One plugin later, and real-time volume rendering of CT-data is working. I told you your hair'd stand on end!

Wheeeeeeee! The plugin system now has double the hair-raising ability. It does all kinds of cute things with plugin dependencies and object pools.

Tomorrow (and Saturday) we're having demos. That means no work will get done, especially since my workstation is going to be part of the frikking demonstration.

In spite of all the administrative and political chores which had to be taken care of, RealWork(tm) won by a great margin. I have formalised my Python/VTK plugin system and can now do things with Python/VTK/C++ interaction which will make your hair stand on end and make me very happy.

This weekend will be dedicated to relaxation, which logically does not exclude liberal amounts of CasualCoding(tm).

I've been busy, very busy. You can download an anti-aliasing patch for XJed by clicking here and you can get an idea of what XJed looks like with anti-aliased font-rendering here.

Otherwise, the meetings just never stop. Fortunately they are interspersed with very interesting colloquia. Still, the net result of all this meeting and listening is that RealWork(tm) has to wait...

My camera now has 8 (2 sets) of NiMH batteries. It is now a happy camera.

I went to visit LKEB today. These guys are doing some seriously cool stuff...

We (office-mate and I) had some demonstrations here at informatics for some of the project leaders. They did look impressed.

I also built some potato versions of my latest unstable jed packages. The reports for vtkFlRenderWindowInteractor are coming in, and it seems that it works perfectly on that other toy operating system.

Whoooooooo! My camera just arrived (quick huh?). I'm happy with that.

Yesterday a whole bunch of us went to KU Leuven and to Materialise in Belgium for an academic visit. It was absolutely worth it... these people are doing some really wonderful things with rapid prototyping (3d printing for the luddites amongst my fans).

The DC3400 won the competition, primarily because of its good linux support and the fact that it has a USB port for communication with a host PC. The Oly only has serial and I couldn't find out for sure whether there was linux support. It does have a very good lens, but the Kodak is considerably cheaper and matches the Olympus very well featurewise. I went ahead and ordered directly from Kodak; if everything goes according to plan, I should have the camera by Tuesday. Yaaaaaaay!

I almost bought an Olympus C990Z (US: D490Z) yesterday, but the stupid shop didn't have any in stock and I didn't feel like haggling about the display model. This seems like a very solid 2MP camera and if reviews can be believed, its only con is that it doesn't have USB. At the moment it's between this and the KODAK DC3400, which also seems like a good buy and is even a bit cheaper than the Olympus.

Come and get it, fresh from the oven! The vtk-fltk thing is at v0.2, muchos thingies fixed and improved. It's weekend. Overall the week has been quite productive, but it's still frightening that it's over. This weekend we're planning some serious shopping...

VTK is building on the Octane; as soon as it's done, I can test vtkFlRenderWindowInteractor on it. It seems there might be problems on Solaris as well, but that I can unfortunately not test (yet). And then there's still the little issue of testing it on windows. Yuck.

I've branched off from FlvtkInteractorWindow (but it keeps my fixes) and started vtkFlRenderWindowInteractor. I have done this so that my ideals of keeping things as simple as possible can be realised without hindrance. Click here for more information.

I rule. Actually mpatrol rules. All your bugs are belong to us! dscas1 is now seemingly rock-solid. In addition, I've upgraded to 2.4.2-ac20 and nvidia drivers 0.9.767... these seem fine, and I've now enabled AGP again. On this SMP work machine, 0.9.6 refused to work stably on a 2.4.x kernel with AGP enabled.

Thanks to the mpatrol memory allocation checker (and specifically its --dynamic mode) the bug has had its little head squashed. It turns out that it was hiding in a very old vtk class that I'd written *mumble mumble*.

These past two days have NOT been nice (due to this very elusive bug) but on the bright side, FlvtkInteractorWindow is now cleaner than ever before (all remotely potential problem areas were fixed) and this will shortly be available from my website.

AAAAARRRRRRGGGGGGGGHHHHHHHHHH I've been chasing the same irritating bug (in one form or the other) for the past 2 days and I'm getting really tired. If tomorrow is a normal working day (please please please) I might get it squashed and then get my life back to normal. Bugs have a strange effect on me.

Yesterday was a bad day for public transport. Down day! Sit! I spent 3 hours on trains and buses for a presentation of around 1.5 hours... at least the presentation (on fMRI) was really good.

I'm currently still spending some time on dscas1: mostly some cleanups and refinements so that I can hand it over to the FE boys to start playing with. Note to world: volumetric data is difficult to work with, purely because of the raw amound of data that one has to shuffle around.

We went to Cascades Japanese restaurant in the Hague last night. The food was absolutely divine and we'll definitely go there again in the future.

dscas1 (my pet cas platform) is getting along very nicely. I had to divert my attention temporarily from the python plugin system in order to finish some work on dscas1 that's been scheduled for Tuesday. Watch this space for screenshots.

Yesssss!!! It works! With the help of David Gobbi, I have a proof of concept of python vtk pipelines integrating seamlessly with c++ vtk pipelines. Interpreted run-time modifiable plugins, here we come!