Minutes from SSS Mtg 2008-Jul-08

Attendees: B.Butler, M.Claussen, D.Harland,    B.Hesman, G.v.Moorsel,
           F.Owen,   M.Rupen,    L.Sjouwerman, B.Truitt, S.Witz

Dave reiterated that the SSS programmers would be concentrating less on usability issues and more on functionality for this calendar quarter, and probably the next as well. The two most important items this quarter are:

1. Producing execution scripts from the data gathered in the OPT that correctly command EVLA antennas and the VLA correlator.
2. Supporting testing of the prototype WIDAR correlator.

Scientists should continue to report any usability issues, but send their findings to Lorant. Lorant will filter these, passing important issues to the programmers, and accumulating those that can be worked on after the planned functionality has been added.

* * *

Brian gave a summary of the status of the generation of execution scripts.

1. We need feedback on whether what we're producing right now is correct or incorrect. Of special concern are those items that impact TelCal and MCAF.
2. Doppler tracking has not been addressed.
3. LO/IF setup should be tested.

* * *

Bryan mentioned that we may need a stand-alone version of model-to-script (m2s), in order to support dynamic scheduling. Many of the values that m2s calculates are point-in-time dependent. The velocity of a source relative to the VLA is one such value, but there are others. Bryan said that we need to sort things out with Barry.

* * *

Bryan, Barry, Ken, and Walter B. are the best candidates for testing the execution scripts generated by our code. Bryan mentioned that we have a program, clexec, that is a simulator for the executor. We probably want to use this program to help with our testing. The output of clexec really requires human inspection in order to understand the outcome of the test.

Gustaaf and Frazer suggested that we add a "validate" button to the OPT that could run clexec and check the quality of the script. Due to the nature of the output from clexec, this may or may not be something we want users to see, though giving users the comfort of knowing their set up and our script generation code will produce a successful observing run is an excellent idea. The programming staff will investigate this possibility late in this calendar quarter after becoming more familiar with clexec.

Bryan suggested demonstrating the use of clexec to the SSS programmers.

* * *

The group had a long discussion on how to handle Doppler tracking. We discussed both short term needs (Ka observing with the VLA) and how, or even whether, we would be doing Doppler tracking with the EVLA and WIDAR. In the short term several approaches are possible, each with merits and drawbacks. Rather than come to any conclusions in this meeting, the scientists will discuss the issue among themselves and then let the programmers know how to alter the user interface to acquire the additional data needed from observers who want to use Doppler tracking.