566151 - Add "Maemo" to UA string

Status: VERIFIED FIXED

(Firefox for Android Graveyard :: General, defect)

Description

[Vladimir Vukicevic [:vlad] [:vladv] (needinfo me, slow to respond)]

Right now, the Fennec on Maemo UA looks like:

Mozilla/5.0 (X11; U; Linux armv7l; en-US; rv:1.9.2.4) Gecko/20100515 ...

This is indistinguishable from any other Linux, which I believe was the intent, but the end result is that now sites are treating Gecko + "Linux armv7" = n900.  For example, if you go to facebook, it'll redirect you to m.facebook.com and offer to install a facebook bookmark for your n900.

I suggest that we add "Maemo" to this (so "X11; U; Maemo; Linux armv7l; ..."), and then work on a standard that allows UAs to tell web sites whether they want a mobile optimized version of a site or not.  Something as simple as a header could do it.

Comment 1

[Matt Brubeck (:mbrubeck)]

Attachment: patch

This changes the OS/CPU string from "Linux armv7l" to "Maemo Linux armv7l" on N900 (and similarly for other Maemo devices).

Note that this will not break existing sites that look for the exact string "Linux armv7l".

Comment 2

[Stuart Parmenter]

We would need to follow https://developer.mozilla.org/en/User_Agent_Strings_Reference -- we could possibly change Linux to Maemo4 or 5 or whatever

Comment 3

[Matt Brubeck (:mbrubeck)]

Yeah, if we want to change the OS/CPU string, then we need to update that spec to make Maemo a separate case from other "Unix" operating systems.

Another possibility is to use "Maemo" instead of "X11" as the Platform string.  X11 is technically accurate but not very useful as a platform identifier for Maemo.  (This is also where we put "Android" in our UA string.)

Or we could include another VendorProductToken like "Maemo/5" in addition to "Fennec/2.0a1pre" (without any spec changes).

Comment 4

[Vladimir Vukicevic [:vlad] [:vladv] (needinfo me, slow to respond)]

Comment on attachment 450474 [details] [diff] [review]
patch

er, this patch a has a lot of non-UA change stuff in it?

Comment 5

[Matt Brubeck (:mbrubeck)]

Attachment: patch v2

(In reply to comment #4)
> er, this patch a has a lot of non-UA change stuff in it?

Oops.  Here's the right patch.

Comment 6

[Vladimir Vukicevic [:vlad] [:vladv] (needinfo me, slow to respond)]

Comment on attachment 450496 [details] [diff] [review]
patch v2

Sure, that UA change looks fine to me, though I raelize now that it's in netwerk code.. you may want to get someone else to ok it as well, maybe dwitte or smoeone who's been looking at UA stuff?  Dunno.

Comment 7

[[e-mail]]

Comment on attachment 450496 [details] [diff] [review]
patch v2

Let's replace X11 with Maemo instead: "Maemo; Linux ..."

r=me with that.

Comment 8

[Matt Brubeck (:mbrubeck)]

Attachment: patch v3: Replace "X11" with "Maemo"

Updated as requested in comment 7, carrying the review.

Comment 9

[Matt Brubeck (:mbrubeck)]

Pushed: http://hg.mozilla.org/mozilla-central/rev/b2bc1dd449b7

Comment 10

[Aakash Desai [:aakashd]]

verified FIXED on build:

Mozilla/5.0 (X11; U; Linux armv71; Nokia N900; en-US; rv:2.0b5pre) Gecko/20100826 Namoroka/4.0b5pre Fennec/2.0a1pre

Comment 11

[Daniel C]

(In reply to comment #10)
> verified FIXED on build:
> 
> Mozilla/5.0 (X11; U; Linux armv71; Nokia N900; en-US; rv:2.0b5pre)
> Gecko/20100826 Namoroka/4.0b5pre Fennec/2.0a1pre

That doesn't make sense. I thought this bug replaced X11 with Maemo, so how can it be verified in a build with a UA that doesn't show this bug as fixed per the summary?