Re: [TLS] Confirming Consensus: Negotiating upper layer protocols

[[e-mail] (Martin Rex)]

Peter Gutmann wrote:
> Eric Rescorla <[e-mail]> writes:
> 
> >WG members, please provide any comments on whether we should take this work
> >on by February 21. Additionally, if you wish to propose an alternative, it
> >would be nice if you could do so soon or at least provide an indication of
> >interest.
> 
> A comment on both of these proposals, to encode their protocols they use an
> ad-hoc, non-TLS-style encoding whose form is rather unclear:
> 
>   Protocols are named by IANA registered, opaque, non-empty byte strings and
>   the list of protocols is serialized as a concatenation of 8-bit, length
>   prefixed byte strings.
> 
> Does this mean the strings use 8-bit chars, the lengths are 8 bit, both, or
> neither?  What's wrong with:
> 
>   opaque ProtocolName<1..2^16-1>;
> 
>   struct {
>       ProtocolName protocol_name_list<1..2^16-1>
>       } ProtocolNameList;
> 
> which fits the way everything else is done in TLS?

Do we really need ProtocolNames > 255 chars?

Btw. the lower bounds of container elements are usually
set to the minimum possible length.  In your example,
the lower bounds of protocol_name_list would be 3, I believe.

How about

   opaque ProtocolName<1..2^8-1>;

   struct {
       ProtocolName protocol_name_list<2..2^16-1>
   } ProtocolNameList;

-Martin