This is an archived snapshot of W3C's public bugzilla bug tracker, decommissioned in April 2019. Please see the home page for more details.
See https://bugs.webkit.org/show_bug.cgi?id=92050 for a detailed discussion. The gist is that "tabindex" on distributed nodes indicates that the author is attempting to include them into the document navigation order, not shadow DOM subtree navigation order, but the heuristic for understanding what to do with the shadow DOM subtree navigation order in this case is weak.
I remembered that 'tabscope' was proposed to whatwg last year. http://lists.whatwg.org/htdig.cgi/whatwg-whatwg.org/2011-November/033775.html That brings me one idea: 1. If a shadow host has 'tabscope' attribute, distributed nodes join the focus navigation of shadow DOM subtree. 2. Otherwise, distributed nodes join the focus navigation of outer document order.
s/tabscope/tabindexscope/
Ooh, interesting. Let me dig into this.
Here's my proposal: 1) Change the spec to make distributed nodes participate in the outer navigation order. 2) Let other tools in HTML (such as tabscope) to deal with scoping of navigation order in other ways. WDYT?
Sounds reasonable for me. My proposal of #1 is same to that. Let's update the spec. (In reply to comment #4) > Here's my proposal: > > 1) Change the spec to make distributed nodes participate in the outer > navigation order. > 2) Let other tools in HTML (such as tabscope) to deal with scoping of > navigation order in other ways. > > WDYT?
What is a distributed node? ( Not definee in Shadow DOM draft )
That are the children of a shadow host, distributed to the insertion points. (In reply to comment #6) > What is a distributed node? ( Not definee in Shadow DOM draft )
(In reply to comment #6) > What is a distributed node? ( Not definee in Shadow DOM draft ) Yes, I was wrong in terminology. They are child nodes, distributed to insertion points.
http://dvcs.w3.org/hg/webcomponents/rev/213b8553fb0c