Geolocation API Level 2 LC comment

Dear Geolocation WG,

Several years ago, there was some discussion on this list about aligning the address format in the Geolocation spec with the existing civic address format provided in RFC 5139, or appending an optional object that could contain an RFC 5139-compatible address when required [1]. As discussed at the time, one of the advantages of using the RFC 5139 format is that it accommodates international addresses, whereas there is no obvious mapping between certain countries' address formats and the format currently used in the Geolocation spec. China and Japan are the major examples of countries with "non-Western" address structures, but even for a country as "Western" as Austria, the mapping may not be straightforward [2].

In the intervening time, there has been some discussion about the alignment of the Geolocation address format with formats in other specs, namely the Contacts API [3], although we understand that the decision has been taken not to align them at this time. The Geolocation format also differs from the OASIS xAL format [4], which is the basis for at least one existing mobile addressing API [5].

Given our experience as co-chairs of the GEOPRIV working group in the IETF, we would suggest that the Geolocation API Level 2 not make use of an address format that differs from existing formats and that does not easily accommodate addresses from large parts of the world. Concretely, we would ask that the Geolocation WG consider aligning their format with at least one existing format (e.g., RFC 5139) or adding an optional object that can capture full RFC 5139 semantics.

Richard Barnes and Alissa Cooper


Received on Wednesday, 11 January 2012 17:15:52 UTC