Computing Session at CM XXXI.

C27 (GSI)



  • Wednesday, 9 December
    • 16:30 16:50
      ...on geometry definitions & CAD converter 20m
      Speaker: Simone Bianco
    • 16:50 17:10
      MC matching with the MVD detector 20m
      Speaker: Tobias Stockmanns
    • 17:10 17:30
      Analysis model 20m
      Speaker: Matthias Steinke
    • 17:30 17:50
      Pattern recognition for the STT 20m
      Speaker: Gianluigi Boca
    • 17:50 18:10
      GEANE stability 20m
      Speaker: Alberto Rotondi
    • 18:10 18:30
      proxy dict 20m
      Speaker: Bertram Kopf
  • Thursday, 10 December
    • 14:00 14:15
      GRID: DC & production tools 15m
      Speaker: Paul Buehler
    • 14:15 14:30
      Tracking muons in Panda(Root) 15m
      Speaker: Stefano Spataro
    • 14:30 14:45
      Standalone track finding in GEMs 15m
      Speaker: Radek Karabowicz
    • 14:45 15:00
      EMC validation studies 15m
      Speaker: Vannirajan Suyam-Jothi
    • 15:00 15:15
      Simulations for the barrel DIRC: status & plans 15m
      Speaker: Dipanwita Dutta
    • 15:15 15:30
      new event generator + kaon multiplicity trigger using TOF 15m
      Speaker: Alicia Sanchez
    • 15:30 15:45
      Lambda Reconstruction with MVD 15m
      Speaker: Leyla Atar
    • 15:45 16:00
      Vertex and Kinematic fitting in Pandaroot 15m
      Speaker: Vishwajeet Jha
    • 16:00 16:30
      break 30m
    • 16:30 18:30
      Developers and users discussion 2h
      Speaker: all
      Hereby a brief report from the discussions between users and developers which corresponding event was held at the December collaboration meeting at GSI in 2009. In the past period, there has been a significant increase of new users who started digging into the new software code of PANDA for their work on PANDA. In view of these new and promising developments, a kick-off meeting was organised with the aim to optimise the communication between the existing code developers, potential new developers, and the users. The aim is to get a clear picture of the objectives and interests of the different groups with respect to the software development and its usage and to see how one could improve structurally the communication among ourselves. Many users, developers, and others, from several institutions participated in the discussion and in general the meeting was found to be very constructive. More specifically, the following items came out of the discussions: A discussion came about the existence or non-existence of a distinction between users and developers. At present, there is no clear separation between users and developers, e.g. users are expected to participate in the code development and developers are involved in using the code for performing simulations. It is, therefore, expected that users do not treat the code as black boxes. On the long term, a clearer separation between users and developers should be foreseen. At the moment, this is not opportune yet for the code development. The minimum requirement users expect from the software is that the example macros are running. Meanwhile, these example macros have been updated and are included in the nightly tests which can be monitored on the dashboard. Furthermore, the minimum required should be that tutorials are up to date and the documentation (Wiki) should match the most actual tutorial. Also, this point has been taken up seriously: tutorials are up-to-date and the corresponding documentation is worked upon. For the documentation, the PANDA wiki system is used. A regular update of the documentation is requested including a description of possible projects which can be taken up by (new) developers. Furthermore, the users are strongly encouraged to write themselves notes and documents in the Wiki environment. Write permission can be asked for by contacted me or any other developer. For the code development, a dashboard environment is used. Here, the compilation, builds, and the results of quality assurance macros are reported based on nightly and experimental builds on a large variety of linux and mac distributions. For a large part, this system is made for the developers to test whether there code changes are working on other systems. It was pointed out that also users are invited to look regularly at the dashboard, since this can also help them to identify problems which have been overlooked by the developers. For questions, bugs, and announcements, the forum and ticket system are used. The forum (follow-up of the hypernews with the option to search and include pictures) are monitored daily (sometimes even hourly) by the developers and already extensively exploited by users. The general impression is that the response time in the forum is extremely fast. In some cases, it is not clear whether the person who started the thread is satisfied with the answers. It is therefore requested that everyone who started a thread, also tries to end it by a simple notification in the forum. Also, it is recommended that in the case the forum thread is not answered, the initiator repeats his or her message or question. Everyone is encouraged to use the forum, even if he or she thinks it might be a very stupid question. Furthermore, also analysis items should be posted on the forum. The more recent "ticket system" should be used to post bugs. Everyone (developers and users) can make use of this system. Those who have an svn account can login using the corresponding login/passwd. It is, however, possible to use the ticket system as an anonymous user. The bi-weekly pandaroot EVO meetings are open for everyone! It will be foreseen in the future to allocate the first (half) hour to questions related to users. This will be announced via the forum (meetings). The users are asked to post their questions prior to the meeting to the forum, such that the answer can be prepared accordingly. A discussion came about the release management of pandaroot. At present, pandaroot has a few branches: 1) development branch; 2) trunk (should compile and build); 3) stable (should compile, build, and run quality assurance macros); 4) production releases. At present, most of the participants use the trunk for their studies, since the developments and changes are fast. A request for more regular production releases was made with strict and fixed deadlines. It was also realised that this might be too ambiguous. An attempt will be made to follow-up this request, though. During the future collaboration meetings, dedicated sessions will be organised to discuss physics analysis results.