Hello All,
I think that "busy" state (as oppesed to "free" state) is not something
absolute. It means "out of resources" but different calls have different
resource requirements. A user might be busy to answer a voice call, but fax
could be received automatically. A private network with narrowband link
might accept just few video calls, yet might more voice calls.
Therefore, we should place enough information, preferrably the entire SDP,
into subscription (SUBSCRIBE request).
Suggestion: 9.3: add,
"The SUBSCRIBE body MUST contain the original session description, to
provide information about resource requirements."
This could be reproduced in Section 5, somewhat like "There are effectively
several queues for calls with different resource requirements".

12.2 Why is it application/call-completion, not text/call-completion?
Shouldn't charset appear explicitly? I'm afraid of different encoding
problems... This MIME type is also more compact.

What is about interoperability with RFC 4235? There is none, but IMO this
should be explicitly stated somewhere. E.g. in 6.1 or 6.2, "Caller MUST try
this procedure first, and if the "call-completion" event package is
unsupported, it may attempt complete the call as in RFC 4235".

Sorry for being away too long (and cross-posting, if any).

Regards,
Anton Tveretin

_______________________________________________
BLISS mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/bliss

Reply via email to