It is much more reasonable to expect the service-provider/enterprise to implement the location conveyance. They'd add the location in their proxies/B2BUAs/ALGs. For example, an enterprise building ALG could add its location before sending the call to an SP.
(Hah! Another obvious idea made un-patentable.) In addition, it's the service provider that will likely perform SIP peering with a PSAP. In the US, that's the typical model in already. Service providers sometimes already know where their users are. Yes, yes, I know this isn't the main point of location-conveyance. But besides all of this: we've got to get the PSAPs capable of reliably using the location provided in the call. The capability to send the location will be far simpler than actually having a PSAP that can accept and use it. On Apr 28, 2007, at 5:19 AM, Jeroen van Bemmel wrote: > > Especially for the use case of emergency calls, would it not be > wise to > select a much more simple approach/syntax, e.g.: > Emergency-Location: lat=x; lon=y The location-by-"reference" (i.e., provide a URI to a PIDF-LO instead of carrying the document itself) should do what you need here. Many UACs already have HTTP servers, which would be compatible with such a system, maintain end-to-end privacy requirements. Or the SUBSCRIBE- NOTIFY version of PIDF-LO retrieval could easily be implemented with a partial multipart-mime implementation, and it could work even through proxies. On Apr 28, 2007, at 5:35 AM, Juha Heinanen wrote: > another reason why it will not get implemented is that sip uas don't > know where they are located. gps does not work well indoors and > mobile > operators at least here have refused to make public coordinates of > their > base stations. Other location-sensing technology may come along, now that there are good reasons for it. Mark R. Lindsey | ECG | +1-229-316-0013 | [EMAIL PROTECTED] _______________________________________________ Sip-implementors mailing list [email protected] https://lists.cs.columbia.edu/cucslists/listinfo/sip-implementors
