Moose Habitat Surveyor Project

Good news, I will continue working on this project over the summer as a CS498 project. As I work toward a deliverable milestone for the end of the semester, I’m working on an MSI installer and testing it on the HP Tablet PC’s (which doesn’t have the “Microsoft DirectX 9.0 SDK (October 2005)”) installed. The installer works fine and seems to register the CPolhemus COM object, but when the app launches I got a File not found exception. At first I thought it was some magic strings I had for loading textures. I refactored all of those to load the texture files from resource streams. I still got the exception after a number of different installations. So, finally I threw a try catch around Main() to get the details of the IO exception. It turns out that the program was failing to load the Direct3DX assembly. I’m not the only one. It seems like a pretty complicated subject. After I install the redistributable that comes bundled with the SDK “C:\Program Files\Microsoft DirectX 9.0 SDK (October 2005)\Redist” onto the target HP machine, then the application launches fine.

I also used the the Dependency Walker (which I remember having it’s own homepage, but which seems to be dead) from the Visual Studio 6 suite of tools. It’s mentioned in What is a DLL?. This let me pin down exactly what runtime DLL’s I needed for polhemus’ PDI.dll which include PiCmdIfc.dll, msvcp60.dll, msvcirt.dll, and msvcrt.dll. I bundle all those DLL’s in the installer. The drivers still have to be installed on the target machine, but the polhemus host and pdi software does not. I might just bundle the drivers for GP.

I gave a final presentation to my peers and instructor today. It was a powerpoint overhead and a video. We had 15 minutes. I came in at 14 minutes. The powerpoint gives a good synopses of the project. If you look closely at the video you see that it is really spliced together. But it covers most of the salient points from the powerpoint and hence the project as a whole. Next week there will be a demonstration day in the computer science lab to show off to faculty and other students. I’ll keep the video playing in a loop and bring all the props for people to play.

Now there’s one week left before finals. The things that seem to be top priority are fixing bugs that are sadly only informally (and ephemerally) defined in my head. For example, I know for sure that the PlantPatchRenderer is measuring the diameter in cenimeters when it should be millimeters. There must be a user manual. I’d _really_ like to have a test/video session in the field with Dr. Spalinger. Plus, there is a laundry list of feature requests _including_ some fundamental changes to the model. We’ll just have to see what get’s prioritized.

It’s hard to resist the idea of working on what has been a very fun and rewarding project over the summer. I don’t feel like I’m going to get a that rare solid sense of closure. I must admit I’m jealous of the next generation of upper division CS students that get to work with this project. Onward and Upward!

Following my project plan, I’m 91% complete which is right on time. It’s nice to be on schedule Dr. Spalinger build a “faux” bush by putting a dead branch in a base platform of his own design and construction. He also has a pole mount for the polhemus device that can be pushed into the ground. MHS is starting to feel like a real life field project!
Read the rest of this entry »