Efficiency in AttributeDefinitions

[Peter Schober]

* Bennett, Steve <s.bennett at lancaster.ac.uk> [2013-04-11 13:18]:
> > The current IdP does not do conditional resolving, yes.
> >
> > There are some hacks to make it do that, if you really want to
> > (check the list archives) and already have the numbers that prove
> > it's necessary
> 
> I'm not particularly keen to start hacking the IdP just yet.

I was not referring to changing any of the code, merely a different
wiring of the configuration (which one could call "a hack", as in
"original", "using stuff in other ways than probably intended",
"non-standard").
It's all in the archives.

> I don't have any figures to demonstrate a need since our IdP see
> quite low traffic and certainly has no performance issues, but I
> wanted to confirm that I wasn't missing any obvious tricks.

If a conditional resolver (i.e., only fetch what is known to be
necessary for a specific relying party) was available I suppose it
would be the default configuration. Or documented at the very least.
-peter