October 30, 2003
Visual Studio .NET Installation
Hi contacted the Computer Science help desk yesterday asking about a Visual Studio .NET installation. I can get a copy of the software and install it on my own machine, but only if I reimage my office workstation to Windows. Right now it is Linux because that provides me with remote access.
The other alternative was to use the one department system that was supposed to have Visual Studio .NET installed on it, but I found out that it was not installed after all. The original system at that location was replaced with one of the new small form factor boxes, and did not have Visual Studio .NET installed. Help has filed a trouble-ticket for this and should get it installed soon.
Until I can access Visual Studio .NET, I'm kind of stuck as far as the Adaptable Video work goes.
October 28, 2003
I haven't been able to get much of anywhere on the System.Security.VerificationException error I'm running into with the MPEG-2 code. I do know that it has nothing to do with the Adaptable Video code, or the specific MPEG video file I'm using, as it still occurs with SimpleMPEGParser.exe and multiple video files. I also found that it appears to be a result of C# not being able to verify the type returned by MPEG2Event.IntraQuantiserMatrix.getDefault(), based on the code at verify_call.c:1088.
I logged onto #dotgnu on irc.freenode.net and ended up talking to someone named leppie. At his/her request, I sent a copy of the SimpleMPEGParser.exe file and somehow the error is related to a T/D436 struct (???).
October 22, 2003
Security Verfication Failure
However, now that I'm working around that, I've run into a different problem with C#'s security model. I suppose it's good for there to be a security model, but I have no idea why it's working the way it is working. Other than external data cannot be trusted from anywhere, given all the stupid kinds of security holes that Microsoft products suffer from. I'm getting this exception when it tries to return a static matrix:
Uncaught exception: System.Security.VerificationException: Could not verify the code
at MPEG2Event.Macroblock.getNext(BitStream, SequenceHeader, SequenceExtension, PictureHeader, PictureCodingExtension, QuantMatrixExtension, DCPredictor, DCPredictor, DCPredictor, Int32, IntPredictor, Int32&) in ./src/Macroblock.cs:170
I'll have to read up more on the unsafe code "feature" of C#.
I've got a call scheduled with Ketan tomorrow morning instead of our usual Thursday afternoon, since Fall Break begins tomorrow at 5pm. I have managed to compile the AdaptableVideo C# code, but am now running into runtime errors. Kind of weird because they are IO errors and I'm not sure what is causing the problem.
October 17, 2003
New Graphs + More Debugging
Ketan and I had our weekly phone call this Thursday afternoon. He's sent me an updated version of the PVR paper for submission to SPIE I need to redo the graphs so they look nicer, and find an updated reference for the bibliography. Since the submission date is the 27th, I need to get this to Ketan by mid-next week.
Ketan also asked me to start looking at putting together an API to the Adaptable Video code so that other applications can make use of it. For example, an API to perform descriptor comparisons and reconstruct one video from two.
Other than that I've been doing some more debugging on my current Adaptable Video code.
October 1, 2003
Ketan and I had a brief conversation today via phone.
As far as my research work goes, I think reachable goals for this semester are frame and coefficient separation and reconstruction, but I'm not confident about completing more than that. Ketan thinks we could write up a hook into some P2P distribution system like Kazaa.
I've also decided that I'm going to try and finish my IP first and then concentrate on the Adaptable Video research. The IP is more important for my graduation, and it also means I can work on one thing at a time, instead of having to multitask so much.