[Users] meeting minutes for 2023-06-01

[Roland Haas]

Present: Zach, Keith, Roland, Steve,  Leo, Peter

SPECv8 contribution
===================

* no specific news, progressing 

ET release coordinator search
=============================

* had planned to include CarpetX and AsterX
* no documentation exists for it yet
* Roland estimates that CarpetX is not closer to inclusion ready than 6
  months ago. CarpetX reviewers 
* paired coordinator approach was useful for Leo in the 2023_05 release
* will pick this up again next week due to small number of participants

Unanswered question
===================

*
  https://lists.einsteintoolkit.org/pipermail/users/2023-May/008930.html
  Roland will forward similar answer about grid structure consistency

*
  https://lists.einsteintoolkit.org/pipermail/users/2023-May/008935.html 
  Roland will respond suggesting that a full C++11 stack is required

Error message shown about missing
http://lists.einsteintoolkit.org/pipermail/users/2023-June/date.html is likely due there not yet being a posting to the mailing list for June and the script cannot distinguish "no posting yet" from "server is down or cannot access the file".

Galley examples
===============

* Steve will update the BBH example
  (https://www.einsteintoolkit.org/gallery/bbh/index.html)
** Zach mentioned ticket
  https://bitbucket.org/einsteintoolkit/tickets/issues/2706/increase-twopunctures-resolution-in-bbh
* Peter had added convergence plot to TOV example
  (https://www.einsteintoolkit.org/gallery/ns/index.html). Convergence
  order is between 1 and 2.

Zach brought up the issue that gallery examples are typically very low
resolution and one often has to increase resolution significantly
observed for example good convergence. This typically clashes with the
desire to keep the gallery examples small enough so that they can be
easily run.

Zach noted that the TOV example is not featured in the cycler on the
main page. Steve and Zach had a discussion about how one could create a
more visually interesting image using VisIt.

Steve brought up work on openPMD reader in Python. Roland will work
with summer student in summer to add mesh refined openPMD support to
VisIt. Will meet with Steve and (NCSA) local visualization expert.

Updated tickets
===============

*
  https://bitbucket.org/einsteintoolkit/tickets/issues/2734/consider-adding-if-statement-functionality
** Steve mentioned current situation in CarpetX where storage is always
   one.
** Steve considers that he may be able to add auto-allocation of
   storage to the WRITE statements, READ would result in invalid data to
   read from

*
  https://bitbucket.org/einsteintoolkit/tickets/issues/2733/twopunctures-contains-globally-visible
** we should fix this


Open topics
===========

* Clang and Flang support. Zach mentioned benchmarks that have clang
  outperform gcc, asks about support for clang and flang in simfactory
  and the Einstein Toolkit.
** Roland says no clang support in the past  due to lack of Fortran
   compiler. Steve and Zach point out that gfortran can be used and the
   there is now also flang.
** Roland mentions that there are "old" flang based on PGI and
   "new" flang based on llvm. Also AMD compiler and Cray compiler on
   Crusher / Frontier are llvm based, so are modern version of the Intel
   compiler suite.
** currently have not knowledge about clang / flang in the Linux
   known-architecture files, only in Darwin files for (older) clang

Next Chair: Steve
Next Minutes Taker: Leo

Yours,
Roland
-- 
My email is as private as my paper mail. I therefore support encrypting
and signing email messages. Get my PGP key from http://pgp.mit.edu .
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 833 bytes
Desc: OpenPGP digital signature
URL: <http://lists.einsteintoolkit.org/pipermail/users/attachments/20230601/9a291a3b/attachment.sig>