[evlatests] Report on AJ329 observations, August 9

[Barry Clark]

AJ 329 was dynamically scheduled, so it goes through the alternate
script generator, obs2py, instead of obs2script.  The organization 
of LST to UT is different on the two, so I am not surprised that the
scans turned up out-of-sync as I only did a couple of spot tests before.
(I appreciate the information - it gives me something definite to look 
into.)  There is a known problem (fixed in the next release) that Modcomps 
schedule by apparent sidereal time, but the Executor uses mean, which
can result in a different synchronization depending on the phase of the 
saros, which I don't off hand know.