Analysis Mtg 240429
From Moller Wiki
Back to Main Page >> Analysis Software overview >> Analysis Meetings
previous meeting << >> following meeting
Meeting information
Agenda
- Updates
- General discussion
Minutes
Participants: R. Conaway, P. King, J. Mammei, A. Sen, W. Deconinck, E. King, P. Gautam
- Ryan shows his current draft of the guide to the root files. We will get this linked to github and to the agenda soon
- Wouter suggests that the document should actually be made as a markdown file within the github repo
- Ryan will circulate the current version via the email list after the meeting
- Wouter brings up the question of having japan-MOLLER as a fork of japan: if people have exisiting forks of japan, they cannot make a new fork of japan-MOLLER. By breaking the fork relationship between japan and japan-MOLLER, people could then make new forks for further development.
- For development workflow, we do prefer having people working within forks, so breaking the fork relationship is useful.
- Wouter raises the question of removing (or reducing) the boost dependency in favor of the C++ std libraries
- Most of the boost libraries (with the exception of program_options) we use have been incorporated within the newer C++ standards, so the boost libraries are no longer under active development. So transitioning to the std libraries should be more robust.
- Wouter suggests choosing C++17 or maybe even C++20 as the minimal standard. Wouter points out that ROOT 6.30 requires C++17, for example.
- Prakash has done some investigation of the RNTuple [1]
- See root.cern/blog/rntuple-update
- With a simple example code, Prakash has found an rntuple is substantially smaller than the corresponding TTree.
- Prakash had some communication with a ROOT developer: RNTuple might be in production root version in late 24 or early 25
- There will be a presentation about "Data Preservation recommendation presentation from Jlab perspective" at the MOLLER simulation meeting tomorrow at 10am.