On Sep 15, 2010, at 1:27 PM, Sam Hooker wrote:
> Are these apps good candidates for authentication and attribute
> acquisition via Shibboleth?
Possibly. One is a ruby/rails application living on an alternative web server, accessed through a proxy declaration in .htaccess . The other is on the web farm.
I've been waiting to write a SHib aware application since I first heard about it at at an all-day NFS sponsored workshop called the National Middleware Initiative Integration Testbed workshop on "Experiences with middleware deployment" in November 2003 (see CIT-STAFF listserv archives, if they still exist, November 3, 2003).
On Sep 15, 2010, at 1:17 PM, Mike Austin wrote:
> I'd much rather not have that data copied elsewhere if possible - just re-query LDAP as needed. Given the scale of our LDAP systems, I wouldn't think it would be a performance problem to query every time.
Both applications are "Register me for..." kinda things. But besides data collection, there will be data mining and reporting, too. Questions like "How many students applied for X in 2010" or "how many Math Faculty came to WebCT workshops is 2005?" With everything 'copied elsewhere' it is trivial to construct SQL queries against the data. Otherwise, I might have to generate Con-the-fly CSV files for download to Excel for analysis (which, in the case of one application, is probably what the clients want anyway, not being versed in SQL).
Finally, all of this sounds like work and complexity. I am a man of simple tastes, and might just do what's quick and easy.