Attendees: Lucas (minutes), Yosef, Roland, Noora, Peter, Steve (chair),Beyhan, Nikolai, Johnny Tsao.

# Review of the ET 2026 Summer School.

- It was held last week at UIUC. - Helvi thanks all local organizers and members of the scientific and organized comite. - 40 people in person, 30-40 people online. - Mix of invited talks and lectures, about 30 minutes each. - Tutorials on CarpetX and the ETK. - Tutorial on Spectre. - Tutorial server was at capacity, which is good, this means good engagement. - We had sessions on different NR codes, - We had talks on career paths outside of academia. - The UIUC organizers will send out emails to presenters asking for permissions for publishing the talk videos. - Next summer school: Organized by Zach Etienne, University of Idaho, summer 2027 - Helvi suggests we have more discussion on using LLMs for NR. - Helvi suggests we distribute short talks by early career scientists better throughout the day. - Lucas comments he likes having tutorials on different codes and how to use them. - Helvi and Roland emphasize for the seniors to reach out to the younglings and talk to them. They suggest we spread out the seniors and the younglings to go out to lunch, for instance, so there can be networking and conversations. - Noora said there was good feedback on having different code presentations. - Beyhan says that presenters should probably have the data ready when presenting. Roland says we could give users complete notebooks, with the notebook output (not final data output) so that users can read them off as presentations are going. - Helvi suggests that we recommend presenters to have backup data, with complete results. Steve agrees. - Roland suggests we give the presenter some special slurm powers, so that they can actually run their demos and not get blocked by the students on the queue. - Beyhan suggests including some of the features that we don't have on the ETK from other codes.

# ET_2026_05 release

- There is now a release announcement. - Steve remarked that the list may be incomplete because he forgot to treat the NRpy repo differently, possibly excluding contributors from there. - Roland pointed out that it is not clear what some of the names have contributed. Steve says he will review his Claude procedure for including names. - Beyhan points out DIRAC/Cosma 8 is missing among supported machines.

## Gallery runs

- Deborah is seeing some issues when running the GW150914 gallery example. The black holes shot to the side, instead of orbiting. Steve believes he knows the cause, which is related to the way that TwoPunctures handles and updates its parameters. His unification of TwoPunctures and TwoPuncturesX have introduced this issue. Yosef suggests adding a regression testing to the suite to make sure it is caught. Steve agrees. - Noora says her example is running.

## Test suite results

-`advanced_features` failing on Anvil and Sunrise - Peter says he will rerun on sunrise. - Zach says he has identified an issue in Nrpy elliptic and Leo is already on it. - Noora says that she will provide the testsuite results for Frontera shortly. She reports two failing tests, `SBH_AxiDilHairID_test (from AxiDil_Evol)` and `sphere_pugh_ppm (from Hydro_InitExcision)`

## Inclusion votes

- CanudaX voted YES for inclusion. - Cottonmouth voted YES for inclusion.

## Unanswered questions on the mailing list

- None that need responses