Ah. You misunderstood what I meant by "more than one IdP"

I mean that more than one Authoritative Party will have claims in an identity 
transaction. For example, I can provide a claim that I am a Canadian Citizen 
with a claim from gov.ca, am [email protected], and a California resident from the 
state of CA.

Here is a more near term scenario:

I sign up to newservice.com and want to use my [email protected] identity to prove 
who I am later on, and give newservice.com to post to Twitter, Facebook, 
LinkedIn and Google+ so that I can spread all my goodness from newservice.com 
everywhere. A user-centric design lets my identity agent get OAuth tokens from 
Twitter, Facebook, LinkedIn and Google+ and select my [email protected] address all 
in one permissions page. Currently I have to bounce to each of those providers. 
What a pain! :)

-- Dick

On 2011-07-20, at 9:16 PM, Allen Tom wrote:

> I only skimmed the BrowserID proposal, but my impression is that the user's 
> email provider is the IdP, assuming that the provider implements the 
> BrowserID protocol. 
> 
> In the case where the email provider has not yet implemented BrowserID, the 
> client uses browserid.org as a fallback IdP. BrowserID.org asserts verified 
> email addresses after verifying the user's email. This is only an interim 
> step and is removed from the loop as soon as the user's email provider 
> natively supports BrowserID.
> 
> Therefore, any email provider can be an IdP,  and there's an interim solution 
> to support users whose email providers haven't yet supported  BrowserID. 
> 
> Maybe I'm totally wrong about how BrowserID works.
> 
> Allen
> 
> On Wed, Jul 20, 2011 at 7:01 PM, Dick Hardt <[email protected]> wrote:
> 
> On 2011-07-20, at 8:47 PM, Allen Tom wrote:
> 
>> That's why I like how BroswerID uses the email address as the identifier - 
>> if the user's email provider was the IdP, then we'd be able to scale past 
>> more than one IdP. 
> 
> You will need to elaborate on that so that I understand where the extra IdP 
> comes from.
> 

_______________________________________________
specs mailing list
[email protected]
http://lists.openid.net/mailman/listinfo/openid-specs

Reply via email to