Followed 'recipie.txt' to the word, and everything seemed to work!

started up cima ok
start up galfa1 ok with /var/diag and /var/levels_a2034

moved to G36p18v87_1a and did simple on, to see where the line was in the narrow band for Galfa1. Seemed to be in the right place.. 
continued with frequency switch calibration, seemed to go ok too.
(THE SFS CALIB IS QUITE LONG ~10 MIN CAN WE SHORTEN IT?)
closed file, and continued with drift.
(file=a2034.0001.fits)


ran script command_drift.

first scan:
I was quite sure that I was on source on tim (just) but CIMA reports that I was not.. I'll need to re-check that I am calculating these things ok.. 
(file=a2034.0002.fits)

second scan:
again, I seemed to get there too late...
however, the interleaving dec position looks good.
(file=a2034.0003.fits)

third scan:
again, I seemed to get there too late...
Also, the new increment in dec for the next set of interleaves looks ok
at least i'm calculating something right.
(file=a2034.0004.fits)

fourth scan
again, I seemed to get there too late...
(file=a2034.0005.fits)

fifth scan
Arrived on time for this one.. with 82 sec to spare.
I REALLY have to find out what is going on..
(file=a2034.0006.fits)

sixth scan
arrived with 102 sec to spare!
(file=a2034.0007.fits)

seventh scan
arrived with 78 sec to spare!... good grief.
aborted this one part way through..
(file=a2034.0008.fits)

NOTE: subsequent testing put crap data into this file.. do not use!!!


The last few scans, which were on time, are the right length, and spacing.. 
I need to streamline it a bit so that we dont arrive late.. and we dont arrive too early.


Okay, I've discovered that I was calculating the positions using YESTERDAYS date, which would account for the fact that the scans were starting 4 minutes late..
Need to think more..stoopid.