[ODE] plans for higher order integrators?

Matthew K Grimes mkg at cs.nyu.edu
Sun Oct 12 18:50:49 MST 2003


I recently had to decide between Dynamechs and ODE for a project in
simulation control. The project will probably require high-accuracy
integration, so I chose Dynamechs, but not without alot of looking over
my shoulder, as ODE seems to be more fully featured and actively
supported. Can somebody (Russ?) tell me if there are short/mid-term
plans to make higher-order (e.g. Runge-Kutta 4-5) integrators available
in future versions? If they are, I'll want to make some extra effort to
keep my project modular with respect to its dynamics engine, so I can
swap out Dynamechs for ODE when that feature's implemented.

Also, it seems like there are a few on this list who, not needing high
accuracy, have switched from Dynamechs to ODE. Are there any pointers
that come to mind, things that you would've written differently in your
Dynamechs app if you'd known beforehand that you'd eventually switch to ODE?

Thanks,
-- Matt




More information about the ODE mailing list