Re: [sipcore] #27: Functionality of "Supported: histinfo" is not clear

["Worley, Dale R (Dale)" <[e-mail]>]

________________________________________
From: [e-mail] [[e-mail]] On Behalf Of Elwell, John [[e-mail]]

My understanding is that the presence of Supported: histinfo triggers the UAS to send back H-I in the response. The issue that Dale raises makes me wonder whether H-I should unconditionally be echoed in the response, since proxies and B2BUAs might find a use for it, even if the UAC does not request it. A UAC that does not support H-I would just ignore the header field anyway.
_______________________________________________

There seem to be two separate "operations" in H-I:

1. a UAS echos in the response the H-I in the request

2. an element updates the H-I in the request, which includes

   2.a. adding an additional hi-entry if the incoming request-URI is not the one specified in the H-I, which itself includes:

      2.a.1. adding an initial H-I with one hi-entry if the request does not contain one

   2.b. inserting an updated H-I into outgoing proxied requests

I've been assuming that both 1 and 2 are triggered by the presence of "Supported: histinfo".  But since the text is unclear, I might be wrong.  Indeed, there is no strong reason that elements that support H-I might not perform both actions even if "Supported: histinfo" is not present, as the presence of H-I in a request or response would not affect the behavior of an element that does not support it.

Dale