The one in the line below requires a lot more work... I'll deal with that one later.

Comment on attachment 8900064 [details] [diff] [review] Part 1: Add an override of EditorUtils::IsDescendantOf() which returns an nsIContent* out argument for the child node found instead of its index in the parent > Bug 1392867 - Part 1: Add an override of EditorUtils::IsDescendantOf() which an overload?

Comment on attachment 8900065 [details] [diff] [review] Part 2: Avoid calling GetChildAt() in HTMLEditRules::TryToJoinBlocks() Really nice!

Pushed by eakhgari@mozilla.com: https://hg.mozilla.org/integration/mozilla-inbound/rev/30fd4ff28505 Part 1: Add an overload of EditorUtils::IsDescendantOf() which returns an nsIContent* out argument for the child node found instead of its index in the parent; r=masayuki https://hg.mozilla.org/integration/mozilla-inbound/rev/a93b3e95a267 Part 2: Avoid calling GetChildAt() in HTMLEditRules::TryToJoinBlocks(); r=masayuki

An nsIContent** outparam that doesn't addref seems like it should be very clearly documented.

https://hg.mozilla.org/mozilla-central/rev/30fd4ff28505 https://hg.mozilla.org/mozilla-central/rev/a93b3e95a267

Pushed by eakhgari@mozilla.com: https://hg.mozilla.org/integration/mozilla-inbound/rev/b4f59a5d13d0 follow-up: Document that the aChild argument won't get AddRef'ed