[Ntp] Antw: [EXT] Re: NTP Security (was NTPv5: big picture)

[Ulrich Windl <[e-mail]>]

>>> Magnus Danielson <[e-mail]> schrieb am 28.01.2021 um 11:27 in
Nachricht <[e-mail]>:
> Philip,
> 
> On 2021‑01‑22 03:34, Philip Prindeville wrote:
>>
>>> On Jan 19, 2021, at 5:24 AM, Magnus Danielson <[e-mail]> wrote:
>>>
>>>> I think the point is that you cannot bootstrap secure time out of
>>>> nothing.
>>> That's a conjecture, yet to be proven.
>>
>> No, that's basic Information Theory.  Lost or non‑existent information
can't 
> be recreated from a void.
>>
>> ...
> Which needs to be properly proven to apply, because we end up shutting
> this WG down if it truely applies.

When "a void" means "maximum entropy" you actually cannot create new
information as creating information increases the entropy.
But fortunately we wouldn't exist in such a state any more, so no reasons to
worry ;-)
Otherwise you are actually not creating information "from a void". Maybe all
out new ideas are just some transformation of information that is already
there...

>>
>>>> You either do full validation using some trusted time source
>>>> (e.g. RTC), or you don't.
>>> RTC is not a trusted time source. 
>>
>> I think I'm more likely to trust that which I can control or verify for 
> myself, than that which I can't.

So you don't trust the RTC initially, but you trust yourself (self-signed
certificate, so to say).

> 
> Which makes it an operational issue, which turns out to fail eventually,
> and when it fails and NTP is your only way to resolve, you do not have
> time and hence you are out of service. That is not leading towards a
> robust solution.
> 
> So, in conclusion, rather than quickly dismiss the notion of being able
> to bootstrap, it needs to be revisited. I through out a proposal to get
> something to test. People did not like that, so now I let others find a
> solution to replace it, and the purpose of my proposal was to provide
> something for people to think and analyze, and find flaws and then fix
> those flaws. What you ended up doing was dismissing the bare thought.
> 
> The lack of interest to solve this means that I see less use and
> therefore less interest in NTPv5.

Out of curiosity: Does anybody know a solution that deduces rough time and
position using a camera pointed at the night sky, using some star almanach?

Regards,
Ulrich

> 
> Cheers,
> Magnus
> 
> _______________________________________________
> ntp mailing list
> [e-mail] 
> https://www.ietf.org/mailman/listinfo/ntp