Re: [Simple] comments on prescaps - extension elements

[Paul Kyzivat <pkyzivat at cisco.com>]

I have other comments on this, but I want to single this one as 
especially important.

	Thanks,
	Paul

Jonathan Rosenberg wrote:
> > 3.23 Extension elements
> >
> >    This section defines how extension features present in "Indicating
> >    User Agent Capabilities in the Session Initiation Protocol (SIP)"
> >    [6]    can be used in this extension.
>
> I know I spoke in favor of this approach in the past, but I never 
> guarantee the self consistency of my argumetns over time, and I now 
> think that this is not a good idea.
>
> The reasons are:
>
>   1. XML has a mechanism for extensibility; this defines a separate one
>   2. an element could appear as an extension as you have defined it, or, 
> if later we do define a matching set of real XML extensions to carry it, 
> it could appear there. You dont want to have two valid ways of carrying 
> the data
>   3. there is an argument to be made that the PA should not generally be 
> conveying infomraiton it doesnt understand, and that is the case here
>
> As such, I would recommend dropping this entire section.
I disagree!

The extension mechanism in callee-caps has means that don't require any 
registration at all. This is very good for cases when there are features 
that are very transient. For instance, it is possible to define a 
feature tag for something like frobitz-sales, or widget-help. And then I 
can use callerprefs to select an endpoint with the appropriate 
capability. This allows end users to take advantage of callerprefs for 
features that matter to them. (Assuming sufficiently smart devices.)

When using presence, I would like the same to hold true. If it is 
necessary to define a new xml schema before I can announce a capability 
for widget-help in my presence then nothing this dynamic can be a feature.

If there is a problem, it is that the draft forbids use of the extension 
mechanism when there is specific schema for the same purpose.

	Paul


_______________________________________________
Simple mailing list
Simple@ietf.org
https://www1.ietf.org/mailman/listinfo/simple