[evla-sw-discuss] inconsistent correlator configurations

[Ken Sowinski]

While I was in Penticton Sonja and I discussed how to treat some
inconsistencies in configuration requests.  I repeat what we decided
to give anyone who cares a chance to disagree.

The issue is how the correlator should respond when antennas which
are declared as belonging to a single subarray request different
baseband, subband, or correlation parameters.  We agreed that such a
request would be an error.  Such an inconsistent collection of antenna
configurations would be rejected in toto for all the antennas which
are determined to belong to that subarray.  The current VLA monitor
and control system and correlator explicitly forbids such a request
and I see no need to support it in the future.

The only exception granted is for antennas which may not have a full
complement of hardware (or VLBA antennas) and are allowed to present
only a subset of the BW presented to the correlator by full EVLA
antennas.  In this case only a fraction of the basebands and subbands
will produce useful correlated data.  The responsibility for marking
the invalid data lies outside of the correlator system; the correlator
will require consistent configurations, even for the basebands and
subbands for which signals may not be present.  Sonja, the last
sentence was not strictly part of our agreement; are you happy with
it?