This week I have been concentrating on writing the disertation. So far I have written the following sections

Research
Analysis (but needs reworking)
Design
Implementation

I hope to write up some of the testing tonight as well as go back to the analysis and improve it.

I think it would be best if tomorrows meeting runs as usual so that what should go in each section of the report can be discused to clarify everything in my mind and so that I can attempt to finish the first draft over the weekend.

This week I have been making slow progress with the report writing and have gotten the 64bit version of the viewer working so that I can profile it in th labs.

Geometry Instancing

Following the meeting with Tyrone this morning I have been evaluating the work required to integrate geometry instancing in to my real-time application. It currently looks like it will require a quite significant juggling of my current draw code to de-layer the way the various LODs are drawn. This is a design flaw of the original implementation and I would like to fix it as it will make the whole system far more flexable but at the moment I am not sure that time will allow for it.

I will move forward with fixing the remaining issues with automatic LOD biasing and add support for automatic distance biasing today before proceeding to the report writing.

Today I have cleaned up the last few issues with the code and made sure all the components interact correctly. I have also created a table of contents for the report along with target word counts for each section to help me focus, hopefully this will get approved in the meeting tomorrow.

I still need to create scripts for the new assets but that is just to improve the visual effectiveness of the product so will be done when I next get some spare time.

Overall im very pleased with the progress this week and I feel my product is far more complete and presentable.

Generator Complete

In another successful day I have managed to write the applet for creating the scripts used to control the offline generation of the discrete LOD levels. It also includes features for testing the model in the viewer by automatically creating a world containing the model and lauching the world as a command line parameter to the viewer.

Viewer Complete

Today marks the completion of the viewing application from a programming standpoint. All known issues have been fixed and it has been polished to a glow. All that remains to do with it is add more model variety.



Hopefully I will be able to get the generator setup app written tomorrow.

Todo List

This week I need to complete the following tasks:
Generator:
Create applet to make simplification scripts.

Viewer:
Add depth biasing.
Add target frame-rate controls.
Add resizing/full screen support.
Fix remaining bugs.
Add more models/variety to the scene.
General polish.

Report:
Contents page and general sections layout.