symba implementation
Hallo! In my project using fargOCA, I am comparing hydro to N-body simulations of planets capturing in resonance. This is a very interesting question currently for the (exo)planet community. I have run into some kind of bottleneck, which most likely derives from the limited N-body solver built into fargOCA (whose implementation in fargOCA is also not very optimised apparently). A much more powerful choice would be a symplectic integrator, such as swift_symba, which is one of the most robust and most commonly used in these contexts. Using symba inside fargOCA will definitely make this study much more self-consistent and allow for a very direct comparison with the typical setup that is used in a lot of papers, which is very exciting. It would be a great way to advertise the code!
All we really need essentially is to link the symba library so that symba functions can be called inside fargOCA. A few years ago Elena (who also agrees that this would be a very nice upgrade to the code) has combined an old fargo version with symba, so she already knows what we need to do once the library is linked. So Elena and I could take care of that part.
For now, we can easily avoid portability issues and just implement this on a branch so that Elena and I can work on it without messing with the master branch. In the long run, this upgrade will easily pay off and make fargOCA pretty competitive, since a very robust N-body solver would make it perfect not only for the kind of things I am doing, but also for different setups like when you have close encounters between the planets, which is something that the current Runge-Kutta solver can't really deal with.
@alainm what do you think?