Re: [Sipping] Re: draft-elwell-sipping-service-retargeting-00.txt

[Dean Willis <[e-mail]>]

On Wed, 2005-10-19 at 09:22 -0400, David R Oran wrote:

> P.S. In terms of the debate over "telling you what happened so you  
> can figure out what the right thing to do is" versus "telling the SIP  
> infrastructure exactly what to do", I have consistently been in the  
> former camp. In fact, that reasoning was one of the things that  
> motivated both the Reason header and the History-Info work. I have a  
> strong preference to avoid putting service directives into URIs  
> except as a last resort, and only then when you know EXACTLY what you  
> want a server to do and can ensure that the request in fact gets  
> routed to that server.
> 

Isn't that sort of like asking a web server to figure out which page you
want to see at by looking at your browser's history list? You must write
some interesting PHP-CGI scripts . . .

Seriously, this fundamental divide is much of what is causing muddling
in SIP. If we had "logic based on backward knowledge only", as Dave is
suggesting, then we'd have a relatively clean (but annoying to me)
model. If we had "logic based on forward knowledge only" (as per RFC
3087), then we'd have a relatively clean (but annoying to Dave) model.
As is, we have a complex model that's annoying everybody.

Once again, I'll note that web servers don't make decisions about which
web page to return based on HTTP-level information about which proxies
have previously seen a request and why they chose to pass it on.

--
Dean



_______________________________________________
Sipping mailing list  https://www1.ietf.org/mailman/listinfo/sipping
This list is for NEW development of the application of SIP
Use [e-mail] for questions on current sip
Use [e-mail] for new developments of core SIP