[evlatests] WIDAR baselines, and time offsets

[Ken Sowinski]

On Tue, 7 Jul 2009, Rick Perley wrote:

>    2) It's 33 seconds, not 32.  The VLA data have a 1.67 second
> granularity, the WIDAR data are at 1.0 seconds.

Old archive records provide a timestamp at the end of the
integration interval; the SDM timestamp describes the center
of the interval.  I expect those differences to be gone away
by the time you see data in AIPS, but you can get close to
a second that way if something is wrong.