Fred Jones wish job list.

• STT Web-Tools: work with John Allison and Steve O'Neale on modifying the STT scripts so that the test-jobs will send logs and status information to the Tinderbox server at TRIUMF. The idea would be to get some useful web-based test monitoring going as soon as possible with a minimal intervention into the existing system testing framework.
• Discuss some practical problems and impediments to successful use of Geant4. Especially, the setting of environment variables and consequent compile-time options -- these are critical settings where a small error or forgetfulness can cause a lot of grief. It would be nice to have a small specialized "environment set-up" tool that is packaged directly with the Geant4 release. I have some ideas of how to make a script and/or GUI to do this, but need to consult with someone expert in Unix and NT processes and shells. Another need is for an "inquiry" method in the G4 kernel which would allow the user to find out as much as possible about how the libraries and executable were built, i.e. a sort of self-certifying function.
• Discuss and learn about "simulation/analysis frameworks", including many of the topics raised by Guy Barrand. This now seems to now be a big issue, with various end-users wanting to use different frameworks. Question: is Geant4 a toolkit, a framework, or both? Although conceived as a toolkit, it already seems to have some of the characteristics of a framework: input shells, macro language, event handling, compiled-in graphics, and maybe soon an analysis category. As a side topic, I have a particular interest in frameworks built with high-level scripting languages rather than C++ or Java. These "loosely coupled" frameworks generally require less programming effort and can be easily modified and extended by the user without rebuilding libraries and executables. If there are others interested in this approach, I would like to exchange ideas.