See https://bugs.chromium.org/p/chromium/issues/detail?id=671907.
The usage counter is: https://www.chromestatus.com/metrics/feature/popularity#EventPath

I think Event#path can be deprecated separately from other Shadow DOM v0 APIs.

> whether we can just change its type to a sequence<> or a FrozenArray<>.

Is this user visible change?

Thanks for the explanation, it's nice to know this has already been discussed before. Is 2.19% low enough a number for us to consider deprecating/removing it?

>> whether we can just change its type to a sequence<> or a FrozenArray<>.
> Is this user visible change?

sequences cannot be used as attribute types, and a FrozenArray would make the returned object frozen, which is a user-visible change. Another possibility is to make it an object instead of an EventTarget[] and handle the conversion to an array/sequence in the C++ side (this should be transparent to users I think).

2.19% is pretty high. :)

I remember that I once tried to deprecate Event.path, but that was objected. :(
See https://groups.google.com/a/chromium.org/d/msg/blink-dev/8_x0OHYQdx0/jquuYXWWLOYJ

> sequences cannot be used as attribute types, and a FrozenArray would make the returned object frozen, which is a user-visible change. Another possibility is to make it an object instead of an EventTarget[] and handle the conversion to an array/sequence in the C++ side (this should be transparent to users I think).

Thank you for the explanation. I understand that this is a user visible change,
however, I think the risk is low. As long as we announced the change at blink-dev@, and having an entry in https://www.chromestatus.com/features, I think it is okay to change. I support it.

Event.path should be avoided, in favor of Event.deepPath(), anyway.

I was considering using an `object' in the IDL file; since this isn't going to be spec'ed, doing so errs on the side of caution and should not have any user-visible changes at all.

I may be overlooking it, but I couldn't find any layout tests using Event#path to double-check my patch. Is there something I missed?

Maybe I've renamed all the usage of Event#path of layout tests with Event#composedPath(). :(

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/5232f5ecbe69c962d633d62c3177f7d5f7e6bbbf

commit 5232f5ecbe69c962d633d62c3177f7d5f7e6bbbf
Author: Raphael Kubo da Costa <raphael.kubo.da.costa@intel.com>
Date: Wed Aug 09 11:27:45 2017

Declare Event#path as an object, not an EventTarget[].

Arrays were removed from WebIDL in 2015, and this is another step towards
removing all usages from Blink.

Since Event#path is part of Shadow DOM v0 and thus not part of any active
spec, take the least disruptive path by turning the EventTarget[]
declaration in the IDL into an object.

This should not have any user-visible effects, as the C++ side returns the
same array it returned before, but manually wrapped inside a ScriptValue.

Bug:  740870 
Change-Id: Idb0574c4e7e1188f1e4202e9e6cd056582a6e4c1
Reviewed-on: https://chromium-review.googlesource.com/605648
Commit-Queue: Raphael Kubo da Costa (rakuco) <raphael.kubo.da.costa@intel.com>
Reviewed-by: Yuki Shiino <yukishiino@chromium.org>
Reviewed-by: Hayato Ito <hayato@chromium.org>
Cr-Commit-Position: refs/heads/master@{#492942}
[add] https://crrev.com/5232f5ecbe69c962d633d62c3177f7d5f7e6bbbf/third_party/WebKit/LayoutTests/shadow-dom/v0/event-path.html
[modify] https://crrev.com/5232f5ecbe69c962d633d62c3177f7d5f7e6bbbf/third_party/WebKit/Source/core/events/Event.cpp
[modify] https://crrev.com/5232f5ecbe69c962d633d62c3177f7d5f7e6bbbf/third_party/WebKit/Source/core/events/Event.h
[modify] https://crrev.com/5232f5ecbe69c962d633d62c3177f7d5f7e6bbbf/third_party/WebKit/Source/core/events/Event.idl

I think this is now fixed without any user-visible changes. I also resurrected one of the tests that used to check for Event.path, so hopefully there's some (minimal) coverage now.