From: owner-its-mgrs-group@lists.colorado.edu on behalf of Burton Fox
[Burton.Fox@colorado.edu]
Sent: Friday, November 09, 2001 4:48 PM
To: ITS Managers
Cc: Malinda MILLER-HUEY
Subject: EduCause debrief

Well, I've had a solid week to digest what I heard and saw at EduCause 2001
and before it all completely escapes me I am going to transfer my notes and
lingering thoughts from the presentations, hall way discussions, and
airplane magazine articles to you.

The Educause conference is the largest gathering of Information Technology
folks from institutions of higher education.  The Educause site is here:
http://www.educause.edu
Many of the papers presented will be available in the coming weeks at
http://www.educause.edu/conference/e2001/.

Monday Session I
Web Portals: What they are, what they do, why you need one
Howard Strauss, Manager of Academic Applications, Princeton University
http://www.princeton.edu/~howard/slides/portals
(must use IE or pages won't display correctly)

Howard forces me think outside of my comfort zone.  Even though I agree with
him on many points, I cannot bring myself to subscribe to his conclusions,
some of which are:
- No more institution centric home page
- There should only be one portal. (don't want the students using Yahoo! or
Excite - we want them to use our portal)
- There must exist -complete- customization available to the user.
Otherwise, they will continue to use another portal that allows them to do
what they want.
- Replaces your desktop
Some of the neat terminology Howard creates:
Cameos: Small pieces of data from larger data set
and most important, the most important challenge isn't technical, it is
requiring all data owners to work together.

Monday Session II
Building the "Next-Generation" Enterprise Application Portal
James Thomas, John Walsh, Indiana University
Indiana's solution is called "Onestart"
http://onestart-ui.edu (there is a guest logon available)
The project is documented at
http://www.indiana.edu/~onestart/project
Major goals
-Sticky authentication (stays with you as you navigate through the resource)
-Role based 'one-stop' shopping
-Emphasis on self-help
-Application integration (not an information portal)
Best practices
- Communicate vision (non techie work)
- Establish clear detailed plan
- Engage core audience (Faculty/Staff/Students)
- Process for Prioritization must be in place
Another neat term: Analysis Paralysis
	Through iterative development one cannot see how the project will ever be
finished
Results
-Better service for students
-Communication & Collaboration
-Cradle to Endowment relationship management
Not tracking how many customers actually customize their view, but a guess
was made of 10%.

JA-SIG meeting
http://www.ja-sig.org
Java in Administration special interest group
Carl Jacobson, U of Delaware
JA-SIG came about because of Carl's need for Java programs and not enough
resources available to create them.  It started out as a way for UofDelaware
to share code with area institutions such as Cornell.  It is now a 'match
maker' resource - not a repository, for applications written in Java that
may be useful to other institutions.  The 'clearing house' model has been
tried before, and this attempt appears to be losing steam.  Vendors are now
included in the listing of available applications and the applications no
longer are available for free.  The real problem is the success of uPortal
which came out JA-SIG using a Mellon grant to develop the Higher Ed
reference specification framework for portals.  Carl said that no one has
provided any applications for 'match maker' in months and most of the time
was spent talking about uPortal.  In fact, this meeting was cut short to
allow participants to participate in the next track session dedicated to
uPortal.  Disappointing.

Monday still, after lunch, l o s i  n g   p  o  w  e  r r
>From reference to reality: going live with uPortal at British Columbia
Dave Frazer, Paul Zablosky
http://www.ubc.ca
Everyone is waiting for a large institution to commit and go live with
uPortal and I thought BC was going to be the litmus test we've all been
waiting for.
Key point: No critical functions are offered through the portal
Well, -that's- fairly important.  All the other stuff just doesn't mean as
much in that light.
35,000 students (2nd largest in Canada)
A reference spec doesn't include much exception handling code.  In
production, 20% of the code handles normal events and 80% handles
exceptions.  This translates to a lot of programming to create a production
service from what is provided out of the box with uPortal.  What am I
talking about, there's no box, no documentation, just a large chunk of java
code and a very active listserv with lots and lots of questions.
Recommendations:
- Allow limited channels, with a minimum dependency on external resources
(unreliable)
- Use the biggest most powerful server available
- Cost? Don't ask.
8 months later - BC had to 'fork' from reference code.  Even though modular
in design, it will be no picnic to upgrade to version 2.

Tuesday morning.
Kick ass presentation on Enterprise Wide Directory Services
http://www.colorado.edu/Committees/DirectoryServices

Great general session on MIT's learning center
http://www.media.mit.edu/physics

Poster sessions:
How to chose a portal
http://www.calumet.purdue.edu/public/ctis/osika/roadmap/index.htm

The last little bit of interesting information was Cal Berkeley's new
pricing model for voice and data.  The went through a whole 'thing' but I
found it interesting that one of the services they began to charge for was
listserv.
(The 'thing' was identify -all- costs, products -- services-- LOB Line of
Business...)

I hope you get a chance to look at some of the web references.
-Burton