user@domain identity form musings

[M. David Peterson]

Hi Martin,

It seems the simple solution then would be to regex '@' and replace with '.' 
so user at domain.com would become user.domain.com
<http://user.domain.com>before being passed to the server...

Adding a bit more to this... Actually, I think this presents an interesting 
opportunity. It just so happened that OpenID.name <http://OpenID.name> was 
still available so I've registered it and will happily accept a request for 
transfer from Brad or donate it to whomever he suggests will be the holding 
body (livejournal?) for the OpenID entities. What this seems to present 
quite nicely a way to both create a network for each implementation of 
OpenID such that setting up a server and verifying its validity would then 
give you the ability to request
whatever.openid.name<http://whatever.openid.name>and have the zone
file updated with the correct IP or domain to redirect the
requests too. By doing this a simple data base of implementers can be 
created, offending members easily tagged (e.g. those who blatantly abuse the 
system to take on an appereance of someone else or for spam related 
purposes, etc.. at the root of each *.openid.name could live a
status.xmlAtom or RSS feed making it quite easy to maintain system
wide stats using
REALLY SIMPLE TECHNOLOGIES....

Brad, if you woul like to guide me to the proper person I would need to 
expect a request to transfer from I will prepare this and "sign" the papers 
as soon as I get them. If, for now, you want me to hold tught, thats fine 
too.. I guess that would give us all a good chance to build implementation 
for this into all the servers for testing.

Cheers :)

<M:D/>

Obviously the ideas could be extended quite substantially...

Either way, I do like the '@' idea... for the masses to embrace new ideas 
they need to feel like its at least somewhat familiar and using the @ email 
directive, even if just for show, seems to me like it could drive it home 
for a lot of people.

I actually have some more announcements regarding the implemenation of the 
OpenID system on the .NET platform and then using this as the primary 
identity scheme for the ChannelXML project -- a community effort to 
decentralize the web. When I finally have had a chance to get things to the 
point they need later today Iwill update the list and takes things from 
there...


On 5/27/05, Martin Atkins <mart at degeneration.co.uk> wrote:
> 
> loune wrote:
> > Hi List,
> >
> > As I mentioned before, the draft specs for the ID system I was working
> > on, instead of using URL, it uses IDs similar to emails. I've just
> > realised that such a format could also be used with with open ID. A URI 
> > could contain the form http://user@domain/ and as long as the consumer
> > passes this form to the ID server with the is_identity param, the ID
> > server can take it and then do whatever verification with it. 
> >
> > So the end result is that multiple people can share a website and still
> > be identified individually.
> 
> I suppose that will work, though since the user/password URL format is
> not technically allowed for HTTP I expect that it will not work for 
> everyone's HTTP libraries and browsers. Internet Explorer doesn't allow
> this since Windows XP Service Pack 2, I believe.
> 
> I would be concerned that some consumers would be unable to accept such
> URLs, and even once they have the links they generate will be unusable 
> for many web browsers.
> 
> _______________________________________________
> yadis mailing list
> yadis at lists.danga.com
> http://lists.danga.com/mailman/listinfo/yadis
> 



-- 
M. David Peterson <aka:xmlhacker/>
http://www.xsltblog.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.danga.com/pipermail/yadis/attachments/20050527/076c889f/attachment.htm