home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Internet Info 1997 December
/
Internet_Info_CD-ROM_Walnut_Creek_December_1997.iso
/
ietf
/
97apr
/
pint-minutes-97apr.txt
< prev
next >
Wrap
Text File
|
1997-05-22
|
6KB
|
125 lines
Editor's Note: These minutes have not been edited.
Minutes of the PSTN/Internet Interworking (pint) BOF
Meeting on April 9, 1997
Reported by I. Faynberg and H. Lu
(Bell Labs, Lucent Technologies)
faynberg@bell-labs.com
huilan.lu@bell-labs.com
The PSTN/Internet Interworking (pint) BOF meeting took place from
21:00 to 22:00 on Wednesday, April 9. The meeting was chaired by Igor
Faynberg (Bell Labs/Lucent Technologies) and co-chaired by Hui-Lan Lu
(same organization) who also took minutes. The meeting was attended by
about 90 people.
The goal of the meeting was to discuss two internet drafts:
1) draft-faynberg-telephone-sw-net-00.txt - describing an Internet/PSTN
connection arrangement through which combined Web and traditional
telephony services can be delivered.
2) draft-zigmond-media-url-00.txt - describing an URL scheme for telephony.
The discussion of 1) was led by Murali Krishnaswamy (Bell Labs), whose
presentation summarized the relevant internet draft. The questions that
were asked centered around the following issues:
1) The services supported with the proposed arrangement and what
types of servers would be connected to the PSTN;
Answer: A combination of traditional internet and telephony
services. A specific example is Web-based Yellow Pages services,
for which a Web server would be connected to the PSTN.
2) Why cannot the Switch/Computer Application Interface (SCAI) platform
be used?
Answer: It can. In fact, it is very similar to IN (the proposed
means of accessing the network), and the protocol used for
accessing IN could be used for accessing the SCAI architecture.
The IN arrangement, however, is more powerful because IN controls
the whole network, while the SCAI application controls a single
switch.
3) Why cannot the client place the call?
Answer: It can, of course. But, unless the client is to pay
for the call--which is contrary to the assumption about the
types of services, the IN will have to take care of it (e.g., the
800 number) sooner or later. It is much better to take care of
it sooner--and right in the middle of the network rather than
at the end point. It also allows the network to optimize its
resources.
4) How does this proposal relate to packet voice?
Answer: It does not.
5) But should we not work in this group on the interconnections between
the IP routers and PSTN in order to deliver packet voice across
the Internet/PSTN boundaries?
Answer: Perhaps we should. But there is no group yet and no
draft on the table today, and therefore we cannot discuss this
kind of interconnection right now.
6) How can the call be placed to the user if user's line is already
occupied [by the Internet connection]?
Answer: The proposal is meaningful only with the use of an ISDN
connection, or voice&data modem, or a digital subscriber line,
or some other arrangement (e.g., two voice lines), which leaves
place for a voice call.
7) So, what are the protocols to be standardized?
Answer: There are two protocols. [As the proposal says:] The
TCP/IP-based protocol between the server and Service Node (or
Service Control Point), and the SMNP-based protocol between the
router and the Service Management System.
At the end of a half-an-hour discussion, a large number of
participants expressed support for carrying out the
work in the IETF, but there was also a sizable opposition to the proposal.
It was agreed to continue the discussion will continue via e-mail. The
self-subscribing e-mail discussion list was to be created by I. Faynberg.
and announced by the IETF. [At the time of publication of these notes, the
list has been created: pint@lists.research.bell-labs.com.]
The discussion of 2) was led by Dan Zigmond (Wink Technologies). Dan
has introduced his proposal for the URL format for telephony. The following
questions were asked:
1) Is the addressing scheme relative or absolute?
Answer: According to the design it could be either. [A suggestion
from the audience was that since the proposed addressing
follows the design of the "mail-to" addressing, it should
support be absolute.]
2) Where will this type of URL be used?
Answer: Click-to-dial, automatic billing, for example.
3) It should be useful to specify--within the proposed scheme--the
attributes for the post-dialing delay, PIN, time to call, and
the type of the phone (e.g., cellular).
Answer: Yes, that could be done.
4) Why not simply use the phone number as is? Why URL?
Answer: To be able to process things in a preset way (i.e.,
have the attributes as discussed in 3).
The meeting expressed strong support for carrying out the work. It was felt
that the proposed work item was too small to warrant a separate working group.
The discussion leading to a standard will continue via e-mail. The
self-subscribing e-mail discussion list will be created by D. Zigmond
(please send requests to dan.zigmond@wink.com) and announced by the IETF.