From julvesta@cv3.cv.nrao.edu Thu Jul 6 13:38:33 2000 Date: Thu, 6 Jul 2000 13:27:27 -0600 (MDT) From: Jim Ulvestad To: Athol Kemball , Greg Taylor , Debra Shepherd , Tim Cornwell , Steven Myers Subject: Minutes of today's testers meeting AIPS++ Intensive Testing Meeting Minutes, 6 July 2000 (jsu) ----------------------------------------------------------- Attendees: Cornwell, Kemball, Myers, Shepherd, Taylor, Ulvestad Discussion was held on the general goals of testing AIPS++ for VLA reduction. This included considerable discussion of the philosophy and status of AIPS++ at present, how to best use the testers' time, and how the testers and AIPS++ personnel should interact. The primary goals appear to be: 1. Test scientific completeness of AIPS++ (e.g., missing applications) 2. Find/identify defects 3. Improve usability 4. Compare results of reduction of same data set(s) in AIPS++ and AIPS 5. Improve documentation ("Getting Results ...") Some other random points ------------------------ 1. Reduction and comparisons should probably begin with relatively simple VLA continuum polarimetry data sets, progressing to spectral-line and high-frequency observations. The AIPS processing should be documented carefully so that the same essential inputs can be used in AIPS++ (e.g., weighting, clean depth). This documentation may be run files or notes of key parameters. AIPS++ processing documentation can be achieved by turning the scripting command on, and later editing the scripts. 2. The freeze for the next release is aimed for 1 October 2000. UV visualization and data editing are particular improvements for this release that should be tested, where the data path is being "broadened." 3. Testers should use the "stable" version of AIPS++. There was some discussion of the best version of the on-line documentation, but there didn't seem to be a consensus. 4. It was noted that work on the display library for visualization and editing is in progress. Some of the functionality is currently available via a pgplotter, and more will be added there as needed for usability. This should be tested and improved, because it is the functionality that will be available at the next release. However, the ultimate goal is to use the more capable display library. 5. There was discussion of specific goals and metrics for the testing, but no metrics more specific than the general goals listed above were agreed on. Action items, due 17 July ------------------------- 1. Next meeting, 10:00, 17 July, room TBD (jsu) 2. Complete compilation and availability of VLA data sets (gbt, dss) 3. Start reducing at least one VLA data set in AIPS for comparison (dss, gbt) 4. Read development plan in AIPS++ Note 234 (all testers) 5. Complete draft of "Getting Results on VLA with AIPS++" (ajk) 6. Develop interaction plan for AIPS++ personnel with testers (ajk, tjc)