[mmaimcal] Re: subarrays

[Michael Rupen]

A couple points:
  * splitting according to antenna capability seems pretty fundamental
    to me.  Am I missing something here?
  * I agree that several of the points I raised were not very quantitative --
    I haven't seen any real discussion yet of the way calibration or the
    finding of phase calibrators in real time might be done, although several
    suggestions involving subarrays have been made.  In print even :)
  * I've been thinking of subarrays as all or nothing -- either you have the
    capability to use 6 independent subarrays, or you don't. If it matters
    how often you use it (apart from the basic question of whether it should
    be made available at all) it would be good to know why, & what tradeoffs
    are useful to think about. 
  * Has there been anything, even at the level of my bare-bones rambling, 
    addressing the question of subarray cost?  Maybe I've been out of the loop
    too long & missed it.  I've chatted with a number of different correlator
    types about this, not to mention Barry, and no-one seemed to think it
    was a big deal, so long as it wasn't really ridiculous (e.g. 1 per antenna).
    So I didn't realize this was a huge cost driver -- is it?

-- Michael