Re: [rtcweb] IdP in RTCWeb

Now I understand  and agree that we cannot change their specification.  
But... do we need to use it?

Igor

On 3/23/2012 7:22 AM, Harald Alvestrand wrote:
> On 03/22/2012 07:17 PM, Igor Faynberg wrote:
>> Thanks! Now I understand.
>>
>> My only follow-up question (and that could probably wait until the 
>> next week) is why not make BrowserID certificate adhere to X.509.  If 
>> we issue self-signed X.509 certificates ourselves, what is in the way 
>> of any IdP doing just that?
> The BrowserID specification is the work item of a group that is 
> entirely outside of this group.
> You can certainly go there (browserid.org) to argue that they should 
> change to X.509 certificates, but we can't decide in the RTCWEB WG 
> that they should change.
>
>>
>> Igor
>>
>> On 3/22/2012 1:53 PM, Eric Rescorla wrote:
>>> Good question.
>>>
>>> 1. BrowserID certificates aren't X.509, and that's all the TLS/DTLS 
>>> will do.
>>> 2. This lets us be agnostic about IdP mechanisms without having to 
>>> change
>>> TLS every time we add a new one.
>>>
>>> -Ekr
>>>
>>>
>>> On Thu, Mar 22, 2012 at 6:22 PM, Igor Faynberg
>>> <igor.faynberg@alcatel-lucent.com>  wrote:
>>>> Eric,
>>>>
>>>> A question:  For the case of BrowserID, I understand that a client 
>>>> gets a
>>>> certificate from an IdP.   If so, why you and I would not use the
>>>> certificates from our respective IdPs in order to authenticate each 
>>>> other
>>>>   in DTLS?
>>>>
>>>> Igor
>>>>
>>>>
>>>>
>> _______________________________________________
>> rtcweb mailing list
>> rtcweb@ietf.org
>> https://www.ietf.org/mailman/listinfo/rtcweb
>>
>