[Geowanking] is this a repeat? IETF IM geo protocols

[Anthony Townsend]

sorry if this is a repeat. dont remember seeing it

-----

http://www.instantmessagingplanet.com/enterprise/article.php/3305451

ETF Closes in on Linking Geographic Info, Presence
January 28, 2004
By Christopher Saunders

Instant messaging brought "presence" -- the ability to tell when others 
are available for chat -- to the desktop. Now, the concept could be on 
the cusp of another, quiet evolution: incorporating location 
information.

Presence as a source of users' status information has been maturing 
over the past few years. It has grown increasingly granular -- moving 
from the terse "Away" to more informative descriptions like "On the 
Phone" or "Off-site with client; back at 2." Groupware, Web 
conferencing and telephony applications have also begun incorporating 
presence information, broadening its impact.

Now, figures in the Internet communications community are working to 
take presence to the next level by creating a framework for merging 
users' location data into their presence information. That's long been 
viewed as a logical add-on to the basic availability data now available 
in most implementations of presence.

But there are important considerations to take into account before 
simply merging the data. Access to users' geographic information needs 
to be subject to user control, much like presence is handled in most 
consumer instant messaging clients -- which generally enable users to 
hide their availability status from certain classes of fellow users, 
such as unknown contacts. Otherwise, everyone on a network could have 
unrestricted knowledge of others' whereabouts without any form of 
authorization.

Within the Internet Engineering Task Force, the Geographic 
Location/Privacy Working Group (also known as GEOPRIV) has taken up the 
task of walking the line between establishing a means of disseminating 
geographic data that is subject to the same sorts of privacy controls 
as presence is today.

GEOPRIV is close to finalizing on a recommendation -- a Request for 
Comments, in IETF parlance -- for just such a system. That draft 
recommendation, authored by Neustar's Jon Peterson and known officially 
as "A Presence-based GEOPRIV Location Object Format," is actually based 
on earlier work done in formulating the basic requirements for presence 
data -- the Presence Information Data Format (PIDF).

"Historically, in the IETF, there was this big IM and presence war -- I 
guess it isn't officially over, yet," said Peterson, who also serves on 
the Internet Engineering Steering Group, the IETF's standards 
leadership body. "One of the fallouts of that was the IMPP Working 
Group ... One of the things it produced was a common core set of tools 
to be used by IM protocols in the ITEF to ensure interoperability ... 
that was PIDF. And this other effort was started on geographic data."

Eventually, however, individuals involved with both IM and with 
geographic data realized that much of the same privacy and 
authorization controls had to be applied to each.

"More and more, we realized that the GEOPRIV problem was really the 
presence problem -- the mechanisms of subscriptions, tracking, 
receiving periodic updates over time, and authentication, who you lie 
to and who you tell the truth to ... All these arguments look the 
same," Peterson said. "Consequently, we just revised the presence work 
... The draft is an extension to the PIDF document that essentially 
adds a new element that permits one to carry geographical information 
with presence."

Initially, a few early drafts and informational documents were created 
by the group. But little work progressed until this month saw the first 
concerted effort to gain RFC status for GEOPRIV recommendations -- 
which would effectively preserve Peterson's draft as a standard.

The latest effort doesn't aim to hammer out a standard for geographic 
information itself. Rather, it's based on current geographic data 
standards, and focuses instead on encapsulating location information 
within presence data, and applying the same sorts of user preferences. 
Geography Markup Language (GML) is the expected location format over 
which the GEOPRIV draft's specifications will be applied.

"There is related work out there, tons of it in the GEOPRIV working 
group for providing more specific policy tools and language ... 
andOpenGIS (Geographic Information Systems) and the GML 3.0 spec seem 
adequate for expressing simple and extremely complex coordinate space," 
Peterson said. "Tons of people work on OpenGIS. We're not experts on 
that. We're experts on things like security and privacy ... At its 
core, [the new draft] is pretty format-agnostic."

According to the language of the draft and earlier reference documents 
on which it was built, the specification document creates only the 
minimum markup necessary for a user to detail preferences concerning 
that external privacy rules be followed, any limitation on length of 
data retention, and any limitation on any retransmission or further 
disclosure.

Peterson said the group aims to bring the document to Last Call -- one 
of the final approval steps before RFC status -- in late February or 
early March.

Assuming the document is approved, goes to RFC, and is widely adopted 
as a de facto specification -- which is fairly likely, considering the 
wide array of industry support within the GEOPRIV Working Group and its 
other affiliated groups in the IETF -- the stage could be set for a 
number of compelling applications.

In addition to, say, relatively mundane uses like being able to 
pinpoint a roving colleague in advance of a big meeting, enterprise 
systems could support rules-based messaging using geographic data. In 
an event of a server failure, for instance, a system could swing into 
action, alerting only IT administrators who were nearby and available.

Peterson, an early figure in Session Initiation Protocol (SIP) and SIP 
for Instant Messaging and Presence Leveraging Extensions (SIMPLE) has 
his own anticipated scenarios. For one thing, VOIP applications based 
on SIP could see a major boost with the introduction of location-based 
data, which would provide necessary infrastructure for 911 emergency 
calling.

"The ability to convey within a SIP 'Invite' message for VOIP the 
location of a caller is critical, and it's been a really substantial 
gap in SIP's story for some time," he said. He also cited gaming 
applications as a personal favorite.

"But what we're trying to do is provide the foundational building 
blocks that people can apply as they want," he added. "The challenge is 
to balance the need to share this information with users' personal 
privacy -- hence this whole concept of GEOPRIV."

Christopher Saunders is managing editor of InstantMessagingPlanet.com.