Hi Tiru,

Thanks for your review.

I have raised https://github.com/cose-wg/draft-ietf-cose-dilithium/pull/25
towards
addressing it.

I'll address the "why seed only" question here directly, I'm grateful for
the opportunity to discuss this with the ietf community once more.

There is a thread which discussed this here:
https://mailarchive.ietf.org/arch/msg/cose/j2Tyni0VPcq9zRlCKUA0iCO1TIE/

Russ and Tim both advocated for alignment (which we had in a previous
version of the draft).
Richard, Neil, and MCR disagreed.

I've not been able to attend WG sessions, so I may lack helpful context
regarding this, but the change was made as result of:

https://mailarchive.ietf.org/arch/msg/cose/mQXoYDTzpyqP-HGBzE2Y-l96g0I/

Daniel weighed in later:
https://mailarchive.ietf.org/arch/msg/cose/ZYg6hwq8tzez0YxX-2Yt78mkjkE/

FWIW, I think COSE and JOSE made the correct call here.
We had an opportunity to make a *single private key representation*, and we
took it.

Regards,

OS, (As an Author)

On Wed, Sep 10, 2025 at 8:32 AM tirumal reddy <[email protected]> wrote:

> Document: draft-ietf-cose-dilithium
> Title: ML-DSA for JOSE and COSE
> Reviewer: Tirumaleswar Reddy
> Review result: "Ready with Issues"
>
> Hi,
>
> I have reviewed this document as part of the Ops Area Directorate's
> ongoing effort to review all IETF documents being processed by the IESG.
> These comments are written primarily for the benefit of the Ops Area
> Directors. Document editors and WG chairs should treat them like any other
> Last Call comments.
>
> The draft is well-written and addresses an important need for PQC
> migration.  I have a few operational and deployment-related observations
> that may help improve the document:
>
> 1. ML-DSA produces significantly larger public keys and signatures
> compared to traditional algorithms. This size increase can create
> challenges for deployments with limited bandwidth, memory, or processing
> capacity.  I suggest adding text to highlight it.
>
> 2. I suggest adding a reference to Section 8.3 of
> draft-ietf-lamps-dilithium-certificates, which explains the rationale for
> disallowing HashML-DSA.
>
> 3. It may be useful to add a note to explain why only the seed format was
> chosen for private keys, given that the LAMPS WG selected the expanded
> private key format to maximize interoperability with existing
> implementations.
>
> 4. You may want to refer to the security considerations in
> https://datatracker.ietf.org/doc/draft-ietf-lamps-dilithium-certificates/
> and discuss if randomized signing is preferred over deterministic signing.
>
> Cheers,
> -Tiru
>
_______________________________________________
jose mailing list -- [email protected]
To unsubscribe send an email to [email protected]

Reply via email to