Re: [TLS] Inclusion of OCB mode in TLS 1.3

[Aaron Zauner <[e-mail]>]

Jack Lloyd wrote:
> On Thu, Jan 22, 2015 at 12:58:08AM +0100, Aaron Zauner wrote:
> 
>> I'm in the process of editing the draft and will add this information to
>> it. Do you want to suggest a paragraph/sentence or do you think this
>> will suffice for implementors?
> 
> Just that statement probably is sufficient, but to minimize the number
> of misinterpretations possible maybe something like this is better -
> 
> """
> The nonce input to the AEAD is described above using the TLS
> presentation language. All values are represented in big-endian form
> when constructing the AEAD input.
> 
> The sequence number of a message is always known to the receiver
> through other means (either implicit protocol state or a per-message
> header in the case of DTLS), so the nonce construction used does not
> require any extra per-message information. Thus the record_iv_length
> is zero (0) for all ciphersuites defined in this document.
> """
> 
Thanks a lot! I'll add those paragraphs to the document. The current
revision has just the single sentence in it.

Aaron