Present: Peter, Steve, Roland, Erik, Leo, Sam, Zach
Minutes: Steve Chair: Peter
The release date is pushed back to Dec 6.
Update on the review of Grail (Peter looked at the test suite). There is an issue with initial data. Shock is exactly at a grid point which gives a rounding issue with intel vs. gnu.
Lucas will be doing the documentation for CarpetX in October.
The tests are not running in the test suite yet. They do get run once in a while (pull request).
Sameer Sende (friend of Steve's) offers a hardware lab with lots of different versions. We could use this for testing.
Intel 19.1.3.304 Sep 2020 on Deep Bayou fails to compile AMReX.
sgrid... is it going in?
Steve asks if we should use the MakeThornList script on the download page? Roland thinks we should add it to the new user's tutorial.
Update on Seed Magnetic Fields? Zach asks to be reminded in a week.
Test suite server has unreliable results? Still the case. Something is glitchy in detecting which tests fail. Balsara is still failing. Email is fixed. Should we email maintainers of thorns that are failing?
Unanswered questions: Peter forgot to reply to Haleigh McPherson.
Zach has been working on constraint damping and improving BSSN. Peter wants to use it to close out the ticket he is working on for McLachlan. He also wants to use it for Sphynx BSSN.
Leo fixed several WVU thorns following Roland's suggestion. Ticket should be closed.
NSIMD is not required by CarpetX, but is nice to have. NSIMD's install downloads something using requests.
2754 can probably be closed.
Do we still need HTTPD? Does it work at all?
Can have tiers of support, tier 1 is actively used, tier 3 moves to unsupported. How do people show active support, etc.
McLachlin has incorrect reads declaration. Ian doesn't have time to look at it. Roland has a patch for Kranc.
Zach says NRPy 2.0 is pip installable that works with Python 3.6 and up.
Next Chair: Leo Next Minute Taker: Roland