When I checked my repository, I noticed that some of the code went missing for almost no apparent reason.
I reasoned that during my clumsy attempt to resolve conflicts in the code, I accidentally deleted a section of the code.
The newest PR adds the code back in with the formatting changes.
Also, speaking about formatting changes, now Github properly recognizes that the header file is intended for use in C (though, it should work in C++ also.)
The results of the very first dynamical friction test are in, and... neither dwarf galaxy was all that effected by dynamical friction. That is, running the simulations with dynamical friction look pretty similar to running them without. But that doesn't mean much yet. I left enough wiggle room open through my approximations that I still can't be sure dynamical friction is unimportant. So now it's time to refine those approximations. And the first place I think we might be losing a lot of dynamical friction are those infinite spikes. That is, we might be jumping over them too hard. Professor Newberg had an interesting idea for how to treat the spikes that I'm going to try, which should hopefully tell us how important an infinite density spike in a caustic ring is.
Oh, did I promise videos in the last post? Since the simulations are almost identical with and without dynamical friction, I'm just going to post one video for along-the-flow friction and one for against-it friction. If you're curious why the videos end before they reach 10 Gyrs, that's because they weren't supposed to go for 10 Gyrs in the first place. I let them end early (actually my computer somehow got unplugged and then ran out of power part-way through visualization) because by the point the video ends the approximation that stars are perfectly bound to the dwarf galaxy has already become horribly wrong.
Videos: http://youtu.be/LxmR9dzr3rU http://youtu.be/y4N9AafOn3w (As before, purple is dark matter and yellow is luminous matter.)
Over the past week or two, we've been working on a rewrite of o2dtk (eek!). There are a few reasons why we decided to do this:
So the project is now formally known as Lead (Pb) and we aim to make a suite of tools for developing all sorts of things with Unity.