Re: [Cbor] Review: draft-bormann-cbor-packed-00 & SUIT

[Carsten Bormann <[e-mail]>]

> On 2020-07-28, at 19:18, Jim Schaad <[e-mail]> wrote:
> 
> [Brendan] One of the questions that I have is the order of compressing and applying security operations.  If you are using a pull parser do you think it is going to be easier (we can talk about more secure later) to do the hashing on the compressed or uncompressed version.  

Whoa.  We normally hash byte strings.
So are you proposing that, say, a compressed version is received, it is uncompressed, re-serialized in deterministic encoding, and then hashed?

> This could be an even harder question to ask with the use of an external dictionary.  There is one potential benefit w/ COSE in that the hash of the dictionary could be included as external data.

There are two aspects here:
— how do we make sure the external dictionary cannot be surreptitiously exchanged by an attacker
— how do we make sure there is an unambiguous way to integrate the external dictionary/-ies into the prefix and sharing tables?

I have a design for the latter that I still need to write up.

For the former, a COSE hash algorithm + the hash would do; the dictionary would be hashed as an encoded CBOR data item.  There is a question whether there should be a way to put in a (compact!) identifier for the dictionary as well, so the hash in the external data isn’t a trapdoor function.

Grüße, Carsten