[Ntp] Re: Grease in Roughtime

[David Venhoek <[e-mail]>]

Hi Marcus,

Apologies for taking some time on this. My concern here is that
suppose we go to a version 2 of roughtime in the future, clients may
for a while send out requests that are both valid as version 1 and
version 2 requests, accepting either response. If the server balks at
seeing version number 2 (because it doesn't exist according to it) and
thus would not, or not respond correctly to such packets, this would
make version negotiation really painfull. So servers in my view should
ignore version numbers they don't recognize in the VER tag, as long as
there is at least one they do know about.

I hope that clarifies.

Kind regards,
David Venhoek

On Mon, Sep 23, 2024 at 10:49 PM Marcus Dansarie <[e-mail]> wrote:
>
> Thanks for the comments!
>
> I just made a pull request on Github
> (https://github.com/ietf-wg-ntp/draft-roughtime/pull/4) that should
> clarify how grease is used in Roughtime. It also addresses most of
> Martin and Kristof's comments.
>
> On 2024-09-17 13:52, David Venhoek wrote:
> > First of all, regarding the ver tag, there is no requirement for
> > servers to ignore unknown versions in that tag. this means that
> > technically it is valid behavior right now for a server receiving
> > unknown versions to reject that packet, even if there is version
> > overlap. This seems highly undesirable
>
> I don't think I fully understand the problem you are describing here
> however. Could you give an example?
>
> Kind regards,
> Marcus
>
> _______________________________________________
> ntp mailing list -- [e-mail]
> To unsubscribe send an email to [e-mail]