GlueX Software Meeting, January 22, 2019
GlueX Offline Software Meeting
Tuesday, January 22, 2019
3:00 pm EST
JLab: CEBAF Center A110
BlueJeans: 968 592 007
- 1 Agenda
- 2 Minutes
- 2.1 Announcements
- 2.2 Review of minutes from the January 8 meeting
- 2.3 Report from the January 15 HDGeant4 Meeting
- 2.4 HDDS Version Management
- 2.5 Task Review
- 2.6 PrimEx Reconstruction and CERNLIB/Fortran
- 2.7 NERSC Running
- 2.8 Review of recent issues and pull requests
- 2.9 Review of recent discussion on the GlueX Software Help List
- 2.10 Manpower Note
- Review of minutes from the January 8 meeting (all)
- Report from the January 15 HDGeant4 Meeting (all)
- Task Review (all)
- Review of recent issues and pull requests:
- Review of recent discussion on the GlueX Software Help List (all)
- Action Item Review (all)
Talks can be deposited in the directory
/group/halld/www/halldweb/html/talks/2019 on the JLab CUE. This directory is accessible from the web at https://halldweb.jlab.org/talks/2019/ .
- CMU: Naomi Jarvis
- FIU: Mahmoud Kamel
- JLab: Alexander Austregesilo, Thomas Britton, Mark Dalton, Sean Dobbs, Ashley Ernst, Stuart Fegan, Mark Ito (chair), David Lawrence, Dmitry Romanov, Simon Taylor, Beni Zihlmann
There is a recording of this meeting on the BlueJeans site. Use your JLab credentials to access it.
- gluex_install 2.9. Mark went over the new features in the latest build-it-all-from-scratch package.
- Several new packages releases, new version set file: version_4.0.0.xml. Mark reviewed the release of new DIRC-enabled software packages.
Review of minutes from the January 8 meeting
We went over the minutes.
- Thomas reports that the revamped MCwrapper is getting a lot of use. A new feature has been introduced: old projects can be copied as a starting point for composing new projects.
- David has revived an old mechanism for updating a database with event range information for raw data files. These data are now being collected as new data comes in. The strategy for back-filling information from old runs is being planned.
- Justin Stevens's merging of the DIRC branches onto the master branch seems to have gone well.
- Debug and optimized builds of halld_recon and halld_sim have been deployed in the build of the latest version set at JLab.
Report from the January 15 HDGeant4 Meeting
We went over these minutes as well.
- All known issues with HDGeant4 have been resolved. There remains a known problem with FDC hit population at the wire-based stage of tracking, but that does not seem to be associated with the simulation code.
- Thomas is going to re-do the studies that led to the first batch of issues with the new code base. We will see how we do this time.
HDDS Version Management
Sean raised several questions about which versions of the geometry we should be using in various contexts. There was some discussion of this over the weekend: Issue #49 of HDDS: TOF plexiglass insert compatibility with DIRC?.
The Lucite shielding for the TOF has been merged onto the master branch, thus the master branch contains both the DIRC and the shielding. The shielding geometry has not been tested in a real simulation; that needs to be done. In the mean time we would like to use a version of the geometry that excludes the shielding. That exists in the 4.0.0 tag of HDDS. Sean will take care of putting this version into the CCDB for the upcoming runs with the DIRC even though the shielding is physically there. When the shielding geometry has been tested, it can be added to the CCDB.
There is still the question of how to include the shielding for pre-DIRC runs (those already taken). Mark suggested we create a branch off of the tag 3.13.0, the last tag before addition of the DIRC, and add the shielding there and tag it. That will allow use of hdgeant with this geometry and provide a source for a file to be added to CCDB where appropriate.
Ashley will work with Mark to create the branch. Sean will take care of the additions to the CCDB.
We reviewed the software work packages. There are two relevant categories: Analysis Software and Software Infrastructure. Mark will summarize status of Software Infrastructure at tomorrow's Analysis Meeting.
PrimEx Reconstruction and CERNLIB/Fortran
David raised the issue of new reconstruction code that PrimEx-η collaborators would like to add. The set of routines are in Fortran and rely on CERNLIB. We need to provide a path for making this addition quickly since the PrimEx run is upon us. However there is a problem with introducing CERNLIB-dependent code into halld_recon; we have just spent a ton of effort to eliminate dependence of halld_sim on CERNLIB by developing HDGeant4. CERNLIB is no longer supported and we would not like to have our code dependent on it going forward.
Mark suggested that the code be added to a branch of halld_recon, keeping the master branch free of CERNLIB dependence. There is an effort to re-write the relevant code in C or C++ making it compatible with the rest of our code base. At that point, the routines can be merged onto master.
David reported that the start of the reconstruction pass at NERSC on Spring 2018 data is waiting for the final pieces of calibration to be put into place.
He also mentioned that it is not necessary that the entire run be run at NERSC. Some of the running could be done at JLab if cycles are available.
Review of recent issues and pull requests
We reviewed the following pages:
The most notable item was the fix to halld_sim to turn on energy deposition for hadrons in the FCAL. Due to an indexing error, only electromagnetic particles were being registered. See halld_sim pull request #34 for a bit more detail.
Review of recent discussion on the GlueX Software Help List
We looked at the list of topics without significant comment.
Dmitry mentioned that he has completed a chunk of EIC work and he may have more cycles these days for work on CCDB, RCDB, and other Hall-D related topics.