centralized DS - contributions?

[Tom Scavo]

On Tue, May 19, 2015 at 7:20 PM, Vladimir Mencl
<vladimir.mencl at canterbury.ac.nz> wrote:
>
> I think there is still a need for a CDS - something federations would
> provide as the backup solutions for SPs that would not be deploying the EDS.

I think the point Scott was trying to make is that you can implement a
CDS function using EDS technology, the conclusion being that two
separate technologies are not necessary.

> And while we do not need this feature at the moment (situating has changed
> and we have our bases covered), the need for multiple federations on a
> single DS may easily come back.

With the rise of eduGAIN, global IdP metadata is already here, but
designing a discovery interface based on federation boundaries doesn't
really follow from that. I don't think that discovery approach will
stand the test of time.

I think the elephant in the room is per-entity metadata distribution,
which will disrupt everything we know (or think we know) about
discovery. I really have no idea how that is going to play out, but
heavy investments in traditional discovery mechanisms are ill-advised
at this point.

Just my two cents,

Tom