503942 - Implement Geolocation Addresses

Status: RESOLVED FIXED

(Core :: DOM: Geolocation, defect)

Description

[Doug Turner (:dougt)]

we should implement geolocation addresses based on the Gear's Address object.

Comment 1

[Doug Turner (:dougt)]

Attachment: patch v.1

Comment 2

[Mike Beltzner [:beltzner, not reading bugmail]]

Marking wanted, but don't know if we'd block. What does this get us?

Comment 3

[Doug Turner (:dougt)]

in addition to getting a lat/long, you can also get this:

where |a| is the objection returned to the geolocation APIs:


a.address.streetNumber
a.address.street 
a.address.premises 
a.address.city
a.address.county 
a.address.region
a.address.county
a.address.countryCode
a.address.postalCode

Comment 4

[Johnny Stenback (:jst)]

Is this in a spec somewhere, or is this still being ironed out etc?

Comment 5

[Doug Turner (:dougt)]

This is still being ironed out.  The last reasonable format for a simplified address as basically the above plus an extra field called "additionalInformation".  I am leaving this field off for now.  If there this does become a recommendation, we can append the field.

fwiw, google chrome and android use the above address object.

Comment 6

[Doug Turner (:dougt)]

http://hg.mozilla.org/mozilla-central/rev/ebea850caba8

Comment 7

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

Looks like this may be causing some mochitest failures on mozilla-central (on Mac OS X, so far).

Comment 8

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

I backed this out for causing mochitest timeouts on all platforms:

http://hg.mozilla.org/mozilla-central/rev/191ef763e892
http://hg.mozilla.org/mozilla-central/rev/df6b05ee9830

Comment 9

[Jim Jeffery not reading bug-mail 1/2/11]

Should be re-opened?

Comment 10

[Doug Turner (:dougt)]

the failure was caused by our test cases not providing an address response.  GLS and other providers should provide this.  I tweaked our code to test the response before assuming it is there.

In bug 482260, we will update the testing code so that we can test with and without a provider sending an address location.

Comment 11

[Doug Turner (:dougt)]

Attachment: patch v.2

Comment 12

[Doug Turner (:dougt)]

http://hg.mozilla.org/mozilla-central/rev/f248aeeaaa3e

Comment 13

[Samuel Sidler (old account; do not CC)]

Doug: I believe you want this in Firefox 3.5. If that's right, please request approval1.9.1.2 on the patch (after checking that it applies).

(Note: You have some approval1.9.1.1 requests out. We'll be triaging those, no need to change them to approval1.9.1.2.)

Comment 14

[Samuel Sidler (old account; do not CC)]

Comment on attachment 388300 [details] [diff] [review]
patch v.1

Not for 1.9.1.2.

Comment 15

[Daniel Veditz [:dveditz]]

Comment on attachment 388300 [details] [diff] [review]
patch v.1

New features/interfaces are generally not appropriate for the stable branch. 1.9.1 approval denied.

Comment 16

[Eric Shepherd [:sheppy]]

Now documented for Gecko 1.9.2:

https://developer.mozilla.org/en/XPCOM_Interface_Reference/NsIDOMGeoPosition
https://developer.mozilla.org/en/nsIDOMGeoPositionAddress

(the latter will move soon, got put in the wrong place, but there'll be a redirect).