Hello all,

trying to get a better idea of the failing tests, I ran the comparison tool (n testsuite_results):

perl ./compare_results.pl ET_2019_03_v0 master results/bluewaters__{1_8,2_8}.log results/comet__{1_24,2_12}.log results/golub__1_10.log:results/golub__1_12.log results/golub__2_10.log:results/golub__2_12.log results/stampede2*.log

and looking at the output there are no new failing tests other than the CT_MultiLevel ones where the cause is known. There's also a Hydro_InitExcision that is now passing on golub but that test has always been tricky (it under-resolves a hydro setup) and the golub test ran on a different set of nodes on the cluster (24 core nodes rather than 10 core nodes) so I am not too surprised.

In summary there are no new unexplained failing tests on Comet, Blue Waters, Stampede2 (KNL and SKX), Golub.

Yours, Roland

Hello all,

...

Present: Bill, Chris, Maria, Peter, Roland, Yosef, Zach

ET release:

• gallery examples: Guiseppe and Helvi are running gallery examples, at least two done. Not sure if will also run BBH example, will cross check with Sam
• HMNS still under investigation, Maria in process of testing on local machines
• machines:

** mike does currently not compile and link, Peter is looking into this ** OSX/Macports fails to even install openmpi-gcc9 ** failures to compile on bethe and wheeler systems at Caltech

• status of tests:

http://einsteintoolkit.org/testsuite_results/index.php ** CT_MultiLevel tests fail on machines with more than 1 thread

• Fishbone-Moncrief: Zach is addressing reviewer comments
• Self-force code: Peter on process of securing missing code pieces for effective source terms, has license and readme file, working on issues with scons with Steve in relation to python 3.5 or higher (needs python 2, README now reflects this). other issues found are being documented on the bitbucket README file

Missing from the minutes:

• the ET release will be postponed until Oct 24th with a hard freeze on Oct 21st

Yours, Roland