[evlatests] flagging questions

[Ken Sowinski]

There have been many conversations of late about the correctness
of flagging produced by mcaf.  I will try to summarize the current
state of understanding in one place.

1.  Starting in mid July there were reports of inappropriate flags
raised for many or all antennas over a period of many scans.  Rich
thinks that mcaf was responsible and the problem has been addressed.

This problem caused a more careful look at data and reminded us of
something old:

2.  There have been ocassional examples of an antenna flagged for a
scan due to an FRM problem.  This is nothing new.  The operators
have long been in the habit of noticing subreflectors that were not
correctly positioned and correcting them by hand.  We should compile
a list of the worst offenders and provide it to the servo group so
they can be checked.  There are accuracy and repeatability tests that
they can carry out.  The tolerances for the FRM are 0.2 degrees in
rotation and 0.2 cm in focus, independent of band.

3.  Some of the flagging nomeclature is not as straightforward as
we might hope.  Joe's attempt to describe an SDM flag table should
try to standardize this in some useful way.  I argue that flagging
nomenclature is more useful if it reflects hardware organization
rather than SDM organization.