Present: Roland, Steve, Yosef, Leo, Sam, Zach
thorns proposed for inclusion =============================
* no updates
Subcyling in CarpetX ====================
* Yosef and Allen find very slow speeds for BBH simulations * Alexandru Dima and Helvi are finding that comparing Lean_public and Lean_public in CarpetX that subcycling is responsible for the majority of slowdown (factor of 6 or so) * Roland suggests to use that as
Subscribe to NR community call ==============================
* no objections
Users mailing list moderation =============================
* currently Peter is the only person moderating users@einsteintoolkit.org * also handles subscription requests * Yosef suggested adding more persons as moderators and add it to the ET call agenda * had some discussion on how to try and improve this process
Unanswered emails =================
* https://lists.einsteintoolkit.org/pipermail/users/2023-July/008978.html Roland will respond
Tickets =======
* https://bitbucket.org/einsteintoolkit/tickets/issues/2669/math-is-not-render... Zach suggests to also provide PDF as fallback on the website
* https://bitbucket.org/einsteintoolkit/tickets/issues/2647/incorrect-weno-coe... Roland to check what existing reconstruction methods in GRhydro are actually doing (interpol or reconstruct)
* https://bitbucket.org/einsteintoolkit/tickets/issues/2739/change-admmacros-s... needs review, code from 2022 and review request from April 2023
* https://bitbucket.org/einsteintoolkit/tickets/issues/2526/mclachlan-has-inco... needs some internal Kranc knowledge, volunteer needed
* https://bitbucket.org/einsteintoolkit/tickets/issues/2653/carpet-presync-tri... had some discussion on where checks for "already valid" happen in Carpet & Presync. Sam did observe that the changes proposed sped things up. Should be re-investigated to see if this still applies.
* https://bitbucket.org/einsteintoolkit/tickets/issues/2637/update-cactus-copy... needs review, mostly finishes up a prior change to README files
* https://bitbucket.org/einsteintoolkit/tickets/issues/963/improve-mclachlan-a... discussion is going on between Peter and Zach where Zach proposes a better and more effective solution to the issue. Hope to revisit this after summer.
Chair 2 weeks from now: Steve Minute take 2 weeks from now: Yosef
Yours, Roland
On Thu, Jul 13, 2023 at 11:02 AM Roland Haas rhaas@illinois.edu wrote:
[...]
Subcyling in CarpetX
- Yosef and Allen find very slow speeds for BBH simulations
- Alexandru Dima and Helvi are finding that comparing Lean_public and Lean_public in CarpetX that subcycling is responsible for the majority of slowdown (factor of 6 or so)
- Roland suggests to use that as
Roland, what do you suggest? The sentence ends here.
-erik
Hello Erik, all,
- Roland suggests to use that as
Roland, what do you suggest? The sentence ends here.
Ah, I was going to suggest this as a topic for the working workshop / hackathon at RIT end of July. Since there are are least 2 groups interested (RIT and UIUC). So forming a group of say 3-4 persons, might work well.
Yours, Roland
users@lists.einsteintoolkit.org