POST geolocation command doesn't obey the spec about chromedriver HOT 13 CLOSED

I'll investigate and see how much work would be involved for this. The set 
method might be tricky.

It looks like it shouldn't be too bad. The command seems a bit strange to me 
however. Why does a user ever want to get the geolocation? Is it expected that 
they will always call set before get? If they call get before set, do I have to 
return the real geolocation position or just some phony position?

Yeah, I don't see why the GET is useful either, but that's how it was designed 
in the Android driver. The main use case is obviously the ability to override 
what a webapp gets back when it asks the browser for location. 

Any update on this one ?

Is this already implemented ? How can we set/get the geolocation ?

This is not yet implemented. I started a patch a while back but haven't gotten 
back to it yet.

Will be in ChromeDriver v20 release

Very cool to finally have this in a desktop browser.

However, it looks like the current implementation expects a JSON object like 
this:

  {"latitude":40.714353,"longitude":-74.005973,"altitude":0.056747}

but the wire protocol spec (and the Android/iPhone implementations) says it 
should be:

  {"location":{"latitude":40.714353,"longitude":-74.005973,"altitude":0.056747}}

Whoops, I didn't notice that it was wrapped in a location dictionary. I'll fix 
that. I  need to run the Java tests for these.