Re: [TLS] Re: Comments on draft-santesson-tls-ume-04/draft-santesson-tls-supp-00

"Stefan Santesson" <stefans@microsoft.com> writes:

>> Eric:
>> 
>> >These drafts seem basically sound. Minor comments below.
>> >
>> >-tls-supp:
>> >You should state that only one SupplementalData field may
>> >be used per handshake.
>> 
>> I agree.  This was pointed out by the Gen-ART reviewer too.
>> 
>
> [Stefan] I've put this on my change log. To be precise in words, should
> the text state that "The client and server MUST NOT send more than one
> supplementalData handshake message each."?

Yes, I think that's right.

>> This is being discussed.  The reason that I would like to see it as
>> standards-track is that the structure supports more than just
>> Microsoft names.  The Microsoft names are simply the first ones that
>> are supported.
>> 
> [Stefan] A agree to the proposed path of the reason stated by Russ. I'm
> willing to tune down the mentioning of Microsoft as it isn't necessary
> for the syntax definitions of this document.

That doesn't really resolve my issue, because it still results in
incompletely specified semantics.

>> >You can do the UPN hint extension exchange and then NOT
>> >send supp_data? That seems wrong.
>> 
>> I agree, if the negotiation is successful, then the supplemental data
>> should be sent.
>> 
> [Stefan] I don't see much value in this direction as the server MUST be
> prepared for the case where the client chooses not to send any hint. It
> may have come to the conclusion that this is not the right server to
> send the hint it intended for some reason. Another reason may be that
> the user has declined sending his user name to this server in a user
> dialog box.
>
> I view sending of hints as completely voluntary at the discretion of the
> client. Even after extension exchange.

Then that that needs to be stated clearly in the specification.

-Ekr

_______________________________________________
TLS mailing list
TLS@lists.ietf.org
https://www1.ietf.org/mailman/listinfo/tls