Re: [Ntp] SNTP and extension fields (WAS: Re: Antw: Re: Antw: Re: Antw: [EXT] NTPv5 Loop Detection without Stratum)

[Danny Mayer <[e-mail]>]

On 9/6/22 5:50 AM, kristof.teichel=[e-mail] wrote:
> > 
> > [e-mail] said:
> > > Yes, but a "packet format" does not mean a fixed-length packet 
> IMHO; instead
> > > if should contain a mechanism for a variable number of optional 
> fields that
> > > can have a variable length and should be parseable even if it's 
> only to
> > > ignore them. I think that is what the e tension fields had in 
> mind, but
> > > somewhat short-sighted.
> >
> > You are tangling 2 threads.  The chunk you are replying to is 
> discussing SNTP.
> >
> > The S is for Simple.  I hope SNTP will work without extension fields.
> >
>
> Wouldn't that preclude use of NTS?
> (It would, and also that of any other security measures other than 
> v3-era symmetric key MAC with unspecified key exchange, if I'm not 
> mistaken...)
>
> If that's correct: can we afford to rule out modern (well-scaling) 
> authentication for any and all SNTP clients?
>
Yes, we want authenticated packets even in an SNTP implementation. 
Saying we don't care about authentication won't get through IETF today.

Danny