Reasons: which command is different in Debian and in CentOS/RockyLinux (alias + extra options)
=> This lead to trouble when vae23/vae24 started with scrontab from bare-metal virgo3
=> command -v is the advised standard (especially on bash)
=> See also https://unix.stackexchange.com/a/85250

After the lxir116 went out of operation there were no runner for the CI
merge stage any longer. Cretae a new runner on run4 which will execute
this stage instead of lxir116.
Most of the changes were done on the GitLab server and the runner node.
With this MR only the environment variables are adapted.

Sergei Zharko authored 2 weeks ago and Sergei Zharko committed 1 week ago

- V0FinderCutter class for the analysis cuts (+ config)
- Updates to offline V0Finder: new cut application scheme
- Updates to offline V0FinderQa: track and particle histograms on different statges
- Updates to CA QA classes: several bugfixes and extra track classes for mCBM (2STS + 1TOF, primary/secondary/all)
- Updates to mcbm_reco_event_L1.C: (beamtime/mcbm2025) for reconstructing digi-events stored in *.rra files
- Updates to tracking parameters for mCBM (TODO: provide common file in the parameters repo)

bba: output alignment matrices for STS,TRD,TOF in the required format; autodetect corresponding geometry paths

Pierre-Alain Loizeau authored 1 month ago and Pierre-Alain Loizeau committed 1 week ago

=> Similar methodology as porting of the standard pulser channel "full stack" monitor

Pierre-Alain Loizeau authored 1 month ago and Pierre-Alain Loizeau committed 1 week ago

=> Done by adding a derived class to the existing monitor overloading the general histos handling
=> Need to add virtual qualifier to some method of existing monitor in order to keep calls in algo generic

Simon Neuhaus authored 2 weeks ago and Florian Uhlig committed 2 weeks ago

Redesign of the mirror holders
	Replace old MirrorSupportBelts by new MirrorFixing which looks like the CAD model and the prototype
Add a small plate at end of mirror support studs to fit the weight
Fixing all overlaps in the geometry
Fixing newly discovered bug in implementation of exit window
	change of geometry hierarchy at exit window
	similar change at entrance window to avoid potential of similar bug
Additionally small changes (small regarding physics impact)
	changes in the volume hierarchy to reduce the depth
	changes in code of creator file to make it more efficient
	replacing some hard coded numbers by variables
	replacing double definitions of same volumes by copying

Souvik Chattopadhyay authored 1 month ago and Sergey Gorbunov committed 1 month ago

changed the function names to capital letter
changed the installation script
changed the installation of cbmroot inside run_ctest_commit.sh
modified Application.cxx so that if run_ctest_commit.sh fails it returns
deleted configEditor class and optimized the
config file added
Applied clang formatting
licence header added
Yaml files are updated and a newline was added
documentation was done for all the fucntions and variables

Souvik Chattopadhyay authored 1 month ago and Sergey Gorbunov committed 1 month ago

removed hist_compare_api binary
applied clang-format
applied clang-format after rebaseing
removed macro/qa/run_commit_out.sh
updated the function name with upper case in configEditor.h
updated all .cxx and .h file with licence header
removed artifacts of the rebase in analysis/PWGDIL/dielectron/conversion/CbmAnaConversion.cxx and sim/detectors/fsd/CbmFsdDigitize.cxx
applied clang-format and added external lib which was deleted

Souvik Chattopadhyay authored 2 months ago and Sergey Gorbunov committed 1 month ago

shell script which can automatically change from one commit to another commit and create the output Ctest root file