[daip] Bug in AIPS BLING

[Eric Greisen]

On 06/25/2015 02:24 PM, Walter Brisken wrote:
>
> Hi Amy and Eric,
>
> I think I found a rather serious problem with the task "BLING".
>
> Immediately upon FITLD I run BLING with all default parameters and:
>    calsour='28884''
>    solint=2
>
> Note that the source named '28884' is the second entry in the SU table.
>
> This runs to completion.  Upon inspecting the output BS table I find
> that for the first processed baseline all of the values in the SOURCE
> column have the correct value (2).  All subsequent baselines have the
> SOURCE column set to 1.  Other values (timestamps, baseline, stokes,
> ...) all look OK.
>
> Data:  /home/parallax/G15MAY19.FITS
>
> I have a temporary work-around (just ignore the source column) but that
> may not be good long term.
>
>
> Let me know if you need any more info...
>
> -Walter

I have found the bug and it will be fixed in tomorrow's MNJ.
The main bug was the fact that the NX table records the center time and 
length of a scan and so in real*4 can get the stop time slightly less 
than the last data point.  The handling of the NX table was odd too so I 
have tried to upgrade it.

Eric