It’s been way too long since my previous update, so here finally a new update. So, what have I been doing all this time?
Well, mostly I’ve been working on developer documentation. It’s been quite a boring task, and as some sort of poor excuse, it’s been uninviting enough to “forget” about the status reports. Anyway, I’ve been shuffling content on the wiki to some extent, creating for the development contents section a new and clearer start page that should help decide and point viewers to the correct place. Three big links which go to building blender, developing blender and contacting the developers. Further some content got reorganised from different places into the wiki (from the www.blender.org typo site). There are still some redirect issues left with some of the links, but these Typo3 issues will hopefully addressed when we get an upgrade to Typo4; at least my attempts to fix those redirects haven’t been very successful. An extra reload in the browser by the user is needed to have the redirect work OK. A similar link collection I’ve created on projects.blender.org to help visitors (potential bug reporters and developers) quickly find our patch and bug trackers.
Apart from wiki work, which will continue of course, I’ve been doing the past few weeks lots of doxygen work. After testing in a few modules I worked on a script to get all our code files tagged and grouped in a structure that resembles our source base. All in all around 2400 files have now been tagged and verified and committed for doxygen usage. Even with a script to help me out with modifying the codefiles it took a long time to go through all the code and have them checked. Anyway, for those who would like to have a nice, browsable introduction to our source code can check out doxygenified blendercode on my site. The current size of this all is 455MB, and that while not all possible doxygenifications have been done! If you don’t want to browse the code through my site you can download a 84MB CHM file: blender.chm. There are CHM viewers for Linux and OSX.
The plan is to write a blog post about the doxygenification on the Blender developer blog, outlining the plans for it in the (near) future.
During the last few weeks 2010 and begin of 2011 I also worked a while on COLLADA, setting up a 2-pass mechanism that I’ll continue working on now that this step doxygenification has been completed. I wrote code for both OpenCOLLADA and Blender to handle stride settings specified in the <accessor> tag for geometry data. The patch works very well for vertices, but needs to be improved a little so it can be used easily also for all other geometry reading (normals, texture coordinates, etc.). On windows this is already secretly built into the OpenCOLLADA lib, but I’ll finalize the code and submit it as a patch to the OpenCOLLADA trackers. For normal export I also received a small library by Morten Mikkelson to help with welding geometry and normal info before doing the actual writing to a COLLADA file.
I managed to do a few bug fixes during this period too, but my coding input has been low due to the documentation effort. I’m poised to get back to coding and bugfixing again! Tomorrow the first day of the last month of my contract with Blender Foundation starts, so we’ll set some nice targets to achieve during this month. Stay tuned.