[LEAPSECS] happy anniversary pips
igb at batten.eu.org
Fri Feb 7 03:34:20 EST 2014
On 6 Feb 2014, at 17:16, Michael Spacefalcon <msokolov at ivan.Harhan.ORG> wrote:
> Steve Allen <sla at ucolick.org> wrote:
>> Taken at face value Google's Site Reliability Team would seem to be
>> arguing for the return to the bad old days of the rubber second.
>> It's hard to believe that Google's Android and driverless car divisions
>> hold the same position, but they haven't spoken.
> Why can't that driverless car maintain two separate and independent
> kinds of time, one physical, the other civil?
It's actually quite hard to imagine scenarios when a car needs accurate civil time
to a precision greater than that of a reasonable man's wristwatch.
The obvious use-case is making sure you obey time-limited traffic restrictions
(close to Google's home, from memory the multiple-occupancy lanes on 101
only operate between specified times) but enforcement of such things has
errors bars of at least five minutes, because that's the spread of reasonable
error on watches, in-car clocks (especially as in the US RDS is much rarer
as a means of distributing time to car systems) and so on.
More information about the LEAPSECS