[evlatests] setting reference antennas

[Bryan Butler]

Ken Sowinski wrote:
> On Tue, 22 Jul 2008, Bryan Butler wrote:
>>
>> Ken Sowinski wrote:
>>>
>>> You lose the pointing and phasing history.
>>
>> could you elaborate on this?  what do you mean?
> 
> My understanding is that pointing and phase results do not persist
> between scripts.  If a script is interrupted after a pointing or
> phasing scan and then restarted the information learned is now lost.

phasing i understand, though i think the use cases for this are few and 
far between.  pointing, hm.  probably only a concern for global pointing 
model runs?

>>> It is clumsy for files in durations.
>>
>> how so?  seems to me the mechanism is exactly the same for durations 
>> or stop times.
> 
> A duration file, if started over, begins at the beginning not where
> it left off.  That addresses my first concern, but means more time
> than was expected is used for the observation.

sure, but i'd guess that 99% of durations files are used during testing 
anyway, so this doesn't strike me as much of an issue.

> 
>>> Someone has to edit a file with the chance of making mistakes.
>>
>> indeed, but we have this vulnerability however we choose to do it, 
>> throughout the system.
> 
> Clicking on a button to select a reference antenna may not be
> as prone to error as editing a file.

assuming rich exposes this in the array operater's interface or something?

seems to me this stretches across so many different programs that the 
benefits aren't worth the development cost.