Re: What about Reverse Geocoding?

Given that it's much easier to come up with a civic address 
representation than to come up with guidance on reverse geocoding, might 
we consider addressing this in v1?

The structure could essentially be a copy/paste from RFC 5139 (proposed 
text below; or choose some reasonable subset of it).  API calls, or 
integration with the Position structure, would be easy to add.

--Richard


-----BEGIN proposed interface-----
interface Civic {
	readonly attribute DOMString country;
	readonly attribute DOMString A1;
	readonly attribute DOMString A2;
	readonly attribute DOMString A3;
	readonly attribute DOMString A4;
	readonly attribute DOMString A5;
	readonly attribute DOMString A6;
	readonly attribute DOMString PRM;
	readonly attribute DOMString PRD;
	readonly attribute DOMString RD;
	readonly attribute DOMString STS;
	readonly attribute DOMString POD;
	readonly attribute DOMString POM;
	readonly attribute DOMString RDSEC;
	readonly attribute DOMString RDBR;
	readonly attribute DOMString RDSUBBR;
	readonly attribute DOMString HNO;
	readonly attribute DOMString HNS;
	readonly attribute DOMString LMK;
	readonly attribute DOMString LOC;
	readonly attribute DOMString FLR;
	readonly attribute DOMString NAM;
	readonly attribute DOMString PC;
	readonly attribute DOMString BLD;
	readonly attribute DOMString UNIT;
	readonly attribute DOMString ROOM;
	readonly attribute DOMString SEAT;
	readonly attribute DOMString PLC;
	readonly attribute DOMString PCN;
	readonly attribute DOMString POBOX;
	readonly attribute DOMString ADDCODE;
}

// For example...
var addr = navigator.geolocation.lastPosition.civic;
// Print a US address
var addrStr = ""
addrStr += addr.HNO +" "+ addr.RD +" "+ addr.STS +"\n";
addrStr += addr.A3+ ", "+ addr.A1 +" "+ addr.PC +"\n";
-----END proposed interface-----


Doug Turner wrote:
> 
> 
> On Nov 11, 2008, at 9:53 AM, Andrei Popescu wrote:
> 
>> Hi Martin,
>>
>> On Mon, Nov 10, 2008 at 9:51 PM, Thomson, Martin
>> <Martin.Thomson@andrew.com> wrote:
>>> Hi Andrei,
>>>
>>> I think that you've got it, but rather than could, I'm saying 
>>> "should".  The API shouldn't assume that the two have the same 
>>> source, but it doesn't need to be.
>>>
>>> Rather than saying that we should revisit reverse geocoding in 
>>> version 2, you should be saying that we should revisit _civic 
>>> addresses_ in version 2.
>>
>> Ok, fair enough. I suspect we'll see more demand for this once this
>> spec opens up to a larger audience, so I believe it'll be one of the
>> top discussion items for v2.
>>
>> All the best,
>> Andrei
> 
> 
> 
> +1
> 
> 

Received on Tuesday, 11 November 2008 18:32:05 UTC