An open discussion list for topics related to the eduGAIN interfederation service.

[Tom Scavo <trscavo AT internet2.edu>]

On Tue, Dec 23, 2014 at 7:49 AM, Niels van Dijk
<niels.vandijk AT surfnet.nl> wrote:
>
> Reading through the eduGAIN metdata requirements (Which I think live
> here:
> http://services.geant.net/edugain/Resources/Documents/eduGAIN_metadata_profile_v3.doc)

I don't think this is an eduGAIN issue but I'll add my two cents anyway.

> I get the impression we have done our best to make IdP discovery as
> hard as possible:
>
> - - We do not require a display name for an IdP at all, neither in the
> native language nor in English.

I agree with you, all IdPs should have an mdui:DisplayName (and
perhaps other user interface elements). The problem can be traced to
the (improper) use of OrganizationDisplayName, which is still widely
used as a display name for *both* IdPs *and* SPs.

> - - We do not require a logo; actually the logo is not even mentioned in
> the document as a could/should. As such there is no recommendation on
> size and on the way to present it (url vs embedded).

Unfortunately, I think this is hopeless. Logos in metadata are fraught
with difficulty (as you say). I think we'd be better off focusing our
attention on the IdP's favicon instead.

> Basically, when literally interpreting the mandatory parts of the
> spec, the only reasonable thing for an SP to do is to fall back to
> presenting the entityID. Typically not something a user will
> understand intuitively...

I agree, this is frustrating. The SP can (and often does) fall back on
OrganizationDisplayName but that only perpetuates the problem.

> I understand that we may want to keep the level of entry as low as
> possible but what users get served by an IdP they cannot find in a
> discovery? Or am I missing something here?

No, you're not missing anything but I don't think eduGAIN is the place
to attack this problem. The only thing I'll add is that per-entity
metadata begs the question whether the prevailing discovery model has
legs. Rather than try to repair what we have, we may be better off
considering discovery from scratch.

Tom