Bug in `2022.01.1` tag

[Former user]

I am using the conda distribution of tempo2 which currently points to and compiles/installs the 2022.01.1 tag.

There's a bug when running trying to fold data with dspsr that runs tempo2 for the predictor. The tempo2 call is failing, and the stderr file contains:

Observatory code 'NULL' not found!

I read the pred.tim file and it contains:

FORMAT 1
 reference 1400.00000000 48442.50000000000000000 0.00000 NULL
 predictor 1241.69112550 59718.08631482351195530 0.00000 ATA
 predictor 1241.69112550 59718.08490721249398447 0.00000 ATA
 ...

the reference has NULL as observatory, so I figured that might be related?

I compiled the master branch, and the same call ran correctly. The "reference" line in the ".tim" file has "ATA" instead of "NULL".

Thanks!

[Michael Keith]

This was introduced when I fixed the predictor code to use the correct TZRSITE, but unfortunately meant that par files with no TZRSITE set broke. You can work around it by updating to the latest version or alternatively adding a TZRSITE (any will do) to your .par file.

[Michael Keith]

Fixed in latest version

[Wael Farah]

Thanks!