynamic unpacking and that seems pretty ugly since you'd have to break
encapsulation to do this. However, the use cases that one can imagine for
dynamic insertion points would be better solved by an imperative
distribution api anyway and are generally advanced enough that v2 seems ok
for them.

On Fri, May 15, 2015 at 5:55 PM, Dimitri Glazkov <dglaz...@google.com>

> On Fri, May 15, 2015 at 5:45 PM, Scott Miles <sjmi...@google.com> wrote:
>> >> How does it work for redistribution
>> We've done some investigation and think it can work.
> Note that distributions work just fine with slots. The only thing that
> doesn't is partial redistributions, and we haven't been able to find a
> scenario that's a blocker:
> https://github.com/w3c/webcomponents/blob/gh-pages/proposals/Partial-Redistributions-Analysis.md
> Please, if you can think of one, let's look at it.
> :DG<

Reply via email to