[daip] SPLIT problem

[Eric Greisen]

Andreas Brunthaler wrote:
> Dear Eric,
> 
> we found a problem with SPLIT in AIPS. The dataset is a VLBA calibrator 
> search, i.e. short scans on ~40 sources of which only a few are 
> detected. We use our script that we have used a few years ago for our 
> calibrator search for the BeSSeL survey. The calibration still works as 
> before, but we noticed something odd when we split the data. We use
> 
> sources = ''
> 
> in SPLIT, to get single source data for all our detected sources (the 
> non-detections don't get a single source file, because there is no valid 
> data). This worked fine in the past. However, if we run SPLIT on the 
> calibrated dataset now, it stops for some reason after 7 sources (3 
> detections and 4 non-detections). If I run split on the uncalibrated 
> (and unflagged) raw data, SPLIT works as before and I get single source 
> files for each source. Is there a feature now that SPLIT stops after 4 
> sources with no data?
> 
> The calibrated data set can be found here:
> 
> ftp://ftp.mpifr-bonn.mpg.de/outgoing/brunthal/bm387z.uvfits
> 
> Andreas
> 
If enough sources in a row failed, then the history card buffer 
overflowed since the pointer in that buffer was not reset in the case of 
failure.  Tomorrow's MNJ will fix this.

Eric Greisen