This is an archived snapshot of W3C's public bugzilla bug tracker, decommissioned in April 2019. Please see the home page for more details.
In particular, UIEvent still has "readonly attribute AbstractView? view;". AbstractView is long dead.
Please link to the changelog.
https://dvcs.w3.org/hg/dom3events/rev/ce11666f2d43
At least in Gecko this returns WindowProxy... Are we sure this should return Window?
See https://mxr.mozilla.org/mozilla-central/source/dom/webidl/UIEvent.webidl
WindowProxy and Window are the same thing in Gecko's WebIDL. There is never a raw Window returned to script as a capturable object; it's always a WindowProxy. I have no clue what the spec behavior of that is, since the spec state of WindowProxy is pretty broken (e.g. it's not an IDL object at all). My suggestion would be to consistently use Window in WebIDL, with the understanding that in practice what you actually get is a WindowProxy...
"understanding" is not a requirement. We need to fix this somehow and it needs to be crystal clear. What is the bug on WindowProxy? I suggest we make this depend on that.
I'm not sure there's a bug (apart from bug 20701), just mailing list threads.
Since WindowProxy is the secure way to return Window (which is the pattern used in HTML5), DOM Events needs to have this be a WindowProxy as well.
Fwiw, since there is no actual IDL for WindowProxy, writing "WindowProxy" in WebIDL is technically just invalid WebIDL...
I changed the IDLs to WindowProxy, and did a pass through the document for the terms "defaultView" which was the term in vogue prior to a more standardized and well-defined window object. Now there is no more use of the term defaultView, and instead Window and WindowProxy are used where appropriate. https://dvcs.w3.org/hg/dom3events/rev/c2837d35f150