Re: [Ntp] Antw: [EXT] Re: Quick review of WGLC for status change for draft‑ietf‑ntp‑update‑registries

[James Browning <[e-mail]>]

On Thu, Aug 11, 2022, at 3:22 AM, Harlan Stenn <[e-mail]> wrote:

> On 8/9/2022 3:51 AM, Miroslav Lichvar wrote:
> > On Tue, Aug 09, 2022 at 12:39:02PM +0200, Martin Burnicki wrote:
> >> Of course you can have a real v4 client use autokey but send a faked v3
> >> version, in which case it might get a v3 response with autokey
> extensions.
> >>
> >> I've never heard that this caused any problem. On the contrary, the high
> >> level of compatibility between v3 and v4 made the slow changeover from
> v3 to
> >> v4 very user-friendly.
> >
> > It doesn't cause problems if you use a compatible client and server,
> > but it's out of the specification. The client shouldn't send an
> > NTPv3 request with extension fields and the server shouldn't accept it
> > and respond.
>
> V3 does not have extension fields.
>
> Sure, somebody could make a v3 response with them, but that's out of spec.
>
> Just like somebody can send a v3 or v4 packet and say it's v1.  That
> doesn't make it "true".
>

Great, someone should go and tell the people who are selling stupid 'NTP'
wall clocks.


> > The Windows NTP client and server use 64-byte MACs. They use NTPv3 and
> > I think that is ok. There are no extension fields yet which could be
> > confused with the MAC. In NTPv4 that would be a problem.
>
> Not in my experience.
>

Not to butcher lagomorphs but it seems to be 68, the adapter code is
ntpd/ntp_sign.c, and in ntpd/ntp_proto.c


> >>> That is not possible if we want to make incompatible changes in NTPv5.
> >>
> >> Yes, that's why I believe the acceptance of v5 will grow very much
> slower
> >> than the acceptance of v4.
> >
> > How long it took for NTPv4 to reach a majority of NTP traffic?
>
> The time it takes is independent of the drama involved.
>
> v3 clients talk to v4 servers perfectly well.  This has been
> demonstrated in the real world, and is also well-documented in RFC 5905.
>

Because NTP version 4 broke the tradition of making breaking changes every
version. It is being brought back.

Also, NTP packets have been 48 octets since RFC 958 and haven't changed the
leap indicator or latter 2/3.