- From: Melvin Carvalho <melvincarvalho@gmail.com>
- Date: Sun, 8 Jul 2012 19:28:58 +0200
- To: Michiel de Jong <michiel@unhosted.org>
- Cc: Kingsley Idehen <kidehen@openlinksw.com>, public-fedsocweb@w3.org
- Message-ID: <CAKaEYhKQ4vVTpEMkhr5LeXt_DFbkG7sp4j3RUrbqw459TLi8xQ@mail.gmail.com>
On 8 July 2012 17:15, Michiel de Jong <michiel@unhosted.org> wrote: > On Sun, Jul 8, 2012 at 6:03 PM, Kingsley Idehen <kidehen@openlinksw.com> > wrote: > > For sake of experiment, make yourself a WebID via the service at: > > http://id.myopenlink.net/certgen . > > yes, i have one, and i like webid in that it's end-to-end, but i think > technologies like webid and pgp can by definition never be more than > an option for power users. > I dont use WebID because I'm a power user. I use WebID because it is the system that best gives me data freedom. > IMO we should give a default option that can work with a simple-to-use > user address and password combination, and then describe a power user > option based on client-side certificates. > > as Markus said, the default option should probably be based on the > OpenID Connect work (which is basically the same as saying it should > be discoverable OAuth i think - openid and oauth are sort of > interrelated). but only saying 'nodes should implement OpenID Connect' > is not enough to solve these two use cases, i think? Scanning over > http://openid.net/connect/ i see no explicit mention of friend lists, > but maybe i didn't read carefully enough? > > Btw, the Lanyrd example is easier to achieve, because it's about who > you follow, which is something your node can authoratively report on. > The AirBnb case is about who follow (and thus implicitly endorse) you, > for which we probably need some sort of signatures? i think this is > built into foaf? > >
Received on Sunday, 8 July 2012 17:29:26 UTC