GitLab

re-factor the sorts activation into a single function to avoid repetition. turns out this is where the drift-before-sort bug was hiding, i.e. the missing drift activation in the for-loop.

Updated the way we converge towards the smoothing length. We now use a Newton-Raphson scheme. This applies to the Gadget-2 scheme.

Modified the splitting rule for self tasks to be less conservative. Added proper documentation of the splitting conditions.

Also create functions to decide when to rebuild and when to split a task. Update the rebuild and split condition to correctly use kernel_gamma.

Add new function to decide on recursion with in a sub_pair task. Makes sure all recursion tests are identical.

change sort task flags to be zero by default and only switch on those that are needed, avoids pre-emptively computing all the sorts for a cell since they may never be needed between two sort cleanups.