Shibboleth Developers

["Scott Cantor" <>]

> I'm bringing this up again because the problem has not been solved and I 
> think there's a bug in the Shibboleth ISAPI filter.

I can't find any evidence of a bug.

> This length of the header variables is not the problem. The problem is 
> that the first variable HTTP_SHIB_SWISSEP_HOMEORGANIZATION to be set is 
> a substring of HTTP_SHIB_SWISSEP_HOMEORGANIZATIONTYPE. So everyone on 
> IIS using "HTTP_SHIB_"header names that are a substring of an other 
> header name would have this problem.

I've tried to test that assumption before and it doesn't hold for me. And of
course the code doesn't operate on substrings, but equality.

> I'd be happy if this misbehaviour could be corrected. For more 
> configuration info see below (AAP and SAML Response snippets).

I would be happy to correct it, but I tried to reproduce this behavior and
could not. I'll try again, but I'm guessing the outcome will be the same. My
guess is somebody seeing the problem needs to debug this, or they need to
add logging statements to try and catch it in the act, so to speak.

-- Scott