1313506 - Set a user agent

Status: RESOLVED FIXED

(Focus-iOS :: General, defect, P1)

Description

[Brian Nicholson (:bnicholson)]

We should define and use a UA for Focus.

Comment 1

[Stefan Arentz | :st3fan | ⏰ EST | he/him]

I would like to suggest that we do the same thing as we do for Firefox for iOS. This is documented at:

  https://developer.mozilla.org/en-US/docs/Web/HTTP/Headers/User-Agent/Firefox#Firefox_for_iOS


This means we will have:

Mozilla/5.0 (iPod touch; CPU iPhone OS 8_3 like Mac OS X) AppleWebKit/600.1.4 (KHTML, like Gecko) Focus/2.0 Mobile/12F69 Safari/600.1.4

Mozilla/5.0 (iPhone; CPU iPhone OS 8_3 like Mac OS X) AppleWebKit/600.1.4 (KHTML, like Gecko) Focus/2.0 Mobile/12F69 Safari/600.1.4

Mozilla/5.0 (iPad; CPU iPhone OS 8_3 like Mac OS X) AppleWebKit/600.1.4 (KHTML, like Gecko) Focus/2.0 Mobile/12F69 Safari/600.1.4


The UA "template" is taken from the current UIWebView UA, and then the Focus/2.0 part is inserted.

Comment 2

[Stefan Arentz | :st3fan | ⏰ EST | he/him]

Gerv, I think you gave us advice when we were did the same thing for Firefox for iOS. What do you think?

Comment 3

[Gervase Markham [:gerv]]

So the question you want to ask is: what bad things would happen if you just used the standard UIWebView UA (is that the component you are using)? Is that what you want from a compat perspective, or are you adding things or taking away things that might be relevant? Do you expect or want Focus to have a market share that actually shows up in UA-based market share figures?

Gerv

Comment 4

[Stefan Arentz | :st3fan | ⏰ EST | he/him]

Attachment: PR: https://github.com/mozilla-mobile/focus/pull/13

Comment 5

[Stefan Arentz | :st3fan | ⏰ EST | he/him]

We decided to follow the same scheme as Firefox for iOS. Adding a Focus/2.0 to it to be able to identify it in traffic logs. If this feels like the wrong approach we can still make changes.

Comment 6

[Aaron Train [:aaronmt]]

We should probably revert this based on zero testing. Do we feel it really necessary to modify the default? 

Getting people to use this product will be hard enough as is, let alone expecting to see the content they expect to see on the web.

Comment 7

[Gervase Markham [:gerv]]

Well, based on your message, we'd be reverting it based on zero evidence. What problems have been seen? Can you link to bugs?

Gerv

Comment 8

[Justin Crawford [:hoosteeno] [:jcrawford]]

(In reply to Gervase Markham [:gerv] from comment #7)
> Well, based on your message, we'd be reverting it based on zero evidence.
> What problems have been seen? Can you link to bugs?
> 

Here is one problem: StatCounter global browser tracking does not know about Firefox Focus, and so it (sanely, alas) assumes that Focus traffic is Chrome traffic. That means all Focus browsing is working against us on one of our key measures. You can verify this by visiting http://gs.statcounter.com/detect in Focus. 

In order to get "credit" for Focus browsing, it would probably be better for Focus to identify itself as Firefox. That might have the added benefit of obscuring the nature of the browsing, a topic of concern in https://github.com/mozilla-mobile/focus-android/issues/445.

Comment 9

[Barbara Bermes [:barbara] - NI please!]

Tracking Protection blocks StatCounter and HitsLink (the analytics that NetMarketShare uses), so even if we had a separate UA, the only pageloads that would contribute to market share is when the user disables the block.

Comment 10

[Gervase Markham [:gerv]]

(In reply to Justin Crawford [:hoosteeno] [:jcrawford] from comment #8)
> Here is one problem: StatCounter global browser tracking does not know about
> Firefox Focus, and so it (sanely, alas) assumes that Focus traffic is Chrome
> traffic. That means all Focus browsing is working against us on one of our
> key measures. You can verify this by visiting
> http://gs.statcounter.com/detect in Focus. 
> 
> In order to get "credit" for Focus browsing, it would probably be better for
> Focus to identify itself as Firefox. 

I think the most likely result of that is that sites would send us Gecko content instead of Webkit content, which would probably not be a webcompat win.

At the moment, we identify as Chrome but with an extra token saying we are Focus. If StatCounter doesn't notice that, the fix is to ask them to notice.

Gerv