too much here but I think you're thinking about Open ID
also related federated identity management proposals
gotta go drink wine and eat burger now
cheers
--> --
> The Sharebro Google Group: for http://sharebro.org and related development
> http://groups.google.com/group/sharebro for archives and options
>
Alex Chaffee - al...@stinky.com
http://alexchaffee.com
http://twitter.com/alexch
On Tue, Nov 15, 2011 at 8:33 AM, Emmanuel Pire <pirem...@gmail.com>
wrote:> From Alex K.>> In terms of identity, and social graph there
are some standards already.>> You should take a look at webfinger for
identity, and FOAF for connection>> lists.> I took a look at both,
they have very good ideas. What do you think should> be the identity:
email like webfinger suggest or openID style with browser> login ?>
> I'd love to hear your thoughts on these services.
My thoughts are, this is a complicated issue and different
applications have different needs and therefore different solutions.
OpenID was designed as an attempt to solve this "federated/distributed
identity" problem and I think it was only partly successful. There are
a few design flaws (what if your chosen OpenID provider dies?) and a
few UX flaws (the original OpenID forced your id to be a huge long URL
string; these have been mostly fixed in recent years, but in different
ways by different providers).
The state of the art now is that most webapps maintain their own
concept of *identity* but may use common means for *authentication*.
E.g. even when you "sign in with facebook" (or google or twitter or
openid or whatever) on a webapp like GetSatisfaction
[http://getsatisfaction.com/login], that webapp still maintains its
own profile and account info for you. And if you link your accounts
then you can *authenticate* with any of your providers but still be
known as the same *identity* once you're in. I think that state of the
art is good enough for me for the time being.
I have big conceptual problems with email = identity. Emails are often
not under a user's control (e.g. j...@comcast.net). Also the same human
usually has many email addresses -- and it's never as simple as "one
for work and one for personal use". Furthermore, lots of people don't
use email regularly; they use Facebook and Twitter and SMS all the
time tho.
I have big conceptual problems with email = identity. Emails are often
not under a user's control (e.g. j...@comcast.net). Also the same human
usually has many email addresses -- and it's never as simple as "one
for work and one for personal use". Furthermore, lots of people don't
use email regularly; they use Facebook and Twitter and SMS all the
time tho.
Not sure you need identity if you are posting your shares to a rss feed. (tumblr, blogger, ect) The rss feed URL is your identity.
--
The Sharebro Google Group: for http://sharebro.org and related development
http://groups.google.com/group/sharebro for archives and options
If you haven't seen it before, WebFinger might be relevant:
http://code.google.com/p/webfinger/
Basically lets you use something that looks like an email address (eg.
m...@lmorchard.com) as an identity.
The webfinger scheme lets hosts (e.g. lmorchard.com) specify mapping
rules from names (eg. "me") to URLs. And, so, services can discover
identity details like this:
http://webfinger.org/lookup/m...@lmorchard.com
--
m...@lmorchard.com
http://decafbad.com
{web,mad,computer} scientist