809547 - Investigate whether we need to preserve Location wrappers

Status: RESOLVED FIXED

(Core :: XPConnect, defect)

Description

[Bobby Holley (:bholley)]

Currently we don't, so expandos can probably disappear (I'm going to verify this). Given that we're nailing down our expando behavior for Location objects over in bug 808608, I think it's worth being deliberate and consistent here.

Comment 1

[Bobby Holley (:bholley)]

Attachment: Tests. v1

Ah, looks like we don't need to because the current implementation of Location
getters in nsDOMClassInfo end up defining it as a value prop (but with a setter)
on the window, so it's never collected.

I don't know enough about the new bindings to know whether we risk regressing
this at all when we move Location to the new bindings. Boris, if you think this
test is unnecessary, please just r- it and resolve the bug. :-)

Comment 2

[Boris Zbarsky [:bzbarsky]]

Comment on attachment 679297 [details] [diff] [review]
Tests. v1

New bindings for Window would make this test fail, but new bindings for Location (which I would hope come first) would make it preserve wrappers correctly anyway...

I think we should land the test.

Comment 3

[Bobby Holley (:bholley)]

https://hg.mozilla.org/integration/mozilla-inbound/rev/e252b7ccec6e

Comment 4

[Ryan VanderMeulen [:RyanVM]]

https://hg.mozilla.org/mozilla-central/rev/e252b7ccec6e