Attendees: Ken Moreland (SNL), Chris Sewell (LANL), Kewei Lu (LANL), Pat Fasel (LANL), Rob Maynard (Kitware), Sujin Philip (Kitware), Chunming Chen (Kitware), Tom Fogal (NVIDIA), Jeremy Meredith (ORNL)
Chris is finishing up his contour tree work with Hamish Carr. So far the implementation has been in directly Thrust. Converting to VTK-m should be straightforward, but he is still deciding on whether and when to make the move.
Pat is working on a tetrahedralization filter. As the data model gets fleshed out, This work will need to become aligned with that.
Kewei is implementing stream surfaces. He currently has an issue with not being able to reallocate arrays such that memory is preserved. We have talked in the past about implementing this, but there is currently no ETA. We also discussed other alternatives to reallocating the array. The current implementation allocates a new array, appends new points, then calls sort. There was some discussion that there might be a more efficient way to copy the data directly to a new array so that there is only one copy rather than an implicit copy (from reallocation) and a sort (which is one of the slowest operations in the device adapter).