Presence and IntComm WG

[Candace Holman <>]

There is still interest in discussing the social context of PIC/ALS and 
writing a paper as a guideline for designers and for other reasons.  I 
attended two adhoc gatherings and wrote some notes.

Candace


PIC Social Context Issues
Quotable thoughts from Member Meeting BoFs 4/19/2004

Attendees: Deke, Alistair, Jeremy, Ben, Candace, Jamey, Randy

1. How many people have announced to you - I don't like being tracked?
        The IETF thinks about it - geopriv
        Targeted advertising is okay sometimes - tell me when/where I can 
get a free slice
        Would like to be able to opt out if/when it gets out of control
        Don't want the data to get in the wrong hands but pizza 
consumption is not a social issue

2. Designs need to include a way to limit the bad things that can happen
        Privacy mechanisms should be considered/discussed
        Privacy levels - information privacy, presence privacy, spam, 
blocked lists, barge-in privileges for your boss du jour
        Call it "sophisticated presence"

3. Public safety
        Hazardous situations or vulnerabilities should be private
        finger daemon was disabled because of possible target enumeration 
- he's in his office but noone else is online?
        Anonymity is not the answer to public safety issues

4. Data mining
        Need to design a way to check who is correlating your information
        Is there a way to prevent correlation, other than anonymity or lying
        Some collections are grand and some are small (in scope, intent, 
nature of data, etc)

5. Limit our scope in what we address in the paper
        Stick to a campus deployment
        But talk to campus representatives of the "real" world (public 
safety, general counsel, etc)
        Different issues for students, faculty, staff
        Without going on a tangent, could use some extreme examples to 
illustrate the issues
                GoogleMail, biometrics, etc

6. What do consumers do to foil the system that carriers provide or the 
designers design?
        Edge vs central architecture - let the implementor choose, let the 
design support both
        Things we haven't even thought of yet
        It is intriguing to think of ways to foil the system
        It is a hard problem to think of ways to foil the system until you 
use it
        It is not always a security problem or a consumer/producer tussle

7. User interface
        A complex language is required to assert presence
        Make it simple for users - design the dials and switches for them 
to express presence
        It's not just about good UI design, modularity, etc
        It's social issues and good design
        As long as we can make designers aware of the issues in a way that 
makes sense to them

8. Commercialization of the internet has turned it upside-down
        Protocol development has changed in so many ways
        Radical changes in adopting new technologies cause issues to crop 
up sooner
        Even though it is hard to predict them, we can't let the issues slide