You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Currently swimmer reactions does not properly handle ghosts and pair discovery: It does it's own n2 pair finding but does not take into account the facts that pairs can be found multiple times, that some of the reacting particles could be ghosts and does not synchronize reactions between nodes. Also the implementation is rather confusing and the details are not documented. In the current state we can not keep it.
The text was updated successfully, but these errors were encountered:
The test case fails without ROTATION. It is also not clear to me, what it does without ROTATION. @jdgraaf@hmenke, is the SWIMMER_REACTIONS feature actually being used?
Based on the issues mentioned in this ticket. I think that, either someone has to commit to clean it up or we will have to remove it.
2886: Swimmer reactions r=KaiSzuttor a=fweik
Fixes#2498.
This was broken in multiple ways (see #2498) and seems abandoned.
Also It uses an interface (access to verlet lists and cells) that we do not
plan to provide in the future.
Description of changes:
- Remove SWIMMER_REACTSION feature and functionality
Co-authored-by: Florian Weik <[email protected]>
Co-authored-by: Kai Szuttor <[email protected]>
Currently swimmer reactions does not properly handle ghosts and pair discovery: It does it's own n2 pair finding but does not take into account the facts that pairs can be found multiple times, that some of the reacting particles could be ghosts and does not synchronize reactions between nodes. Also the implementation is rather confusing and the details are not documented. In the current state we can not keep it.
The text was updated successfully, but these errors were encountered: