*VidMid-VC Conference Call*
November 19, 2001
*Attendees*
Egon Verharen (chair) - SURFnet
Tyler Johnson - UNC Chapel Hill
Samir Chatterjee - Claremont Graduate U.
Steve Olshansky - Internet2
Art Vandenberg - Georgia State
Michael Gettes - Georgetown
Nadim El-Khoury - UNC Chapel Hill
Albert School - OARnet
John McNair - Tennessee
Jill Gemill - ViDe/Alabama at Birmingham
Ann West - Internet2
Ken Klingenstein - Colorado/Internet2
Ben Chinowsky (scribe) - Internet2
*Discussion*
This was the first VidMid-VC call to be held over H.323. While only about a half-dozen participants were able to join via video, the phone bridge worked perfectly, so that no one was excluded from the discussion for technological reasons. There was general agreement that VidMid-VC should continue using H.323 for its biweekly teleconferences.
After approving the minutes of the November 5 call, the group discussed topics for the upcoming face-to-face meeting at North Carolina. Egon listed the items that should come out of the NC meeting: 1) a framework for the VidMid-VC workplan, and 2) a description of a reference model of the VidMid-VC architecture, including the LDAP object classes it will use. [AI] Tyler will bring a draft reference model of the VidMid-VC architecture to the NC meeting. [AI] Ken will send out a draft list of VidMid-VC deliverables.
The group discussed possible resource-discovery issues for VidMid-VC to
pursue. [AI] Tyler will document his ideas on a DHCP-like capability for
videoconferencing, for discussion at the NC meeting. While there was
considerable interest in this capability, there was also agreement that the
resource-discovery issues that need to be addressed are considerably broader
than this, including:
- finding users and intermediate nodes such as MCUs
- helping an unauthenticated user find their local security domain so they
can authenticate
- finding user attributes for authentication and authorization
- finding out what "bandwidth environment" you're in, and adapting to that
environment as it changes over time
- finding a gatekeeper when it's necessary to switch to a phone bridge in
the course of a videoconference
Jill noted that VidMid-VC will need to verify that the architecture it develops is capable of being expanded to accommodate mobility. Ken asked if there was consensus on pursuing federated administration, and Egon replied that while it seems likely that this approach will work for VidMid-VC, this still needs to be confirmed. Tyler noted that the telcos are moving forward with "building phone numbers into DNS" using ENUM. Tyler also noted that some in the videoconferencing world, while recognizing the need to be able to reach a variety of services via a standard phone number, would like to see this done with LDAP rather than DNS. However, the movement in both the SIP and H.323 communities appears to be toward DNS-based dialing, with LDAP used for authentication after resource discovery is accomplished. Ken suggested that instead of discouraging the use of DNS, VidMid-VC could focus on well-defining how information should make its way into DNS, e.g., via batch rather than continuous updates. Egon noted that DNS is being used to discover gatekeepers, not endpoints, and Jill observed that this parallels email delivery: whoever@wherever.edu takes you to the wherever.edu server, which in turn knows how to get to whoever.
There was a short discussion of SIP-H.323 issues. Tyler noted that Orit Levin will be at the NC meeting, and described her focus as "taking all the telephone-like stuff out of H.323", thereby making H.323 more SIP-like and saving RADVISION from having to develop two separate product lines. Egon raised the issue of protocol transparency -- how important is it for the user to see "click here to reach X" rather than "click here to reach X via H.323" and "click here to reach X via SIP"? Michael asked if transparency should be accomplished via H.323-over-SIP or SIP-over-H.323, or whether the system should figure this out at connect time. No consensus was reached on these issues. Samir predicted that both H.323 and SIP will be around for a long time, and strongly recommended that whatever VidMid-VC develops handle both, as well as the interaction between them.
The group skipped its planned review of the updated scenarios document; [AI] All will review the updated scenarios, and send comments to the list. The group reviewed action items:
[5-November - Karen, Jill, and Ken will try their contacts at Microsoft
about participating in the NC VidMid-VC meeting; all will cc Steve, who will
handle further communication.]
Done; no responses yet.
[5-November - Egon will try to contact David Nash at Polycom about the
NC VidMid-VC meeting; he will cc Steve.]
Done; Steve is waiting to hear back from Polycom.
[5-November - Steve will contact Doug Sicker to discuss folding other
SIP directories work into his.]
Done; Ken and Steve are meeting with Sicker tomorrow.
[5-November - Samir will summarize the state of SIP directories work
for the NC meeting.]
Still to do. Samir noted his agreement with Henning Schulzrinne that there
is no work to be done with object classes for SIP.
[5-November - Egon will write a summary of the VidMid-VC session at
the VIMM.]
Still to do.
[5-November - Egon will review the scenarios with Art and send them
to Ben C. for posting.]
Done.
[5-November - Art will add a discussion of VC/VoD interaction to
the scenarios.]
Done.
[5-November - Samir will find out if and under what conditions
dynamicsoft can make its SIP stack available to VidMid; if that doesn't
work out, he will look into porting his SIP client to a standard open
SIP stack.]
In process; Samir noted that he's made two contacts at dynamicsoft.
[5-November - Ken will get Internet2 IPR staff to discuss SIP licensing
with Columbia.]
Still to do.
[5-November - Egon will ask Ted if he has been able to contact Microsoft
and Polycom/Accord about participating in VidMid-VC.]
Done; Egon hasn't heard back from Ted.
[5-November - Jill will find out if her colleagues who attended TeleCon
made any contacts for VidMid-VC.]
Done; Steve is following up with Jill's contacts.
[5-November - Steve will re-poll the list about using
videoconferencing for future VidMid-VC calls.]
Done; Steve noted that the few additional responses he's gotten
are all in favor.
[5-November - VidMid-VC will hold its next regularly-scheduled call
(Nov. 19) over H.323; Tyler will provide dial-in information and operate
the MCU.]
Done.
[22-October - Samir will send the list a short summary of
approaches to coping with NATs and firewalls.]
Samir is working on this.
Finally, Steve noted that he's trying to find out if he should try to get VBrick (http://www.vbrick.com) involved in VidMid-VC, VidMid-VoD, or both. [AI] Jill and Steve will find out more about VBrick.
*Action Items*
[AI] 19-November - Tyler will bring a draft reference model of the VidMid-VC
architecture to the NC meeting.
[AI] 19-November - Ken will send out a draft list of VidMid-VC deliverables.
[AI] 19-November - Tyler will document his ideas on a DHCP-like capability
for videoconferencing, for discussion at the NC meeting.
[AI] 19-November - All will review the updated scenarios, and send comments
to the list.
[AI] 19-November - Jill and Steve will find out more about VBrick.
[AI] 5-November - Samir will summarize the state of SIP directories work
for the NC meeting.
[AI] 5-November - Egon will write a summary of the VidMid-VC session at
the VIMM.
[AI] 5-November - Samir will find out if and under what conditions
dynamicsoft can make its SIP stack available to VidMid; if that doesn't
work out, he will look into porting his SIP client to a standard open
SIP stack.
[AI] 5-November - Ken will get Internet2 IPR staff to discuss SIP licensing
with Columbia.
[AI] 5-November - Ken will ask Philippe approximately how much money would
be required to integrate H.323 and SIP into VRVS.
[AI] 22-October - Samir will send the list a short summary of
approaches to coping with NATs and firewalls.
[AI] 22-October - Egon will contact OpenH323 about participating in
VidMid-VC.
[AI] 22-October - Tyler will look into Siemens's involvement in
videoconferencing-related technologies, especially SIP.
[AI] 24-September - Ken will send Philippe information on possible
funding for integrating H.323 and SIP into VRVS.