I’m finally getting my notes together from the user group meeting this week. I really enjoyed being able to meet in person with people who I only knew online from the newsgroup. Thanks to Bill Okubo, Harold Cline, and the rest of ITT VIS for putting together a great event.

I wasn’t as diligent as I wanted to be about taking photos, but I did upload a couple to Flickr.

Congratulations to Ken Bowman for winning the Image Contest!

So here are some notes about the talks:

Keynote address: “Sweatin’ the Small Stuff: Visualizations of an Unseen World Using IDL” by Rob Dimeo, Ph.D. (NIST Center for Neutron Research, Gaithersburg, MD)

Rob gave a great talk about some interesting problems encountered while doing neutron scattering. He has a book coming out soon, Simple Quantum Visualizations in IDL; check his website for the free download in the future. Starting a trend followed by many of the other presenters, Rob outlined a few suggestions on what he would like to see in IDL:

  1. Do not abandon iTools, but do improve documentation and training.
  2. Focus on interoperability with other software packages.

“HDF Update” by Elena Purmal (The HDF Group, Champaign, IL)

Elena talked about tighter integration between the HDF Group and ITT VIS on new releases of HDF 5. I am personally experiencing the pain with this as our main simulation software is moving to HDF5 1.8 and IDL is stuck back in HDF5 1.6.3 (but HDF5 1.6.7 if you go download it from the ITT VIS website!). She also mentioned OPeNDAP, but had to run to another meeting right after her talk before I talk to her more about it.

“IDL at LASP” by Chris Pankratz and Karen Simmons (LASP, Boulder, CO)

Karen Simmons presented an amazing history of IDL starting even before David Stern. She even produced a paper tape containing one of the first routines in what would later become IDL. Chris Pankratz focused on the (many) later space missions controlled at LASP.

“A Look Ahead: IDL 7.1 and Beyond” by Bill Okubo and Chris Torrence (ITT VIS, Boulder, CO)

Bill and Chris laid out the plans for IDL until 2012 (although it got significantly vaguer after IDL 7.1). Highlights for IDL 7.1: the visualization manager introduced into the IDL Workbench, an iTools procedural API, and Windows command line. IDL 7.2 was labeled “visualization plugins, part 2”, IDL 7.3 with “Applet builder and applet UI tools”, and 2012 with “enhancements for programmers,” but there weren’t too many details for out that far.

Slither: The IDL to Python Bridge” by Jason Ferrara (Jacquette Consulting)

This was a very cool demo of essentially a Python bridge, nearly identical in functionality to the Java bridge. Slither is a commercial product that allows calling Python code from within IDL (unfortunately, I think I’m going to need the other direction). This is good news that the IDL export bridge technology can support an outside developer creating something like this.

GPULib” by Peter Messmer, Ph.D. (Tech-X Corporation, Boulder, CO)

Peter outlined the capabilities of GPULib, an IDL library for using your CUDA-enabled graphics card to achieve performance gains of up to 100x speed ups. With a few $1500 graphics cards, it is possible to create a desktop computer that would have placed on the top 500 supercomputers five years ago.

New Horizons Geometry Visualizer” by Henry Throop (SWRI, Boulder, CO)

This talk was the winner of the 2008 IDL Application Contest. The New Horizons Geometry Visualizer is a web based application using IDL to do the calculations for visualizing what the New Horizons spacecraft (destination Pluto and beyond) is looking at. This was a cool applicaton, but I think the main point was that IDL doesn’t provide much help for developing applications on the web.

“A Large IDL Application” by Richard Azuah (NIST, Gaithersburg, MD)

Richard talked about his experience creating an iTools application. His suggestions for improving IDL:

  1. The widget toolkit is dated: need “worksheets” i.e. a WIDGET_TABLE that works better, rich text support, and cut and paste that works on all platforms.
  2. Need better font support for mathematical symbols.
  3. Need better control of the default components added to a new iTool (when creating a new type of iTool).
  4. Improve iTools Developers Guide.

“Object Lessons: Confessions of a Direct Graphics Junkie” by David Fanning (Proprietor of Coyote’s Guide to IDL Programming)

David gave a great explanation of how to create a class definition and how he uses objects to quickly create visualizations by combining objects representing different displays of data (I would say this was putting an object graphics interface onto direct graphics, but I think David would probably cringe at that). This lead up to his announcement that he was open sourcing his Catalyst library (content coming soon!). I also heard my first coyote story, very entertaining!

“Discoveries at the Edge of the Solar System” by Marc Buie (SWRI, Boulder, CO)

Marc gave a great talk on finding Kuiper belt objects using IDL software to aid human processing of images. Library routines mentioned are available at his website. His suggestions for IDL:

  1. Improve keystroke support in draw widgets.
  2. Stabilize the format for publishing integrated help; it’s not worth learning how to do it if it changes quickly to a different format

“IDL Obsession” by Ronn Kling (Kling Research & Software, Marshall, VA)

Ronn talked about a series of cool topics: his AVI plugin, a hybrid Workbench perspective, N-dimensional ellipse fitting with weighting, and a new, faster method for determining whether a point was inside a convex polygon. I thought the new method for points in a polygon was really cool, but I was also interested that people are starting to share some of the many Workbench settings that are export-able.

Revolution IDL: 3D Visualization” by Eduardo Iturrate (ITT VIS, Boulder, CO)

Eduardo demoed a cool tool that I’ve mentioned here before. This really came alive with Eduardo behind the controls. I think this has a lot of potential, especially (but not limited to!) as a learning tool since you can see the source code of a scene that you create interactively.

Slides of the presenters who agree to have their slides distributed will be posted. UPDATED: here they are. Also see David Fanning’s summary.