To edit this file or the web pages, please do: svn checkout https://github.com/pencil-code/website/trunk/www Topics for Pencil code User Meeting in Graz 2023: ============================================================== % Notes: "-" means that this item is still open (nothing has happened) % "+" means that it is already in progress (people working on it) % "*" same, but high priority % "/" postponed % "v" problem solved % "x" means that it has been dealt with % [assigned Persons] feel free to add yourself % # http://pencil-code.nordita.org/UserMeetings/2023/agenda.txt - promote practice of making init routines additive. - Newsletter: feedback - Summary from the Pencil Code Steering Committee - restart problems with param.nml being outdated (add details in manual) Find out & document what practice exactly leads to problems - Status of HDF5: why is it slow, and how slow is it? ---------------------------------------------------------------------------- OLD ITEMS: - review of auto-test sensitivities (accuracies too high) AM1, resolved for now. - reference.out versus reference.out.double AM1, write documentation on how it works (Philippe, todo) - updates of various data bases (pencil-code/license/developers.txt and https://www.nordita.org/~brandenb/pencil-code/PCSC/committers/list.pdf, for example) AM1, discussed, but no action taken - explain in manual how multiple special modules work AM1, discussed (Matthias, todo after the meeting) - Phython script for multiple simulations to scan parameter space PM1, discussed (Fred, todo: create documentation) - should we rename some more README files in sample to README.md AM2, discussed, Philippe needs to adapt 'pc_auto-test' scripts. - to add src/pc_build_details (Status; Philippe?) AM2, dicussed. Plan is to scan Makefile.local and suggest papers to cite. - user-specific initialization (Status: Philippe) *** AM2, Philippe+Matthias clarified the situation and how to continue - make pc_read_xyaver independent of proc*/dim.dat AM2, discussed, to postpone. - user-specific settings for compilation: read_all_videofiles (and what does pc_videofiles do?) AM2, discussed, postpone to next office hours, Matthias to work on that. - Documentation on how to use ParaView or YT (3D visualization tools) AM2, postponed, someone interested needs to take care. NEW ITEMS: - register_report_aux versus farray_register_pde AM1, postpone - file size in samples: EULAG sample too big AM1, partly fixed by Matthias - How to deal with >5 years inactive owners? AM2, discussed, will send them email - initialize with data partly coming from an input file => use pc_write_var to create a VAR0 and the run => implement and consolidate functions in initcond to read external file AM2, discussed with Alberto+Axel+Matthias+Philippe - guide for MAC users to install PC (XCode compilers complain) PM1, solved (Antonino writes text for the quick start guide) - HDF5: break up, slices (Fred Gent): future i/o strategies => backwards compatible or conversion tool PM1 (long-term: Fred, Philippe needs to implement new strategies) - HDF5 IO reading of persistent variables => make robust against changes in number of processors => seeds need a different treatment (depending on processors) PM1 (intermediate-term: Fred, Matthias, Philippe will fix that) - generalize scale factor reading (Philippe+Axel+Alberto+Ramkishor) PM1 (Philippe will create module with generalize profile reader) - EoS and energy module combinability: general re-design PM2 (Matthias: make energy module more general on a branch) - Output of snapshots at arbitrary times => develop a mechanism to have multiple times in a new "times_list.dat" => code checks for this file and writes out at the desired times PM2, discussed (Ramkishore: to be done) - use of ODEvars (Matthias+Axel+Alberto) PM2, discussed (Matthias: add new functions in the IO modules) - GPU acceleration feedback PM2, discussed (Matthias likes to get feedback) - How to enforce certain coding standards => how to deal with shared variables, registering of variables, etc. Fr-AM2, discussed, extend the coding style with coding standards (Matthias) - Higgsless implementation in non-relativistic code Fr-AM2, discussed (Ramkishor+Alberto: working on it) - documentation/tests of relativistic options in hydro and MHD Fr-AM2, discussed (Axel+Alberto+Ramkishor+Yutong: working on it) - early_finalize flag: to stop concurrency of processes earlier => proposal to rename to "lconcurrent" Fr-AM2 (Matthias: working on it) - explain how run PBS scripts => advantage: easier to give many options, no long commandlines => Philippe to draft for quick start guide. will be solved - How to deal with broken auto-test: Joern asked the question => PCSC should take action if nothing else works => proposal to move long-time broken autotest to "broken" section in "pc_auto-test" (Philippe: pro, Matthias: contra)