Toerless Eckert <t...@cs.fau.de> wrote:
> On Thu, Mar 13, 2025 at 04:08:23PM +0100, Michael Richardson wrote:
>> > If public=true it means that the syntax and semantic of the binary
>> blob > is publically documented.
>>
>> I'm not sure this is useful to do it that way. It doesn't matter
>> whether or not it's publically documented, but rather whether a
>> Registrar has been told about it. A Registrar made last year would
>> not know about public documents that came out this year.
> I can have a 30 year old NMS and i can still make it processes MIBs
> that are newer, because it can configure MIB files onto it. Same thing
> for data description files for any such extensions.
Yes... that's what I mean "told about it" --- ideally it could be configured
to support new extensions.
>> At the same time, a vendor-proprietary extension might be told to an
>> operator under NDA, which would not make it public.
> Yes, a public bit is only an incomplete helper. Registrar would only
> allow for voucher to pass if it can decode the blob. If it does not
> have the definition file for the blob, it can prohibit the oucher (or
> if we figure out how remove the blob) and log an error. In this error,
yes, the presence of blobs that are not understood could be considered
suspicious. But my point was that "public" is the wrong semantics.
"known" (or even, "allowed") is what we want to say.
>> >> https://github.com/anima-wg/voucher/pull/81
>>
>> I've defined a leaf (attributed) called "manufacturer-private", which
>> is a mostly opaque byte string that the manufacturer can put anything
>> they like. I say "mostly", because the suggestion is that it's a bstr
>> (CBOR), which can contain any valid CBOR. Or, in JSON land, it's
>> JSON, having been base64URL encoded to a string.
> I'm fine with everything, i am just representing the fears i have from
> other IETF participants that do not like the MASA concept in the first
> place and fear side channels. Hence my suggestions to make this side
> channel as acceptable as feaible.
My experience is that many assume "MASA" must be controlled by the factory in
a hostile foreign regime. I find that sad, and they want to use another term.
All sorts of side channels are possible; I agree that one might want to be
concerned, but ultimately, there are other methods to create them.
--
Michael Richardson <mcr+i...@sandelman.ca>, Sandelman Software Works
-= IPv6 IoT consulting =- *I*LIKE*TRAINS*
signature.asc
Description: PGP signature
_______________________________________________ Anima mailing list -- anima@ietf.org To unsubscribe send an email to anima-le...@ietf.org
