Newbie questions are posted on [EMAIL PROTECTED] Do NOT post them on the main sip list. Also, as a general rule, please consult the faq and the archives since many of your questions have been answered countless times before. Some comments inline: > -----Original Message----- > From: Varadarajan Srinivasan [mailto:[EMAIL PROTECTED]] > Sent: Monday, June 04, 2001 11:21 PM > To: [EMAIL PROTECTED] > Subject: [SIP] SIP Queries > > > Hi All, > > I have some basic queries regarding RFC2543. I have used pdf > version of the draft bis02, > and the page nos are for the same. > > 1) Is it not the Route header and Via contains redudant > information? > > Via is needed only to set route or just to send responses in the > same path as that of the > request.After route is > set, I think proxies can take this info from route > header and via > contains redundant > information. Via is always used to forward responses. Route is always used to determine where to forward requests. > > 2) Are Max_Forwards header and ttl parameter, mean > similar things > ? Can we take > one out? No. Max-Forwards is a sip level hop-count through SIP proxies. ttl is an IP layer router hop count for multicast. > > 3) In chapter 16, RFC2543[100], in example, Why Via headers of > the response 180 > Ringing to the > Invite request (in the previous page) are not the same?? Its an error. Its already been corrected in bis-03. > > 4) In chapter 16.5, [106](Page 106) of RFC2543, under the SIP > message 487 Request > Terminated, it is written as the response should terminate at > Proxy. I think it is Client. No, it is the proxy. The proxy is a forking proxy and sends the best response(s) upstream, which do not include the 487. > > 5) Why statefull proxies require TCP? They don't. Proxies that do TCP have to be stateful, which is not the same statement. > > 6) In Appendix A, in minimal implementation, [110] of > RFC2543, for > Clients, > it is said Contact header is not a minimum requirement. Without > that how do you > register? SIP can work without registrations. Anyway, bis-03 mandates Contact for a minimal implementation. > Why BYE is not a part of minimum requirement?How will you > terminate a call then? There are some sessions which don't need a BYE. A basic client does require BYE though. > > 7) In Appendix A, in minimal implementation, for > Servers, OPTIONS > is categorised as > Minimal requirement. > I dont think so as it is only a sophistication. It is correct as written. > > 8) > > INVITE 11:02:32 > UAC/UAS 1 ------------------------------ > UAC/UAS 2 > > > 180 > RINGING 11:02:34 > < > ----------------------------- > > > CANCEL 11:02:36 > > ----------------------------- > > > > 200 OK for INVITE > < > ----------------------------- > > A user 1 invites another user 2. It gets the 180 > ringing and later > wants to cancel the > invitation by sending > cancel method. Before, the Cancel method reaches user > 2, it picks > up the call and > hence sends a 200 OK. > Now, at this situation, what to do next? UAS sends 200 OK for CANCEL. UAC sends ACK for 200 response to INVITE. Call is established. caller may hangup call with BYE if it wants. > > 9) I have attempted certain clarrifications on RFC2543 for > Appendix I. I just want the > group to check whether mine is correct or open the > problem to the > group. > > i) In Appendix I, [127] of RFC2543, II paragraph, the author > specify that the request > URI for > Register is just the domain name wihtout @ but a format > of request > line demands that > it is a URL which requires @ in its syntax. > The person who raised this query, is suggesting that To: header > field can be processed > instead of Request URI by > REGISTRAR and Request URI can be omitted in the > REGISTER method .. > I think > this is right. No. > > ii) In another clarrification in Appendix I, the last but one > query before Appendix J, > the person who raised the query is WRONG. Table 5 does contain > www-Authenticate as a response as well > as Request header. You should not make such strong statements about who is wrong, and who is not, since you are wrong here. The table did indicate this as a response header, but no text existed to describe its usage as such. THis comment is talking about that text, not the table. In any case, response authorization using www-authenticate has been deprecated in favor of rfc2617 mechanisms. -Jonathan R. --- Jonathan D. Rosenberg, Ph.D. 72 Eagle Rock Ave. Chief Scientist First Floor dynamicsoft East Hanover, NJ 07936 [EMAIL PROTECTED] FAX: (973) 952-5050 http://www.jdrosen.net PHONE: (973) 952-5000 http://www.dynamicsoft.com _______________________________________________ Sip-implementors mailing list [EMAIL PROTECTED] http://lists.cs.columbia.edu/mailman/listinfo/sip-implementors
