AIDA Developers Workshop, SLAC 2007

Main Outcomes

  • AIDA 4.0 - The next release of AIDA will be 4.0. Since the major version number will be incremented we will not require strict backwards compatibility with the existing 3.3.* releases, but we will try to minimize and document incompatibilities.
  • Geant4 Documentation - We will try to combine the existing Open Scientist documentation with the JAIDA/AIDA documentation for Geant4 into one set of documentation which clearly describes how to set up AIDA for Geant4. We should try to get all the Geant4 README files updated to point to this documentation, instead of containing often outdated information on a single AIDA implementation.
  • Persistency
    • AIDA implementations should support plain XML, zipped ASCII XML, binary XML, zipped binary XML. These files will all be called .aida files. Implementations will be able to read any files. When writing we will need to specify which format to use
      • Currently we specify the file as xyz.aida, storetype=xml, with options compress=true/false
      • We can add storetype=zip,fs (unzipped file structure), option binary=true/false
    • We discussed storing partial AIDA objects, for supporting large clouds and tuples. Significant progress has been made on this issue. AIDA-222@bugs
  • Functions
    • We discussed the meaning of codelet, and the role of the function catalog. We agreed on a common kinds of codelet, we need to nail down the exact syntax.

User Requests

We held an AIDA Users session.

  • Guy gave an impressive demo of Open Scientist :
    • demo of the installation and usage of the osc_batch, osc_vis, osc_g4 kits.
    • demo of the creation of an AIDA program storing an histo at various format.
    • demo of the interoperability with ja3
    • demo of the usage of the interactive onxlab and opaw program to view histos.
    • demo of how to use the osc_g4 kit to do some visualization of a G4 application.

Two user requests were discussed

  • Easier access to tuple columns. This may just be an issue of updating the AIDA examples to show how to use the new Tuple Columns introduced in AIDA 3.3.
  • Ability to produce random numbers from an IFunction representing a PDF. This could probably be provided as a utility on top of the existing AIDA interfaces.
  • No labels