Re: Location Requirements - Data Details

Maybe James means the use of the 64-bit NTP timestamp format, as in 
RTCP. That, I believe, is a more reasonable requirement. Whether this is 
better than the alternative of a time-zone based spec (Wednesday, July 
17, 2002 17:45 +500) is debatable, but probably way beyond the detail 
justified in a requirements document. (Since the latter may accidentally 
leak timezone information, the NTP format may have privacy advantages.)

I agree that requiring a particular synchronization method is not 
appropriate. For example, a GPS -based beacon won't have NTP, but it 
will have time that's much more accurate than your average NTP-synced 
clock.

Randall Gellens wrote:
> At 7:15 PM -0500 7/17/02, James M. Polk wrote:
> 
>>  I firmly believe the timestamp should be a requirement (using NTP)
> 
> 
> I'm not sure we should require use of NTP.