Summary Notes: IDV Steering Committee
22 February 2006

The IDV Steering Committee held another teleconference on February 22nd, 2006. Participants were: Rich Signell, Mark Laufersweiler, Mark Seefeldt, Mike Voss, Steve Aulenbach, Bill Fingerhut, Tom Whittaker, Don Murray and Jeff McWhirter.

Rich kindly offered the use of his WebEx server so that Don and Jeff could demonstrate new features in IDV version 1.3b2(1). Everyone agreed that this "desktop sharing" application was extremely useful for this purpose and that we should explore using it more. Participants were using Windows, Linux and MacOS.

A suggestion was made that the UPC should investigate hosting a WebEx server or similar cross-platform, desktop-sharing environment on its web site. (Note: from previous discussions about using such a facility for collaborations for training and support, I also encourage this.)

To begin with, Don and Jeff demonstrated and discussed three new feature areas:

  1. The Transect Control for showing multiple parameters in cross-section-like displays
  2. Synthetic time animation to allow you do define your own time steps
  3. automatic updates of data and associated displays

Work the last few months has concentrated on these three items in preparation for the T-Rex experiment participation.

The ISL (IDV Scripting Language) was also discussed and some issues clarified. Jeff raised the question of whether it would be better to use Jython for this, but the group seemed to think that a small, well-defined "language" based in an XML framework would be easier.

An issue with the Linux installer was discussed. (Subsequently, Don made a new release with a corrected version of the Java runtime.)

Geographic subsetting of grids in the properties was also reviewed as a way to allow for large amounts of gridded data to be used. This was viewed quite favorably by the committee. A suggestion was made to allow the user to type in the lat/lon bounds as well as using the rubberband box.

After realizing several of the topics discussed would and could benefit the community, the Committee decided it would be a good idea to set up a forum for the community that could be a place to post and search for "articles" containing solutions to problems and/or ideas.

The committee would like Unidata to investigate setting up forums on the Unidata web site--definitely for the IDV users but perhaps for other groups as well.

A review of the priorities list from last year was done. Don and Jeff explained progress in each of the areas where applicable.

To discover each member's high-priority items, we polled the committee:

Steve:

  1. How to reach secure servers
  2. user interface usability
  3. WRF portal developers are trying to customize the IDV GUI. Jeff said that their input is being used to make changes to the IDV framework to make it easier to customize. It still requires manual editing/configuration of XML.
    - there is some documentation but not much at this point
    - although good up-to-date user doc is needed, we discussed that most people just don't read the documentation

Mike:

  1. Interested in scripting -- will try using the ISL
  2. Catalogs -- wants to access the higher resolution grids in THREDDS top catalog on motherlode. Don explained the problems with this -- the group suggested working with THREDDS folks to make a grid-only catalog.
  3. For climate data: computing things like zonal means
  4. Saving data in a bundle
  5. Unidata needs to support the IDV more. The entire committee agreed that if IDV is going to be the future of Unidata applications for visualization and research, more staff is needed to add the necessary features and provide end-user support.

Mark S: Most of the T-Rex work has benefited him

Mark L:

  1. Apple/Mac graphics problems need to get fixed. He is putting pressure on his Apple Tech Rep -- others are encouraged to do the same.
  2. User problems with trying to read arbitrary (non CF compliant) netCDF files. Can the netCDF group develop a GUI to generate NcML that wraps "bad" files?
  3. How do we break through the barrier of people who are using other applications?

Bill:

  1. Reading old bundles with multiple view managers makes too many windows -- Jeff is working on this
  2. Would like a User Pref to set up the number/type of Displays initially without using the "default bundle" method (which "seems too permanent".

Rich:

  1. Handle memory better when stepping times (his 400x400x40xNtimes arrays)
  2. Resampling problems with staggered grids
  3. CF issues - has proposed CF standards for staggered grids.
  4. How are we going to handle "user submitted" code for the "core?"

The telecon lasted for nearly two hours, but everyone seemed quite enthusiastic about the progress the developers have made and the evolution up until now.