Re: [ntpwg] call for adoption (draft-dfranke-ntp-data-minimization)

[Harlan Stenn <[e-mail]>]

On 3/27/17 1:50 PM, Daniel Franke wrote:
> On 3/27/17, Harlan Stenn <[e-mail]> wrote:
>> It would help if you properly quoted your context.
>>
>> I don't see a '32' anywhere in what Aanchal said, above.
> 
> The context I should have included was from Miroslav upthread:
> 
>     My suggestion is to remove Precision from that list and add before
>     that paragraph: Precision SHALL (or SHOULD?) be set to 32.
> 
>     Anyway, do the all other fields need to be zero? As long as they are
>     constant and not unique to the individual clients, as in most SNTP
>     implementations they currently are, it wouldn't be a problem, right?
> 
>> Assuming this is about the poll interval, I have 2 comments:
>>
>> - Bluntly, OpenNTPD would not be my first choice for a source for NTP
>> protocol experience or knowledge.
> 
> Regardless of your opinion of OpenNTPD design choices, the facts
> remain that it's in wide deployment and sets the precision field to
> zero.  Those facts alone make it a good choice for standardization
> since:
> 
> 1. Matching an existing implementation leads to one less opportunity
> for fingerprinting.
> 2. Existing widespread use of this value means we can be confident it
> won't break anything.

If they made a suboptimal choice we should take this opportunity to make
a better choice.

My previous discussions on the usefulness of this sort of fingerprinting
are still unanswered.

There is no reason to believe that one arbitrary value for this is
somehow "magical" and a different arbitrary value would be dangerous.

>> - There is likely benefit to knowing the client's poll interval in
>> certain situations so the server can recommend changes to the poll
>> interval.  This becomes more difficult if the server doesn't know what
>> the client's poll interval is.
> 
> You were confused here about which field we were discussing. Zeroing
> the poll interval is already a MAY in the current draft, partly
> because we want to allow for exactly this sort of thing.

OK.

-- 
Harlan Stenn <[e-mail]>
http://networktimefoundation.org - be a member!

_______________________________________________
ntpwg mailing list
[e-mail]
http://lists.ntp.org/listinfo/ntpwg