Shibboleth Developers

[Thomas Lenggenhager <>]

Is it correct, that the ApplicationID is in fact the ID of the
Application Domain?

BTW, the document about application domains is no longer at the location
announced in the e-mail last October, but I could find a cached copy at
Google
http://www.columbia.edu/~wassa/application.domains/app.domains.html

> I've got a simple draft schema for configuring the mapping between resources
> and application IDs in the target.
> 
> Here's a sample:
> 
> <ApplicationMap xmlns="urn:mace:shibboleth:target:appmapper:1.0"
>       ApplicationID="http://jstor.org/InCommon";>
> 
>       <VirtualHost Name="www.jstor.org" Port="443" SSL="true">
>               <Path Name="admin"
>       
> ApplicationID="http://jstor.org/InCommon/admin"/>
>       </VirtualHost>
> 
> </ApplicationMap>
> 
> There's a default ID at the root that you override at any point by inserting
> VirtualHost elements (with hostname/port/scheme) and then Path elements.
> Obviously you can nest Path elements in each other to whatever depth is
> needed.
> 
> Constructing the above, you get https://www.jstor.org/* mapping to the
> default app ID except that https://www.jstor.org/admin/* maps to the
> overridden app ID.
> 
> I'll throw some XSLT together to include in the package to spit out the URL
> mappings, shouldn't be too difficult (unless somebody wants to give me a
> script, hint hint).
> 
> The ApplicationID values are URIs, and would normally be the SAML 2.0
> ProviderID values used by a target/SP to issue messages, and dereferencing
> the URI would usually return a SAML metadata document.

Doesn't this mean, that in your example
        http://jstor.org/InCommon
and     http://jstor.org/InCommon/admin
return SAML metadata documents? With these URLs I would more expect a
normal web page, the application or one of its services.

Wouldn't it make more sense in real life to use something like
        http://jstor.org/AppID/InCommon
and     http://jstor.org/AppID/InCommon/admin

Or did I got something wrong with the concept of SAML metadata?

Beyond these questions I like your proposal.

Thomas