[evlatests] Update on the timing tic.

[Vivek Dhawan]

This a condensed record of recent activity.

In an attempt to diagnose the EVLA time offset, adjustments were
made to the NTP (network time protocol) service yesterday. This
broke the fringes on EVLA antennas for some times/antennas. The
breakage was undone today.

A more permanent solution is in the works. Ken and James Robnett
have more information which they can summarize after the dust
settles.

Software and maint time took up most of the ~24hr period in which
the NTP errors occurred. These errors are easily caught by tests
in place already.

Upon examination, only one science project had a little affected
data, which can be easily removed in the usual calibration and
flagging.

The main problem (the occasional time offset on the whole array),
is likely still with us. It has not been seen since Feb 21.

Vivek.