[evlatests] Time error fixed.

[Ken Sowinski]

On Wed, 18 Feb 2009, Ken Sowinski wrote:

> There is more work to do to understand what happened.  A simple 
> diagnostic would be useful.

I consulted with Rick.  A simple test that could be done during
startup would be to map a known calibrator with respect to a nearby
calibrator.  If time is wrong then the position for the known
"target" will be wrong.  Can an AIPS script be put together to
reduce such a simple observation and squawk if it gets the wrong
answer?  With suitably bright calibrators a snapshot of a few
minutes should be enough.