-
1.6
add ability to initialize particles in clouds well, cubic clouds. different initial shapes will be added as they are needed, but the basic structure is here now.
-
1.5
Add elements for more thorough postprocessing using C++ codes. I *think* there's no need to move to 2.0 right now.
-
1.4
add a better version of distributed particles
-
1.3
add distributed particle solver. I haven't extensively tested the new code, but it should be reasonably stable. With this version, a bunch of fancy C++ features are being used, more than before, so a series of efficiency tests are to follow, and then the shiny new features are to be thrown out if they prove costly... I do believe the use of base classes makes maintaining the code an easier job though.
-
1.28f51d992 · update todo ·
v1.2 Particle data is now kept in a separate .h5 file. This sort of change is on the border of a MAJOR version change, but I figured that core functionality for children of NavierStokes is not actually affected, only possible particle postprocessing stuff, and in a negligible way.
-
1.1
Generate consistently named executables When using the .launch() method, the structure of the code is more or less fixed, therefore it makes sense to keep the executable name consistent, as it can be safely reused for any simname. Full control over the executable name is kept when just using .run(), since different species of particles may be added for instance, thus making different codes incompatible.
-
1.0.1
Bugfix release. Executable scripts should handle source fields correctly now.
-
1.0
production ready version Well, I've been using it for production work for a while now, so I may as well make it official. Anyway, I suppose there are still many bugs, but I don't see them.