Shibboleth Developers

[Scott Cantor <>]

Ian Young wrote:
> That's not to say that it might not be worth pushing in that direction
> anyway, as otherwise you're responsible for the fork forever.  But I
> wouldn't make any plans that depended on resolving the problem any time
> soon, say for Shib 2 time.

Partly it's a time pressure thing (we really are pretty close to a stable 
beta, believe it or not), but I also know that as time goes along, we'll 
need upkeep of the makefiles and source code to keep up with Linux and OS X 
and so forth, and I have no confidence I can get that to happen if I don't 
control the code.

So forking is actually less work, unfortunately.

> Yes.  Because eventually you want this package to be someone else's
> problem again.

And it moves us toward a more portable model where I could eventually pull a 
Tomcat and wrap the logging so that we could change it radically every 
release, with the goal being total lack of output on initial install, like 
Tomcat has achieved.

(Ahem, that's a joke.)

-- Scott