I tried some runs with fish looking in a wider radius for cells they could move to (fishMoveDistParamA changed from 50 to 500 or so) and the code gradually used up all the memory. (May go away if we adopt the R-tree method for finding neighbor cells?)
Actually it may be active memory use, not a memory leak. Once all the RAM was filled up, the processor slowed down... so it probably was not unused garbage filing RAM.
I tried some runs with fish looking in a wider radius for cells they could move to (fishMoveDistParamA changed from 50 to 500 or so) and the code gradually used up all the memory. (May go away if we adopt the R-tree method for finding neighbor cells?)