Minutes for the Einstein Toolkit Meeting, 20250731. https://docs.einsteintoolkit.org/et-docs/Main_Page#Weekly_Users_Call 9am CDT Thursdays
Present: Peter D, Leo W, Bill G, Deborah F, Maxwell R, Johnny T, Beyhan K, Hector I, Keith D, Yosef Z
Chair Leo, Minutes Bill
* BBH Gallery example (Zach and Peter) Peter, still not yet done, been too busy.
* Unanswered Questions on the mailing list https://www.einsteintoolkit.org/tools/unanswered.php
** 25 July SimFactory placeholder macros. Confusion with the @XXX@ macros and where they are set and changed.
** 23 July Share Your Feedback... Seems that it should not be in our mailing list, refers to software and network tools(?).
* Open Tickets https://bitbucket.org/einsteintoolkit/tickets/issues?status=open&status=...
#2773, CarpetX thorndoc confused in InterLatex.pl. Assigned to Steve.
#2867, WaveToyX examples have HDF5 output error. Not assigned.
#2866, Problem building ET on MN5: OpenCL (Marenostrum 5). Not assigned.
#2764, PUGH tries to free memory not allocated by malloc. Assigned to Beyhan K. More work to be done.
#963, Improve McLachlan accuracy. Original post 2012. Peter, we are trying to compare and see how it compares with NS simulations as well.
* Tickets for Review https://bitbucket.org/einsteintoolkit/tickets/issues?status=new&status=o...
Not much discussion.
* Any Other Business
** Leo, question maybe for Steve, does anyone use the ET in clusters but through apptainer or singularity images? Steve does this. Leo is trying to do that and working with his cluster folks. Bill is also interested in that. Leo trying to setup with base layer that is operating system and then compilers and then external libraries, say external Lorene, and last a thornlist or par file. The last image would behave as ET would. Has it working locally but when tries with MPI on cluster with multiple nodes it hangs, problem with internode communication. Peter, Steve has been able to run MPI jobs with containers...maybe using the right MPI tools with Slurm to run them.
** Johnny, question also for Hector, variables needed for CCE Exporter. Want to use the CCE Export thorn, but do not write the time derivatives Lapse and Shift. Deborah, MacLachlan does write it back. Lean at Illinois, folks were working on something similar, Roland knows how to do this. Something about the time stepping. Hector, tried adding tensors to the Kranc file and stored them, whatever was done to write them to ADMBase, but see NaNs even though the evolution code does run normally. Maxwell, Baikal has a similar problem with dt_alpha and use the right hand side, some subtlety on the iteration, because what was saved is a fourth-order RK step. https://einsteintoolkit.org/thornguide/EinsteinAnalysis/CCE_Export/documenta...
** When is the developer hack-a-thon? Peter, maybe in two weeks.
Next Meeting, Thursday, 7 August 2025. https://docs.einsteintoolkit.org/et-docs/Meeting_agenda
fyi, bill gabella