Users April 2026

users@lists.einsteintoolkit.org

7 participants
19 discussions

Meeting Minutes for 2026-04-09
by Lucas Timotheo Sanches 09 Apr '26

09 Apr '26

Attendees: Peter Diener, Lucas T. Sanches (minutes), Steven R Brandt (chair), Roland, Jordan Nicoules, Cheng Hsin Cheng, Maxwell Rizzo, Nikolai Wallin. # Pre agenda topics Because Jordan joined, we started discussing issue 2927. He found a bug with GCC versions. Steve thinks that this is due to a bug when CST gets generated into code. He says he should be able to find the problem. We also discussed issue 2912, due to the Llama inconsistent volume forms. Jordan says he has a fix for the issue though, so he promised to send a PR to fix it soon. # ET_2026_05 release ## Gallery runs - The BH-NS gallery example is done now. - No other gallery runs have started. ## Test suite runners - Anvil and Sunrise are giving problems on multiple thorns: 1. CCE export. Peter looked at the failures. Two HDF5 are generated but the test suite obviously cannot compare binaries, so it relies on a python script. It's likely that Python is the root issue here. 2. CanudaX NP Scalars 3. CottonmouthBSSNOK. TwoPuncturesX is failing everywhere. Steve says the tests should be updated. Some tests require matter, which we don't have for CarpetX on the release. ## Thorn Reviews - BHaHAHA: Peter wants to create a par file with it included so he can understand all the options. Zach says he would like to chat about it and Peter wants to include it in SPHINCS BSSN. - Boost: Peter tried to download and put it in his thornlist, but it is failing. Roland says we should actually decide what we are going to provide from boost. So far, it seems we need System, file system, math, parameter parsing, and FUKA also needs python, but that will be a headache. - Cottonmouth: Lucas presented some prelim. results of a new Z4c recipe. Steve stated our intention to have this thorn proposed for inclusion. Steve also presented timing results. # ET Workshop at UIUC - Registration is open. - Speakers have been invited. Many have responded. # Atlassian ticket migration - No updates as of yet, we still have time. # Kyoto initiative - IllinoisGRMHD had volunteers on their behalf. Zach is going to meet with them to discuss details. They may join next ETK call # Mailing list questions - Bug in QuasiLocalMeasures: Jordan will let his collaborator make a ticket. - Range error setting parameter, on cluster: Jordan discussed this during the start of the call. A ticket will be made about it # Tickets - 2928: Steve asked the community's opinion on adding a gpu option to simfactory. Lucas and Cheng-Hsin would like that. - 2923: Maybe upgrading ADIOS2 fixes the issue, but Cheng-Hsin had not time to test it yet. - 2910: Apply. - 2921: Cheng-Hsin has further comments, he will add it.

1 0

Einstein Toolkit Meeting Reminder
by users＠einsteintoolkit.org 08 Apr '26

08 Apr '26

Hello, Please consider joining the weekly Einstein Toolkit phone call at 9:00 am US central time on Thursdays. For details on how to connect and what agenda items are to be discussed, use the link below. https://docs.einsteintoolkit.org/et-docs/Main_Page#Weekly_Users_Call --The Maintainers

1 0

Agenda for Thursday's Meeting
by users＠einsteintoolkit.org 06 Apr '26

06 Apr '26

Please update the Wiki with agenda items for Thursday's meeting. Thanks! https://docs.einsteintoolkit.org/et-docs/meeting_agenda --The Maintainers

1 0

Bug in QuasiLocalMeasures
by Marco Brito 02 Apr '26

02 Apr '26

Good afternoon, I found a bug in the QuasiLocalMeasures thorn, in the computation of the ADM momentum. In the source code file qlm_analyse.F90, in line 361 there is a plus sign instead of a minus sign, contradicting the ADM momentum formula: P^i_{ADM} = \frac{1}{8\pi}\lim_{r\to\infty} \oint_S (K^i_j - \delta^i_j K)dS^j [1] As expected for slices where K \neq 0 one was getting a wrong result. The correction is easy, just replace the plus sign by a minus sign. In any case you can find a patch file in attachment. Should I in any case file a ticket or a Pull Request somewhere regarding this? Best regards, Marco Brito. [1] M. Alcubierre, Introduction to 3+1 Numerical Relativity, International Series of Monographs on Physics (Oxford University Press, Oxford, 2008).

1 0

Meeting minutes for 2026-03-26
by Nikolai Wallin 02 Apr '26

02 Apr '26

Release: -Cottonmouth in master, Doc directory is non existent. This needs to be fixed. Currently the fastest according to Steve. -Two Punctures X should be added. Currently added to master -Probably need to update the CI system to ensure numpy is installed. Currently failing -Overall release is making progress, we need to get the reviews out of the way. -Need to add some work to the website front page. -Peter is reviewing BHaHAHA -Steve is reviewing Gallery- think thats done -Lucas and Beyhan reviewing knudaX -Lucas has started review -Cottonmouth - Zac and Deborah will review, Zac has done a preliminary AI overview -Roland will review KapperX -Boost - Steve is reviewer -Two Punctures X Zac is reviewer -Round of testing can now begin in earnest, we have the complete set of thorns in place -Steve is trying to run on Frank(enstein) to test on many types of GPUs -Apptainer no longer needs root. Announcements: ET workshop is June 15-18 -Travel grants will not work via reimbursement Atlassian is retiring Bitbucket Cloud Issues- Everything will shut down by August -Jira or github? -Zach is going to test moving some into a dummy github page -Should try and keep issue name the same -If not done by August 1st we’ll switch to a Jira migration Kyoto initiative -A few hydro codes will be represented -Jay and Zac are already involved for AsterX and IllinoisGRMHD (CPU) -Beyhan voiced interest in representing GRHydro alongside these two -Steve would also like to contribute -We will revisit this around once a month to keep up to date Unanswered questions on the mailing list - None Tickets: 2929: Loss of convergence with subcycling in time 2909: Remove TAGS from interface.ccl 2923: Adios2 compilation -Steve replied has not heard back. Should already be fixed 1775: -In the works, but we’re limited to versions -Its rather large being 200 MB of source code 2921: -No updates from Lucas, waiting for review 2922: -Seems to be a bug in cuda 12.9 -Only a very hacky solution works

1 0

Meeting minutes for 2026-04-02
by Steven Brandt 02 Apr '26

02 Apr '26

April 2 Present: Peter, Steve, Roland, Lucas, Deborah, Beyhan, Maxwell Rizzo, Cheng-Hsin, Zach, Nikolai Wallin Chair: Deborah Minutes: Steve Release seems to be on track The TOV Gallery is updated Previous one is too big because too much was uploaded Need to link BH-NS collision to the gallery page TwopuncturesX is having trouble Cheng-Hsin says the Teukolsky-Lean test--fails on 2 procs, but not on 2 threads Is output written in consistent order? Cottonmouth - Split into 5 parts to obtain optimal, Zach asks whether these can run in parallel? Zach says that makes the point that higher order often needs different splittings Peter asked about the CPU performance CanudaX review, Lucas has started, needs doc folder (as does Cottonmouth). BHaHAHA review, Peter says it has a lot of good documentation New Gallery--all done, ready to go Boost--Roland has a script to reduce the tarball to 30M or so. Boost has its own build system, bjam. Hooked into make's parallel build system. Considered splitting it into multiple tarballs. Newest boost doesn't have boost::math::tools::eps_tolerance, so can't use it. Workshop and School SOC meeting this afternoon Atlassian retiring bitbucket issues thingy Zach will move that in May Kyoto initiative comments? Who should we ask for updates? Zach, Philip Moesta, Beyhan, and Jay are involved. Zach is happy to be the Kyoto initiative liason. Unanswered emails: Range error email, new issue created, using a different compiler made it go away Tickets: ADIOS2 - Cheng-Hsin - looks at the build logs, int is size 4. On Delta. Has had the same issue on the local Ubuntu laptop. 2909 - Reminder to Roland. "A question that sometimes drives me hazy: am I or are the others crazy?" --Einstein

1 0

Meeting minutes for 2026-03-26
by Nikolai Wallin 02 Apr '26

02 Apr '26

1 0

Einstein Toolkit Meeting Reminder
by users＠einsteintoolkit.org 01 Apr '26

01 Apr '26

1 0

Range error setting parameter, on cluster
by Jordan Nicoules 01 Apr '26

01 Apr '26

Dear all, I've encountered the error below while trying to run a parameter file on a cluster. Surprisingly, for the same parameter file (and the same version of ET: 2024-05) on my local work station, this does not occur. This appears very early in the execution, at parameter setting. WARNING[L2,P0] (Cactus): ParameterSetReal: Unable to set real 'Coordinates::h_radial_1' = '4' not in any active range WARNING[L1,P0] (Cactus): Major error in parameter file '/home/jnicoules/Work/ET/Simus/parfiles/HairyBH/HairyBH_sol02_MLBSSN_llama_run02bis.par' line 94: Range error setting parameter 'Coordinates::h_radial_1' to '4' In the param.ccl of the Coordinate thorn, the range of this parameter appears as * :: "negative turns off stretching" Changing the * range statement to *:* makes it work on the cluster. Is this some known issue? Is there something specific to look for, in the Cactus configuration for instance? So far, I haven't thoroughly tested ET versions, different clusters etc, but I can try to provide more information if needed. Best, Jordan Nicoules

2 10

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

Users April 2026