> > IF the object is to find a way of piggy-backing on the back of a
> > technology (OpenID) around which there is now a very
> significant level
> > of mainstream (i.e. outside the education sector) interest [...]
>
> It's not clear to me that "the mainstream" sees OpenID as
> mainstream, e.g.:
>
> http://news.cnet.com/8301-13577_3-10111404-36.html?tag=mncol
Yes, agreed.
Actually... fwiw, I don't strongly disagree with anything you say here.
It's about making best guesses at future directions and then responding
accordingly. In my view, 'limited experimentation' can be a 'strategic
change' (though not necessarily an overly bold one) if it is done with
strategic objectives in mind.
Andy.
> > [whitelisted OpenID federation]
> > may not play 100% directly to the current OpenID ethos of open,
> > decentralised, user-centric identity provision in the short term...
>
> ...is in fact the antithesis of those original OpenID goals...
>
>
> > but
> > so what? It's not an unreasonable expectation that, in the longer
> > term, we will see developments around the trust infrastructure
> > associated with OpenID and that will either mitigate
> against the need
> > for a federation (of the kind we might put in place now) or will
> > modify the way in which it works. Indeed, by playing in
> that space we
> > will be more easily able to contribute to such developments rather
> > than watching from the outside.
>
> There's a fair chance of that, but this is the magical
> mystery tour quality of OpenID: if you don't like it today,
> you can expect it to be the name for something radically
> different tomorrow.
>
> > By moving in the direction of OpenID, even if it is only
> perceived as
> > a partial move, then we potentially get significant benefits as a
> > community (in terms of mainstream adoption) and we offer our users
> > much greater likelihood of a seamless experience as they
> move between
> > different phases of "lifelong learning" and across educational and
> > non-educational online spaces.
>
> I also would hope to get some mileage out of OpenID global
> identifiers for these purposes as it settles down over time
> (and if it doesn't self-destruct as the open-everything true
> believers tussle with the whitelisting pragmatists -- it
> remains to be seen how that will play out) but trying to
> build a "hybrid OpenID / Shib federation" any time soon would
> look to me like building on sand. I.e., if I'm going to
> piggy- back, I'd like it to be on something stable. In the
> current environment, limited experimentation in this area
> seems a better option than bold strategic change.
>
> Fiona.
>
|