[Users] Meeting Minutes 2021-05-27
Gabella, William E
b.gabella at Vanderbilt.Edu
Thu May 27 11:19:22 CDT 2021
Present: Atul K, Bill G, Steve B, Yosef Z, Karima S, Peter D,
Leonardo W, Zach E, Ken S, Maria BH, Miguel G, Roland H
Chair: Atul K, Minutes: Bill G
* Before agenda items? (like quick help questions, etc)
* SPEC benchmark contribution
[https://bitbucket.org/einsteintoolkit/tickets/issues/2470 task] [RH, SB]
Steve, Roland asked SPEC how things are going, they outlined what more
they want from
us. They want to be able adjust workload to keep a 40 core Haswell
machine busy for
many minutes. They could integrate Cactus fine, adjust config.h for
...they will do that. When testing is done, will start on step 2
proposal, workload is
step 3. Their goal is for Serial max memory 1.8 GB and run for 8 min,
for Parallel 40
threads 8 GB and 10 mins. They are asking for advice on adjusting
Nothing is urgent but there are things to be done. See above ticket, #2470.
* Release Planning [ZE]
(Hendrik) Lorentz is the release name.
Zach, It is coming along. Release Announcement draft has been shared among
Simfactory should work with Python 2 or 3, compatible with Python 3
stated in the
announcement. List of changes since last release should be double
Steve, Maintainers database exists inside the repo. Zach sees
scripts to turn that into other formats. Developers.py creates HTML,
will also read developers.txt . Release helpers should go into this
file for the current
release and taken out for next release. Handling this file should be in
Planning, etc. See http://einsteintoolkit.org/developers.html and
Zach has been updating the Release Planning documentation
https://docs.einsteintoolkit.org/et-docs/Release_Process . One advice
is to word wrap
at line 72 for the email formatting, he added an online tool to do that.
Branches created tomorrow, so any minor changes have to happen today.
Do we use SVN? Steve, maybe. Roland, we do not have any Subversion repos
anymore, some checkout function does use them.
Atul, Zenodo ticket #2524 should be addressed
Zach, Zenodo may not be updated by May 31st. But the DOI from them is
part of the
bibtex key we ask people to cite. Nice if there, but not checked out
for new release,
so can be updated later.
Zach, Will edit the main ET page, http://einsteintoolkit.org/index.html
. Remove the
ICERM announcement, keep North American Workshop, and Release Announcement
in the center.
If you are part of the release team, gallery example checkers included,
your names at
** Gallery Examples [ZE]
Steve, it runs and finished a bit early, so will run it to the end---ran
out of time.
Working with Kuibit to generate the plots vs Mathematica. Waveforms and
Peter, re-ran it last night and still works fine. Will update web pages
Bill, Runs fine with and without Presync. Updated the Bitbucket / web
pages, had to
remind myself how to upload the tarball, what credentials to use (create
a APP key on
bitbucket but use my username not SSO email, go figure).
Atul, Done but not finished the Bitbucket updates.
Poisson equation] [MG]
Miguel, finished the test and it was fine. Upload the tarballs and
files, but need to
update the links on the web page.
Atul, We should re-run the examples again with the new Release...in June
Maria, did look over the documentation for it and wrote something on
Roland should check it out. Roland, some items in the ticket #2413 are
related and not Power related---will clarify in the documentation.
Maria, Power works
with BBH but not so easy with BNS. Roland, you can run it with BNS but
need to specify
more items. Atul, add in documentation how to change the frequency with
and about quasi-local measure. Roland, want to avoid too much
running ET in the Power documentation.
bugs of severity major or worse]
Steve, #2533 is not fixed yet. If two ActiveThorns that is the problem,
but need better
Yosef, #2534 Mol never worked well with Grid Arrays, but had a code that
did work and
it currently does not. Has not worked with 2020 releases, did in the
issues had missing header and maybe Presync problem. Steve, if we do
fix it we need
to add a test for this issue/problem so we do not break it again.
open tickets] flagged for the next release
#2526 is not a blocker and will not be ready by the release.
#2469 Leonardo, I iwll push hard on this today.
#2488 Atul, has a minor tag so not critical right now.
Zach, Last commits should be minor and not change functionality and
should happen today.
Will work with Steve and Roland on the Zenodo, etc. But please read the
Announcement soon, want to finish it.
Bill, we should mark the tickets (which are listed as Tasks and not
Tickets) for the gallery
examples as resolved when you are through with them. They will be
re-assigned for the next release.
question on mailing list]
Seem to be handled.
open tickets sorted by update time]
#2530, Roland has some ideas to fix it, not high priority and will not
do anything before
the release to avoid breaking anything.
tickets ready for review]
Six tickets most still need reviewers. Please volunteer.
#963, Peter, did look at it. It is a patch to an old version of
McLachlan that is no longer
current. Has to wait until after the release.
Chair and Minute taker for next weeks:
June 3, chair Steven B, minutes Miguel G
June 10, chair Peter D, minutes Atul K
Research Assistant Professor
Department of Physics and Astronomy
Nashville, TN USA
More information about the Users