Date - 2000.06.02
Tester - S.T. Myers (AOC)
Platform - Linux (kernow)
Version - Weekly (1.4 build 91)

Note: bugs are denoted by lines with prefix:
>>>BUG
and queries are denoted by lines with prefix:
>>>QUERY

-------------------------------------------------------------------------------

GOAL: Use map tool to make and calibrate a simple image.  Check look
and feel of map tool.

1. Activate map construction from Tool Manager.  Use spanner menu to
   input measurement set 19991219X.ms (polarization calibration data)
   and "send" to Tool Manager. Create.

   Look at the "solvecal" function menu.  It does appear that I can
   bring up the various managers though the respective spanner menus.

2. Bring up "CalibrationlistManager" under "calibration" entry.
   --> Bring up "CalibrationManager" under "calibration1" of list manager:
       --> t=10 sec
           table=19991219X.caltable
           select='FIELD_NAME=="1331+305"'

   Actually, at this point go back to CatalogGUI and browse the
   measurement set 19991219X.ms as I forget what is in it.

>>>BUG: I went to the "WINDOW" Menu of the Tool Manager to try and
   find the CatalogGUI that I had previously made and then
   minimized. When I clicked on the "dc" entry it brought up another.
   The first one is not listed in this window (even after refresh)
   even though it was created from the spanner menu on "map".  The
   WINDOW list should include all tools created by aips++.

   Returning to what I was doing after looking at FIELD table:

   --> calibration1 = CalibrationManager
       --> t=10 sec
           table=19991219X.caltable
           select='FIELD_NAME=="1331+305"'
           CREATE

   Did this do anything? Nothing got entered in the "calibration1" box

           SEND

   Now there is something there... <calibration>

       CREATE
       SEND
       calibration1 = <calibrationlist>

3. Now tackle the "selection" entry of solvecal:

   --> selection = SelectionManager
       --> freqsel = FreqselManager
           --> nchan = 1
               start = 1
               step = 1
               CREATE
               SEND
           freqsel = <freqsel>
       --> fieldnames = ['1331+305']
       --> spwids = [1]
       CREATE
       SEND
       selection = <selection>

4. Now the "sourcemodels":

   --> sourcemodels = ModellistManager
       --> source1 = '1331+305'
           model1 = ModelManager
           --> catalog (try from Perley-Taylor 95)
               CREATE
               SEND
           model1 = <model>
       CREATE
       SEND
       sourcemodels = <modellist>

   I have no idea if this worked!

5. Finally the "solvers"

   --> solvers = SolverlistManager
       --> solver1 = SolverManager
           --> t = 10
               phaseonly = False
               table = 19991219X.soltable
               append = False
               CREATE
               SEND
           solver1 = <solver>
           CREATE
           SEND
       solvers = <solverlist>

6. Now we GO

   It seems to be doing stuff, like selecting, but is going through
   all sources?  Lots of "Field: not found" errors.  I JUST WANTED
   3C286 which is in the data!  Is doing the first source 0423-013
   and not converging, so try to ABORT...

   I have a feeling I lost my map tool, and its still running.  Wait
   a while...

   Its still running!  I may have to abort out of aips++ to stop it.
   I see no other way.

   BANG!

>>>BUG: THERE *MUST* BE A CLEANER WAY TO ABORT A WAYWARD TASK.
   SOMETHING FROM THE TOOLMANAGER WOULD BE NICE.

7. Post Mortem

   Although I did not get as far as making a map or finishing
   calibration even, I have seen enough to be worried about the
   direction this is taking.  If anything, this is almost more
   complicated than just going through the steps manually!  I think
   that this "wizard" approach may not work for us in the end, as
   the process is not linear enough to work well as a series process.
   Or maybe the problem is that not enough indication is given in the
   menus on the order in which things should be done.

   I will try to think on possible better approaches over the weekend
   (I have actually been thinking about this over the past month).