Shibboleth Developers

[Scott Cantor <>]

The HS is now logging the usernames it maps to log4j, and I clarified
the page it generates to explain why not to click the button. The replay
attack is partly addressed in the SHIRE for now, and the thing should be
more robust with regard to this for the time being. We'll revisit later.

I've made most of the small fix-ups I had on my list, and I think all
the main AIs for 2.5 are done, so we should probably look at branching
early next week. If I have time I'll prepare new tarballs I can test out
this weekend.

As others have noted, this alpha will not interop with the last one
because of changes to the way the POSTed assertions are signed and
verified, but it also runs the signature part much quicker now. It's
virtually instantaneous, and the bottleneck is now the AA, and the poor
caching by the SHAR, which Derek will of course deal with with his
design.

-- Scott

------------------------------------------------------mace-shib-design-+
For list utilities, archives, subscribe, unsubscribe, etc. please visit the
ListProc web interface at 

    http://archives.internet2.edu/

------------------------------------------------------mace-shib-design--