Subject RE: publisher interface study
From <J.Paschoud@xxxxxxxxx>
Date Mon, 21 Sep 2009 10:03:43 +0100

I agree that a common name and branding would help.  This was a clear
conclusion from the excellent Publisher Interface Study report by Rhys
Smith.  Brand and infrastructure can be independent.  (I'm sure there
must be some non-Internet-world examples out there but I can't think of
one now.)
However, we need to appreciate that the actual providers of federation
services in each country (usually the NREN such as JANET in the UK) will
naturally have a desire to maintain their own branding.  This is evident
in the UK by the fact that we still like to call eduroam "JANET Roaming
Service" or JRS (reminiscent of "Fog in Channel, Europe cut off"  - but
please Google for this yourself if you're not old enough to know where
it comes from ;->).
Perhaps the key problem we need to solve is how a universal brand can be
presented alongside information about the parties involved in a
particular transaction (i.e. IdP, SP and WAYF).  "A resource protected
by eduID, provided by Elsevier, Licensed by Bloghampton University, with
discovery mediated by JANET(UK)" is *definitely too clunky!  - But
something designed right to convey this information to an end-user could
also help to identify who to call when something doesn't work.
I raised (at a JISC meeting for potential project bidders last week) the
issue that usability and UI design  - including of the 'discovery
problem' and the more sophisticated bits of FAM with which users may
interact in the future (devolved authority, etc) is significantly
missing from the key themes that JISC is trying to address in its' next
round of IdAM development.  This needs to be more than assumed or
implicit because it's crucial to the whole thing becoming a dominant
John Paschoud
Projects Manager & InfoSystems Engineer
LSE Library
T: +44(0)20 7955 6129
E: J.Paschoud@xxxxxxxxx



	From: Nicole Harris [mailto:n.harris@xxxxxxxxxx] 
	Sent: 21 September 2009 09:27
	To: refeds@xxxxxxxxxx
	Cc: Josh Howlett
	Subject: Re: [refeds] publisher interface study
	Hi Josh
	Couldn't agree more - the 'name bit' is a small part of the
battle.  This needs to be combined with a lot of other work that is
going on - such as the work Rod is doing on the WAYF etc. etc. and
detailed guidelines on how publishers should implement both the 'eduID'
name to replace the plethora of 'institutional login', 'shibboleth',
'federation' that is out there at the moment, and to improve the general
user flow.  We are hoping to appoint some professional UI people to look
at these guidelines. 
	I think the problem that any 'academic' login might work in some
places and might not in others is always going to be a problem because
at the end of the day it depends on whether your institution both has a
subscription to that service and uses federated access.  Again, this is
all going to be part of sensible fail messages etc. etc. 
	I don't necessarily think that common brand, or at the very
least 'recognisable name' as I'm not sure we are sophisticated enough to
be a brand (!) without common infrastructure is a problem.  OpenID is in
exactly the same position as a service with a common brand but
distributed architecture.  I think we have one up on them in terms of
scale of use by SPs at the moment though :-)
	Hope that helps
	Josh Howlett wrote: 

		Hi Nicole,

			We are nearing the end the of the consultation
period for the 
			publisher interface study and as yet have had no

		First I should say that this is, in principle, a fine
		However - and perhaps I have misunderstood this proposal
- I'm curious
		whether a common branding is a Good Idea in the absence
of a common
		If it is not possible for a typical SP to obtain an
assertion from a
		typical IdP (say, because they had no federation in
common), then this
		might end up disappointing users. You can imagine a user
		confused when the 'eduID' button works for some SPs, but
not others.
		I see that a smart embedded WAYF could help,
particularly for those SPs
		that only care about certain IdPs.
		In summary, I certainly support the idea and think we
should start
		working on the branding ASAP; I'm just saying that I
don't think this is
		enough to solve the general case. Perhaps that's been
implied, but I
		haven't seen it said explicitly.
		JANET(UK) is a trading name of The JNT Association, a
company limited
		by guarantee which is registered in England under No