From julvesta@cv3.cv.nrao.edu Fri Aug 4 17:21:19 2000 Date: Fri, 4 Aug 2000 16:57:18 -0600 (MDT) From: Jim Ulvestad To: Tim Cornwell , Athol Kemball , George Moellenbrock , Debra Shepherd , Steven Myers , Greg Taylor , John Hibbard Subject: AIPS++ Meeting Minutes, August 1, 2000 AIPS++ Users Meeting Minutes, 1 August 2000 (jsu) ------------------------------------------------------------ Attendees: Cornwell, Kemball, Moellenbrock, Myers, Shepherd, Taylor, Ulvestad, Hibbard (by phone) Agenda -------------------------------------------- (1) Test reports (2) Suggestions -- Plotting, editing, other? (3) Documentation status (4) Near-term plans -------------------------------------------- An AIPS++ Users meeting was held on 1 August 2000 in order to review progress. Discussions focused on calibration, plotting, and flagging issues, as well as the versions that should be used. Due to some uncertainties, Kemball sent a message the previous week describing the hierarchy of versions and the recommendations for testing. The bulk of that note is reproduced below: --------- 7/27/00 note from Kemball --------- Dear VLA testers, A reminder concerning the hierarchy of AIPS++ builds may be useful at this point in the testing: daily (built Sun-Thu in the evening) : Daily development snapshot. Not recommended for end-user use. weekly (built Fri in the evening) : Weekly development snapshot. Useful for testing, but not to be relied upon. stable (every few weeks) : A designated weekly build which passes all automated testing. Good for end-users. release (every six months) : Last released version, including all patches. Good for end-users. We need to avoid the use of daily; the code there is changing quite rapidly and it is not meant for end-user access. If you need to use something in daily and there is some other problem with the daily build that day, I suggest the following approaches: i) wait until the next daily or preferably weekly build, or: ii) test another part of the system in weekly or stable as part of the same reduction. This will probably be the most productive approach overall. [ ... ] ---------- End included note ---------- Some of the other salient points from the meeting: - "Daily" should not be left up overnight, because the system update at midnight will do nasty things. - PGPLOT environment variables have been changed. There is a conversion under way from PGPLOT to the TV viewer system. This is currently in "daily," will be in "weekly" next week. - Discussion of flagging gave the information that something like TVFLG is in the plan for a bit later this year, while the aim is for flagging to be possible in all plots in 2001, possibly by April. - An X-band calibration conversion problem was discussed--in "casually" edited data, almost no solutions converge, while there are some bad solutions even in rigorously edited data. Kemball is looking at the convergence problem. It is thought that the fact that AIPS does RR and LL separately, while AIPS++ does all polarizations together, is not the problem. - Taylor plans to try polarization calibration later in the week. - Documentation was discussed, including the philosophy behind "Getting Results" and "Getting Started". There was a consensus among the testers that a separate VLA-specific chapter that is something like a cookbook is a very high priority before the next release. Hibbard will try representing the random user away from the AOC to see if he can figure out what to do from the current documentation. - Work is under way on the MAP tool, to replace the IMAGER tool. Kemball will let the testers know when MAP is ready to try. There was some concern that there is not a lot of time for testing and documentation of this tool before the next release. - Next meeting to be held on Tuesday, August 15, at 10 a.m., in the AOC upper conference room.