Present: Maria Hamilton, Peter Diener, Karim Shahzad, Roland Haas, Steve Brandt, Atul Kedia, Bilal Hameed, Kenneth Sible, Miguel Garcia, Yosef Zlochower, Frank Loffler, Zach Etienne, David Bradley
Agenda
* Obtaining old releases via GetComponents
* Peter - issue #2488 reported by Frank Loffler, not present * Steve - problem with svn.einsteingtoolkit.org too old to work. What is needed to make old thorns to work, is (1) update svn server, or (2) port the thorns. Upgrading svn is the easier to do. * Roland - proposes rewrite of the thornlists from the old repository, to correct them to use the git checkout and point to the correct repository in github, and turn off svn server.
* Kuibit toolkit * Peter - Kuibit toolkit announcement email * Roland - Python package for post-processing simulation data, will be an upcoming presentation * Peter- delegates person to invite Gabriele Bozzola, the package developer.
* SPEC benchmark * Peter - SPEC benchmark contribution thorn choice status * Steve - administrative issues with SPEC contribution: who receives the money, how to pay a student, what code to be proposed to include * Roland: proposed to be received as a gift to an account of the University. Choice of thorn in work, compiling McLaughlin, if not will send a survey with the 3 codes suggested, one will be proposed for inclusion in the SPEC benchmark
* Obtaining old releases via GetComponents again * Frank - online now, explains the issue, suggests alternatives. * Roland - reiterates that is not a simple search and replace. Thorns were moved between branches and repositories. * Frank - proposes to put on a student on it, and have it tested that the old versions work. * Peter - student in mind? No suggestion. * Steve - why don't just submit it as a patch if the fix is easy. * Frank - more complicated, should effort be put in upgrading the server, or instead upgrading the old thorn list, and let server die. * Steve - mailing list will have to suffer if we don't maintain the old server * Frank - less likely to break than the svn stuff on the old server. Would be nice to clean it up and get rid of svn.
* Unanswered questions on the mailing list * Peter - unanswered questions on the mailing list, error messages in the makefile * Roland: fix email question empty repository * Peter - ETK 2010 download error * Frank - first release after moving to github. Fix: point GetComponents correctly, remove http links * Peter - who will send the email? move to next - instability par file * Roland - we do not have it, need to write code to produce ID * Steve: reply email no one has such thing
* Open tickets * Peter: anything to discuss * Zach - too late to propose new thorns? He is the release manager * Terrence new thorn Maxwell equations 2 thorns: ID and evolution ADM like and BSSN-like. generate thorns and put them in WVUThorns arrangement. CarpetX testing. Between Scalar Wave and BSSN. * Update from Ken - parses BSSN equations, still need to double check with BSSN eqs not parsed.
* Old tickets ready for review * Steve: has a pair of scripts (1) check what Cactus needs, (2) check the SPEC configurations, should be in utilities directory.
* Minute taker and chair * Jan 28 Miguel - minutes, & Zach - chair * Next week: Yosef- chair, Steve - minute
* SimFactory2 ignoring threads * Roland: - deliberate: reports as 8 cores, but they are not physical cores? Not limited by that choice and ignore the hyperthread? * Steve - procs is nr. processes, it will be divided by the num threads?
Maria _____________ Maria C. Babiuc Hamilton, Ph.D. Professor, Department of Physics College of Science, Marshall University, 1 John Marshall Drive, Huntington, WV, 25755 Room S 257, Phone: (304)696-2754