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

[Nicole Harris <harris AT terena.org>]

On 07/07/2014 12:15, Lalla Maria Laura Mantovani wrote:
> At 12:06 07/07/2014, Nicole Harris wrote:
>> I do think that adding IdPs to eduGAIN has an more significant impact on
>> how we currently operate and it might be worth thinking more about why
>> you are exporting them and what it is you expect from doing that. I.e.
>> if GARR doesn't expect the UK federation to republish those IdP entities
>> in the UK, why is GARR putting them in eduGAIN?
>
> GARR puts in eduGAIN entities that opted-in and expect to exchange
> metadata with other entities that opted-in as well.
> We are ready for this. 
Sure, but it is important that we keep "presence in the metadata" and
discoverable / usable service separate.  Otherwise you just get in to
problems with workarounds as Peter says.

 I think that the correct action here is to remind the SP
http://scitation.aip.org/ that it should not consume the metadata tagged
as "eduGAIN" at this point in time.  It would also be a good opportunity
to ask them if they wanted to sign up :-) I don't think trying to impose
import bans on edugain metadata is the answer.  At the end of the day,
there is very little difference between 2 separate streams and one
stream with 2 separate tags in it - they are both different ways of
flagging that 'this stuff is different'. Yes it is more likely that an
SP might ignore the edugain tag in metadata than chose to consume an
interfederation aggregate without knowing what it is, but that is
possible.  It also presents opportunities though :-)

http://scitation.aip.org/ are being lazy overall with discovery - for
example all the UK colleges appear in their dropdown and I very much
doubt many of them subscribe to AIP publishing.  So I really think this
is lazy discovery not a symptom of how the UK are publishing.

It is also important to remember that many federations are doing
additional filtering of downstream metadata so reciprocity can never be
guaranteed. There is also nowt in eduGAIN that says "we will only
publish our downstream eduGAIN metadata to those who appear in the
upstream".

Overall though I think that is great that we are starting to see these
problems as it means that eduGAIN metadata is really out there in the
wild.  It also brings up that question of longterm responsibility for SP
support and management in the longterm with eduGAIN, which I think is an
interesting point. 

I really wish we had lower-cased edugain :-)