[humanities-dev] Federated ID for TEXTUS

[Jonathan Gray]

Agreed. The login example you sent + logging in with Google account
was extremely smooth.

If we can make it this easy + document the process somewhere on site
then I think this is a good way to go.

I love the idea of using OpenID - just want to make sure we take UI/UX
into account, as well as the fact that we can't expect our users to
have encountered this before and the possibility that they won't
necessarily be very technically savvy.

Worth blogging about on TEXTUS blog and soliciting for input from the
digital humanities community on what other people have found/done?

J.

On Thu, Feb 23, 2012 at 10:15 AM, Christian Morbidoni
<christian.morbidoni at gmail.com> wrote:
> I think how you implement the UI and login workflow is very important,
> and it only partially depends on the use of OpenId. For example you
> don't really need to redirect, but you can open a login popup, which
> makes users experience nicer. Then you can also avoid to register to
> openid account in your website, it could be done automatically (as in
> the link I shared in the previous mail).
>
> However that post raised interesting points. Google changing their
> OpenId urls would be a problem, again I wonder if browserid could
> solve this, as it bases of the email as an identifier.
>
> best,
>
> Christian
>
> On Thu, Feb 23, 2012 at 11:02 AM, Tom Oinn <tom.oinn at okfn.org> wrote:
>> On 23 February 2012 09:34, Jonathan Gray <j.gray at cantab.net> wrote:
>>> On Thu, Feb 23, 2012 at 9:23 AM, Jonathan Gray <j.gray at cantab.net> wrote:
>>>> Could also be worth speaking to CKAN team about why they phased out
>>>> support for OpenID on TheDataHub.org?
>>>
>>> See also:
>>>
>>> http://allinthehead.com/retro/351/openid-has-failed-so-what-s-next
>>
>> This article utterly misses the point. OpenID in the original form it
>> was envisaged is no longer used, but sign-in with Google, Y!aho!o!,
>> Facebook is everywhere and the first two of those *are* OpenID, they
>> just present the login UI differently. I agree that telling academics
>> they need an OpenID account would be a recipe for confusion, but how
>> about a google account, for example?
>>
>> Personally I'd rather trust google with my login credentials than the
>> OKFN, they're bigger and have a much higher stake (due to being rich)
>> in keeping themselves away from privacy related lawsuits.
>>
>> OpenID, at least with google, can also give us verified email
>> addresses (through OpenID attribute exchange) should we need them for
>> notification without requiring users to go through a confirm loop,
>> it's a lower friction system all round.
>>
>> It'd be interesting to see what proportion of our target user base
>> already has an appropriate account (Google and Yahoo being the most
>> significant).
>>
>> Tom
>>
>> _______________________________________________
>> humanities-dev mailing list
>> humanities-dev at lists.okfn.org
>> http://lists.okfn.org/mailman/listinfo/humanities-dev



-- 
Jonathan Gray
http://jonathangray.org