NRAO Applications Users Group Meeting - DRAFT MINUTES

Date: 2007-05-30 (Wednesday) 
Time: 1300 MDT (1500 EDT)

*******************************************************************

Minutes:

Send additions and corrections to smyers

1. NAUG News

   o Users Committee report (Joe)

     Joe gave a presentation and a CASA demo.  The committee
     seemed satisified with progress and our short-term plans
     for the beta release.

2. CASA Status Report

   o Status of Mac distribution?

     Should have Intel and PPC versions available now.  Try 'em.
     Still some issues with running 32bit installations on a
     64bit machine.

   o Other distributions

     Strong desire was expressed for support of a free Linux distro,
     preferably Fedora Core (which many outside organizations, such
     as LANL, use).  Other options are Scientific Linux and CentOS.
     We'll need some machines for a build farm to support these.
     Joe and sys group will investigate adding one additional OS.
     Concerns raised about this effort detracting from other high
     priority things.

3. ALMA

   o CDR5 report (Joe)

     Joe and Nicole gave presentations.  Were not called back for
     grilling.  Some issues raised by Lewis and Crutcher, such as
     outside collaboration on CASA in US community, and parallelization.

4. EVLA

   o SAGE meeting report (Michael/Bryan)

     Lots of concern on development of algorithms for big EVLA 
     reductions - these are research problems and not pure code
     development.   Discussion on use of NRAO clusters by users.
     Possibility of use of University Centers (with clusters) for
     offload of Science Center processing and storage?  Panel
     seemed split between those who wanted pipelines, those who
     didnt and those who didnt care.  Joe commented that we have
     started doing simulations of EVLA-sized datasets and things
     seem to scale so-far (just doing current style processing).     

5. Main Event - Presentations and Discussion Items

   ***************************************************************

   https://wikio.nrao.edu/bin/view/Software/2007-May-30

   - Alpha patch status

     See beta planning page and checked items for progress
     towards the June 15 release (June 12 freeze):

     https://wikio.nrao.edu/bin/view/Software/BetaPlanning

   - Joe and Kumar Charlottesville Travelogue

     (and general progress by the stay-at-homes)

     o integrated profiles (box, polygon)
     o polygon statistics
     o interactive clean = now rectangles only 
     o image analysis tasks
       * imhead = currently modes get/put/stats 

     Discussion: It was generally thought that statistics should
     be elsewhere than imhead (the min/max in the header is just
     for convenience).  Some discussion that I couldnt hear and
     so will ignore.

     See wiki page.

     ACTION: more work needed on ia stuff, imhead in particular. 

   - Task/Parameter List (Ed):

     Tasks names --- preliminary proposal sent to Joe (Ed/Steve), 5 changes:
         
        clrstat ==> clearstat
        contsub ==> uvcontsub
        correct ==> applycal
        listobs ==> listsummary
        smooth  ==> smoothcal

        Comments?

     Parameter names --- preliminary proposal by Ed discussed in sub-group

     ACTION: accepted.

   - Globals or Not?

     Too much has been said already.  As I (Steve) see it, there are 3 levels
     that we are interested in:

        global - all params are global, and are remembered however they are set
                 (THIS IS THE CURRENT SITUATION - "global means global")

        inter  - only parameters set interactively (in casapy using the 
                 <param>=<value> Python mechanism) are remembered.  Params
                 set explicitly in a function/method call are not remembered
                 (but are saved in the <task>.last file). (This option possible
                 soon if desired.)

        local  - only "environment" variable params are global, all others are
                 task specific (NOT YET AVAILABLE.  Is a requirement for ALMA.)

        The "global" option is what we have, and would be the DEFAULT
        in casapy.  We have a mechanism for the "inter" option
        courtesy Kumar.  Down the road, we may implement the "local"
        option.

        Proposal: have Kumar implement his "casaglobal" parameter to allow
        the global/inter options.  Since this is like an environment variable,
        it would of course be "global" in all options!

        Counter Proposal: make this an environment variable.  We should have
        a dictionary of casaenv variables (like those in SD/ASAP) that can
        be set outside of casapy or inside.  And "casaglobal" should be one
        of them.

     Discussion: lots of noise, no decisions.  Kumar is implementing
     a "casaglobals=True/False" where True gives the current behavior
     (true globals) and False causes the task params to revert to the
     task-specific defaults after execution (though the .last file 
     contains the values used).  The executives feel that this is 
     sufficient, and is not a show-stopper in any event.

     ACTION: further discussion on this closed.  We do want a "tget"
     function though!

   - Logging

     Should log output go to screen in casapy --nolog mode?
     Currently it goes to casapy.log file (which you can monitor
     manually if desired).

     Desire was expressed for an intermediate --termlog option that
     directs logger output to terminal.  The --nolog option should
     have at least high-priority and user-requested (e.g. from list
     tasks) output directed to terminal.

     It was also requested that there be an option in the list* and
     other similar tasks to output to a text file (like AIPS DOCRT=-1
     and OUTPRINT=<filename>), maybe via an "outprint" parameter.
     This output should be free of the extaneous logger stuff (like
     times and tool origins).

     ACTION: Joe will look into this.

   - Data selection 

     No time for discussion on this.  The following document was sent
     to CASA developers for comment.  They will go ahead and merge this
     into their plans where appropriate.

     ACTION: no further action by NAUG needed.

     Preliminary proposal sent to Joe (Ed/Steve) for "selectdata" params.
     Key ones are "spw" and "chanrange" which take selection strings 
     for ranges etc.  See NAUG Note at:

     http://www.aoc.nrao.edu/~smyers/naug/notes/2007-05-30-myers-selectdata.txt

   - Documentation (Steve,Ed):

     No time for discussion on this.

     ACTION: Steve and Ed will continue handling this.  Awaiting next
     batch of task changes.

     The latest version of the cookbook is available on the CASA and
     NAUG pages, or directly at:

     http://www.aoc.nrao.edu/~smyers/naug/doc/cookbook/casa_cookbook.pdf

     Last update was May 14 (only minor fixes).  Next version is due
     May 30 (some Synth Cal improvements).

   ***************************************************************

   Future Topics

   o Final selection scheme (George,Sanjay)

   o suggestions welcome

6. CASA Development Targets and Schedule

     - See project office page:
       http://projectoffice.aips2.nrao.edu
       and the CASA Wiki
       https://wikio.nrao.edu/bin/view/Software/CASA

7. Upcoming meetings and deadlines:
   o 2007 June 15?    CASA Alpha patch 1.  NAUG testing starts.
   o 2007 June 18-21  NRAO 50th Anniversary Symposium (CV)
   o 2007 June 22-24  Meeting "Transformational Science with ALMA: 
                      Through Disks to Stars and Planets" (CV)
   o 2007 July 8-12   Debra and Joes Excellent Garching Adventure
   o 2007 Aug  15?    CASA Alpha patch 2
   o 2007 Sep  30     CASA Beta Release!
   o 2007 Oct  1      Blancmanges from Skyron invade Earth

   The agendas for past NAUG meetings are archived at:
   http://www.aoc.nrao.edu/~smyers/naug/agenda/

   The minutes for past NAUG meetings are archived at:
   http://www.aoc.nrao.edu/~smyers/naug/minutes/