[daip] forwarded message from John Benson

[Eric Greisen]

------- start of forwarded message (RFC 934 encapsulation) -------
Received: from walrus.aoc.NRAO.EDU (walrus [146.88.2.26])
	by zia.aoc.NRAO.EDU (8.9.3/8.9.3) with ESMTP id PAA06044;
	Thu, 20 Jun 2002 15:01:30 -0600 (MDT)
Received: (from jbenson at localhost)
	by walrus.aoc.NRAO.EDU (8.9.3/8.9.3) id PAA21068;
	Thu, 20 Jun 2002 15:01:29 -0600 (MDT)
Message-Id: <200206202101.PAA21068 at walrus.aoc.NRAO.EDU>
X-Sun-Charset: US-ASCII
Content-Type: text
From: John Benson <jbenson at aoc.nrao.edu>
To: jromney at walrus.aoc.NRAO.EDU, egreisen at walrus.aoc.NRAO.EDU
Cc: jbenson at walrus.aoc.NRAO.EDU
Subject: job 6034 - bk090b 
Date: Thu, 20 Jun 2002 15:01:29 -0600 (MDT)


I looked again at job 6034 that caused problems in AIPS. Here is 
what I found :

- -The job as built by cjobgen has four rows in the FQ table, but
 only rows 0, 1, and 3 are used during the job. That's fine, the
 correlator would have run okay. We expect that some job script tables
 may have rows that are not actually used.

- -The analysts changed the job script, changing the fq_row = 3 to
 fq_row = 2 in the observation tables, and shortened the FQ table to 
 three rows. The now row 2 was the row 3 in  the original job script.

- -The correlator ran okay on the modified job script. But since cjobgen
 makes a companion FITS file (.CL) of Tsys, phase-cal, and other tables that
 are referenced to the FQ table in the job script, changing the job script
 FQ table was a bad idea. The analysts cannot change the indexing in the
 FITS tables in the .CL file so they are mismatched. 

John...

------- end -------