Re: [Sidrops] Some remarks on draft-ietf-sidrops-rpki-rsc-06.txt

[Job Snijders <[e-mail]>]

Dear Theo,

On Fri, May 20, 2022 at 08:10:26PM +0200, Theo Buehler wrote:
> This mail was stuck in the moderation queue for ~10 days. Most of it is
> obsoleted and has been addressed by draft-ietf-sidrops-rpki-rsc-07.txt
> 
> The one remaining point is this. I would still appreciate some
> clarification:
> 
> > 4. FileNameAndHash with optional filename field.
> > 
> > The draft only says that the filename field is optional. This seems
> > strange for a list of files whose hashes are supposed to be checked.
> > No rationale is given. I cannot find any text that explains what an
> > implementer is supposed do with a naked hash without corresponding
> > filename. Example use, a permission to ignore or an instruction to
> > report would all be helpful.

The FileName being optional in FileNameAndHash was done at the request
of some folks who envisioned workflows in which content is directly
addressed and not stored on a filesystem. An example could some kind of
challenge/response where only the RSC file is exchanged without any
additional files.

I think your suggestion to ask implementers to report naked hashes to
the relying party operator is a good one. We can add this in -08.

Kind regards,

Job