IHEP: Y. Dong, Z.H. Li
SDU: X. Huang
INFN: G. Ambrosi F. Gargano, F. De Palma, M. Duranti, V. Formato, N. Mori, L. Pacini
(+ many other people, please report any missing name)
- N. Mori, V. Formato and L. Pacini gave presentations about the architecture, status infrastructure and current usage of HerdSoftware
- X. Huang gave a presentation about the HERDOS prototype based on SNiPER and HEP packages (DD4HEP, PODIO)
- M. Duranti pointed out that having two different solutions then it would be useful to start comparing them on quantitative basis (speed, output size etc.). It is not clear if we are ready for this at the moment.
- N. Mori asked if DD4HEP allows for small tweaks in the geometry without having to write an entire XML file. X. Huang confirmed that it is possible.
- X. Huang asked about the usage in HerdSoftware of two different frameworks (GGS and eventAnalysis) for doing simulation and data processing, and asked if it is possible to merge them. N. Mori answered that this situation is due tyo historical reasons, and that it is in principle possible to merge the two but up to now there has been no need for this, so there are no plans at the moment.
- X. Huang asked why EventAnalysis use text files for job configuration, which is a rather old approach compared to modern approaches like using python. N. Mori answered that it is a design choice since configuration scripts written in a programming language tend to be cluttered with computation which are not essential to the job configuration; however, a configuration system for EventAnalysis based on Root scripts is present but is currently disabled due to bugs, and can be resumed quickly if needed. Python could also be integrated in the configuration system, but it would require time and it's not planned at the moment.
A ddodle poll will be prepared by N. Mori to have another meeting in about two weeks, depending on the amount of new things to be discussed.