Minutes of the PSTN/Internet Interworking (pint) WG Meeting on August 27, 1998 Reported by: S. Bellovin (AT&T Labs-Research) smb@research.att.com and I. Faynberg (Bell Labs, Lucent Technologies) faynberg@bell-labs.com, based on the notes taken by Hui-Lan Lu (Bell Labs), Lawrence Conroy (Roke Manor Research/Siemens), Jonathan Rosenberg (Bell Labs), and Lev Slutsman (AT&T Labs). The PSTN/Internet Interworking (PINT) WG meeting took place from 15:30 to 17:30 on Thursday, August 27. The meeting was chaired by Steve Bellovin (ATT Labs-Research) and Igor Faynberg (Bell Labs/Lucent Technologies). Hui-Lan Lu, Lawrence Conroy, Jonathan Rosenberg, and Lev Slutsman, to whom many thanks, took notes and helped to prepare this report. The proposed agenda of the meeting was as follows: 15:30-15:35 Chairs Introduction and Agenda Bashing 15:35-16:05 Scott Petrack Issues relevant to the PINT Protocol Profile: Additions to SIP & SDP Parameters - pint_context, usage, Forward & backward compatibility Requirements for carriage via SIP infrastructure 16:05-16:35 Lawrence Conroy Issues relevant to Profile Context & Use: Configurations Registration behavior Proxy behavior Practicality of service naming Privacy, SIP, and service requests Open Issues: * Nuisance Calls * Malicious Call Interception * Assuring Equal Access * Anything else? 16:35-17:05 Hui-Lan Lu, The status of the counterpart work in the ITU-T Greg Ratta 17:05-17:30 Chairs Where we go from here: New PINT time table MIB work In the introduction, the chairs summarized the status of the PINT work: 1) completion and approval by the IESG of the PINT informational RFC, which has been presently undergoing the editorial changes requested by the RFC editor; 2) completion of the first draft of the PINT Protocol; establishment of "liaison" (or, more precisely, mutual representation with the ITU-T Study Group 11 Working Party 4 (re: Intelligent Network). It was noted that some confusion was caused by the SS7 BOF preparatory discussion, which was cross-posted to several lists, including the PINT list. Neither the IP telephony requirements nor suggestions concerning the future PINT work were among the issues to be discussed within the scope of the present PINT charter. Concerning the agenda, the chairs informed the meeting that no TINA representative was able to come to the IETF; the extra 15 min. were appended to the last item of the agenda. It was also noted that the the official representative to PINT from WP 4/11, Prof. Mi of China PTT, could not come to the 42nd IETF meeting. Hui-Lan Lu (who represents PINT in SG 11) was asked to respond to questions regarding the SG 11 side of the story. (The official document was submitted to PINT by Greg Ratta on behalf of SG 11.) The agenda bashing resulted in two proposals: 1) Henning Schulzrinne (Columbia University) requested a 15 min. slot to address the SDP issues; the request was granted. 2) Fred Burg (AT&T) requested that the use of H.323 for PINT be addressed; however, no one wished to make a presentation on the subject. There was a comment suggesting a need to consider "interworking between SIP and H.323" in the future work. On request of AD, the discussion was postponed until the last item on the agenda (future work), but, again, no one brought the subject up then. The discussion of Scott Petrack's presentation raised the following questions: a) Use of the country-specific freephone numbers (can, for example, '1-800-flowers' be specified as a parameter?). This issue needs to be resolved in the mailing list discussions. b) What if a wrong SDP parameter (i.e., 'wave') is specified in the body of the message? How to prevent that? The proposal was made not to deal with that issue in a standard, but leave that to implementations. A reference was made to the version 9 of the SIP I-D for the use of warning headers inside responses. It was decided to follow up on the mailing list. c) Can the status information (specified in the new draft of SIP) be provided both synchronously (to the request) and asynchronously? The answer: Yes. In fact, the duration for which the event notification is provided can be specified. d) Concerning the reference to the Draft ITU-T Recommendation Q.1241, is it used for normative or informative reference (the normative reference would be inappropriate)? The answer: the reference is only informative. e) How to deal with the "Invite" redirect? It was noted that this problem is synergetic with that of the IPTEL Gateway location and therefore should be left out of the PINT Profile. (Another synergy with IPTEL was in the use of the Request message.) An issue was also raised whether specification of the fax format is relevant to PINT. The response was: Yes, we need to specify the format because we are transferring the source data with the request. It was emphasized that this has nothing to do with Internet faxing--the fax delivery is carried purely by the PSTN. Scott will provide the clarification text in the next version of the draft. Another question was on relation of the PINT Server and Executive System to the "Service Node" of the PINT Charter. Scott Bradner noted that at the time the charter was written that term was used, but now it is clear that was not the best choice. I. Faynberg seconded this observation, mentioning that the ITU-T SG 11 came to the same conclusion, and noted that the relation of the Executive System to Service Node and other IN physical entities will be explained in Hui-Lan's presentation. The discussion of Lawrence Conroy's presentation raised the following questions: a) Are conferencing and multicasting supported in PINT? The answer: Not in the present phase. Only requests for two party calls are supported by the PINT Protocol. b) Can same services be delivered by IP telephony? The answer: It is likely, but this is not the PINT WG concern. The discussion of Henning Schulzrinne presentation (which proposed simplified session description) raised an issue of support (with the proposed description mechanisms) of invitations scheduled for some time ahead. This issue was deferred for the follow-up on-line discussion. The discussion of Hui-Lan's presentation raised no issues. The architecture she presented was used to point out the entities (the SCF, SDF, SRF, and the SSF/CCF), whose combination maps into the Service Node mentioned in the PINT charter; it was noted that other combinations (e.g., SCF and SDF, for the Service Control Point) were also possible. During the discussion of the reviewed milestones, Steve Bellovin pointed out that the completion of the work on the PINT MIB is necessary for advancing the status of the future PINT Protocol RFC to Draft Standard. A suggestion was made that this work be synergetic with that of the SIP MIB (the PINT extensions would make the PINT MIB). The suggestion was supported; the final decision was deferred to the "MIB doctors" who will pick up the work. At the end of the meeting, it was suggested that the future PINT meetings will dedicate more time to discussions (and less time to "tutorials"). The suggestion was supported by the chairs. It was also noted that the decisions are actually made in on-line discussions, not at the meetings.