Dear All,
Following one of the inter-gk conference calls, I picked up an action to find out what is happening in the IETF E.164 area. Although I wasn't actually at the BOF this is what I have managed to pick up from a colleague.
Firstly, a working group is to be created, which is likely to be (will be?) chaired by Scott Petrack.
The details of the scope have not been fully fleshed out yet. The input to the process will be an E.164/telephone number. Whether local short-cut codes and private networks are supported is still to be decided. The output has not been decided, but could range from a simple IP address, to something that may include alternate E.164 number, to something that would list a full set of remote endpoint capabilities.
A number of proposals were discussed most of which had the theme of: convert E.164 number/telephone number to domain name, then convert domain name to IP address. The latter stage may involve something like the gateway location work.
One proposal (Falstrom(?)) is to use DNS NAPTR records to convert E.164 numbers to domain names. To do this each digit would be separated by a dot, the sequence reversed, and then the domain name e164.int appended to it. e.g.:
+44-1473-646436 -> 6.3.4.6.4.6.3.7.4.1.4.4.e164.int
This could match multiple NAPTR records. The NAPTR records map to a domain name and include with them what service is associated with that domain name. e.g. The above address could map to:
IN NAPTR 100 10 "s" "h323call+N2R" tele2.se IN NAPTR 100 10 "s" "sipcall+N2R" tele3.se
The first one indicates that the h323call+N2R service is at address tele2.se, and that there is a SRV record (indicated by the "s") for it. The other values indicate priorities. (Not sure what the +N2R indicates)
A set of these queries would be returned to the querier, and it could select the service it was interested in and would do a DNS SRV look-up to convert this to an IP address.
One issue with this is that it has a single top-level domain. Huitema suggested that multiple organisations should be able to provide the top level mappings, which he called a conflicting claims method.
Other proposals had variations on this theme, some of which included references to other databases such as X.500, LDAP and so on (Brown).
This seems quite sensible and is well worth keeping an eye. Apparently not much has happened on the mailing list since, but this could be because Scott hasn't produced the minutes yet!!!
Hope this helps,
Pete ================================= Pete Cordell BT Labs E-Mail: pete.cordell@bt-sys.bt.co.uk Tel: +44 1473 646436 Fax: +44 1473 645499 =================================