Just reading Andy’s post of the same title. I think that you could argue (if you wanted to play Devil’s Advocate, or are particularly partial to arguing) that Athens has actually been a bad thing in that it has been too effective, and actually held back investment in  other (perhaps more institutionally based) authentication/authorisation solutions in the UK. I’ve always wondered why solutions like ezproxy have much higher takeup in the US than in the UK – and Athens is surely the answer?

On the Shib front, although it is clearly where we are going with JISC at the moment, I can’t help but feel that we really ought to be seeing demand driven from somewhere other than library resources. For access to library resources in the UK HE sector, Shib seems like overkill – it certainly goes way beyond anything we need to do in terms of controlling access to this type of resource at the moment.

Shibboleth was originally championed by the Grid computing contingent in JISC, but this seems to have disappeared a bit recently – or I’ve just stopped paying attention. For example the ESP-GRID project http://www.jisc.ac.uk/whatwedo/programmes/programme_middleware/project_espgridjuly04.aspx – this was meant to report in March 2006, but the project website seems empty.

Anyway, based on some work I’m currently involved with which is looking at e-learning across 3 institutions, I can see some potential for Shib – at least in the next few years. Here, you can imagine Shib being used to allow access to relevant resources depending on your role in each organisation. I don’t think that the ‘personal learning’ environment will be realised fully in the next 5 years, so some time yet for federated authentication/authorisation to be of use.

Also, there is a question – will ‘personalised’ mean not hosted? Perhaps HE institutions will be the providers of personalised learning portals (i.e. the environment is personalised, and perhaps transportable, but provided by a single institution) which will allow consumption of relevant material from learning objects etc. across a federation – then something like Shibboleth might make perfect sense.

Just to go back to an earlier comment of Andy’s, that he was worried that blogs might stifle discussion. I started to leave this posting as a comment on the eFoundations blog, but ended up blogging it instead. The problem with this is that its a hell of a lot harder to follow discussion when it stretches across several blogs than when it is focussed on a single blog.