526209 - provide a way for chrome JavaScript to change navigator.userAgent on a per-page basis

Status: RESOLVED INCOMPLETE

(Core Graveyard :: Embedding: APIs, enhancement)

Description

[Benoît]

To alleviate the problem of web browser discrimination from our users, I'm developing a Mozilla extension that spoofs the user agent string on a per-page basis.

Currently I have an JavaScript XPCOM component that intercepts HTTP requests and changes the User-Agent header if necessary. This works for websites that only use server-side scripting to determine what web browser is making the HTTP request. However, this doesn't work for websites that, in addition to checking HTTP requests, have JavaScript code that checks navigator.userAgent and decide based on the result what code to (not) execute. This is the case for FaceBook, for example.

However, there is currently no way for chrome JavaScript to change the navigator.userAgent object on a per-page basis. If this would be possible, then temporary spoofs of the object for websites like FaceBook would be possible.

Comment 1

[Robert Kaiser]

Actually, I think navigator.userAgent for any "page" should just always match the User-Agent header sent in the HTTP request that loaded this document. That should make the spoofing solution only have one point where to set this, and would make us stay consistent.

Comment 2

[Robert Kaiser]

timeless tells me his work in bug 334875 should provide a way to do this - now we only need to drive that work forward...

Comment 3

[Ben Bucksch (:BenB)]

window.navigator is implemented in nsGlobalWindow.cpp. I.e. that's custom JS code that's executed for every invocation (to my knowledge), and nsGlobalWindow.cpp should be able to know its URI, so I don't see why this wouldn't be possible to implement without too much trouble.

Comment 4

[Robert Kaiser]

If I read the code correctly, that would sounds like it would take the user agent string itself from the HTTP channel, actually, but Benoît said that wouldn't work correctly. Interesting - but then I might misread the code, as I basically don't have a clue about C++ and might misunderstand it.

Comment 5

[:Gavin Sharp [email: [e-mail]]]

I don't think bug 334875 helps with this at all.

Comment 6

[timeless]

gavin: in theory it should. if navigator is a per window object it can be subsumed by one which handles userAgent specially.

Comment 8

[Benjamin Smedberg]

Marking a bunch of bugs in the "Embedding: APIs" component INCOMPLETE in preparation to archive that component. If I have done this incorrectly, please reopen the bugs and move them to a more correct component as we don't have "embedding" APIs any more.