Re: [e2md] Dean's Proxy-Shill Version of the Problem statement

CNAM is either way .. if the data is going to be used in a public DNS tree
yes there probably should be pointer that requires authentication. For all
the infrastructure EWNUM systems no..its just a look up function in a closed
trusted tree. 

-----Original Message-----
From: e2md-bounces@ietf.org [mailto:e2md-bounces@ietf.org] On Behalf Of Ray
Bellis
Sent: Friday, May 14, 2010 4:26 AM
To: Dean Willis
Cc: E.164 To MetaData BOF discussion list
Subject: Re: [e2md] Dean's Proxy-Shill Version of the Problem statement

> Just how far away from that approach would we be willing to explore?
> Might we, for example, consider a model where the data is held somewhere
> outside of ENUM, and ENUM just contains a pointer to the data?

That depends on the data.

For Send-N, absolutely not.  The data is trivially small, and getting it
returned as quickly as possible in response to a query for an incomplete
number is a specific design goal.

For the example given during the BOF of a record describing a ring tone,
absolutely!  That should clearly take the form of a *reference* to the
ring-tone data, and not the data itself.

For unused, maybe a hybrid approach - a small amount of data might be
necessary in the actual record (i.e. a "reason code", combined with a
reference to more detailed information).

Ray

_______________________________________________
e2md mailing list
e2md@ietf.org
https://www.ietf.org/mailman/listinfo/e2md