[evlatests] Software time June 14

[Barry Clark]

Down to the next level of problems.

Flagging was not working right.  Data remained unflagged while antennas
were slewing, both for VLA and EVLA antennas.  There were brief intervals
of flagging in the middle of the slew.

There were two incidents where the data became odd (non-closing, and,
mostly, down from full correlation.  This lasted about fifteen seconds, 
and ends with a ten second gap in the data stream, at which time it 
starts up correctly.  Ken thinks this is some internal glitch in the
correlator, and that it also occurs with the Modcomps in control, at
the rate of a few a day.