Re: [security-services] Authentication Methods - Proposed changestocore-29

[Jeff Hodges <Jeff.Hodges@sun.com>]

I wholeheartedly agree with Hal's first four comments and wish to see them
addressed. 

thanks,

JeffH
-----

> Hal wrote:
> 
> I take exception to a few things that differ from what I explicitly changed from core-29.
> 
>       -----Original Message-----
>       From: Hallam-Baker, Phillip [mailto:pbaker@verisign.com]
>       Sent: Wednesday, April 03, 2002 4:20 PM
>       To: Hallam-Baker, Phillip; 'Hal Lockhart'; 'Philpott, Robert'; security-services@lists.oasis-open.org
>       Subject: RE: [security-services] Authentication Methods - Proposed changes to core-29
> 
>       Proposed complete text:
>        
> 
>       Subject Confirmation Methods are defined in the SAML Profile or Profiles in which they are used [SAMLBind]. Additional methods may be
>       added by defining new profiles or by private agreement.
> 
>       The following identifiers refer to SAML specified Authentication methods. Where Base64 encoding is specified the data is encoded as
>       specified by [RFC 2045]. 
> 
>        
> 
> There are no places where Base64 encoding should be specified. (See below) Therefore this sentence should not go here. If there are other Base64 encoded
> items earlier in the document, perhaps a similiar reference to [RFC 2045] should be included at the appropriate place.
> 
>       1.1.1. Password (Pass-Through): 
> 
> First, I assume this is supposed to be 7.1.1 etc. 
> I explictily specified leaving out the "(Pass-Through)" In what sense is this a pass-through? The System Entity sent a
> password to the Authentication Authority. The AA confirmed that the password was correct. The AA later issed an
> assertion with an Authentication Statement in it. The AA did not pass-through the password to anybody. Please drop
> "(Pass-Through):" from the title as I specified previously.
> 
>       URI: urn:oasis:names:tc:SAML:1.0:am:password
> 
>       The authentication was performed by means of a password.
> 
>       1.1.2. Kerberos 
> 
>       URI: urn:ietf:rfc:1510
> 
>       <SubjectConfirmationData>: A Kerberos Ticket 
> 
> No, no, no. There is no such thing as Subject Confirmation Data. What we are defining here is Authentication Method. It goes in the AuthenticationMethod
> attribute of the Authentication Statement. Look at the schema. (lines 667 & 1814 in core-29) There is no <SubjectConfirmationData>. There is no
> Authentication Data. There is no data of any kind, just the Method identifier. That is the point of this entire revision.
> 
>       Theauthentication was performed by means of the Kerberos protocol [RFC 1510], an instantiation of the Needham- 
> 
> Should be a space between "The" and "authentication". (I'm surprised your picky grammer checker didn't catch that one. ;-)

[..snip..]