Heavy Flavor Averaging Group Phone Meeting

Wed, 10 September 2003, 7:00 am PDT

Minutes:

• Jim Alexander, Marina Artuso, Hal Evans, Andreas Hoecker, David Kirkby, Urs Langenegger, Vera Luth, Olivier Schneider, Marjorie Shapiro, Jim Smith.

• Jim Smith reported on difficulties of preparing rare averages for summer 2003: large number of quantities to average, labor intensive, quality control problematic. Problems agravated by experiments missing deadlines.
• Proposed HFAG policy for conference updates: "Experiments should provide all results to be included in average at least one week before start of conference. This is a firm deadline."
• Yoshi and David will work to persuade collaboration managements of the importance and benefits of respecting this policy.
  
• HFAG members should help ensure that when their collaboration will miss a deadline, HFAG is at least aware that a result is in the pipeline.
• It is crucial that we respect the confidentially of the collaborations in providing advance information about their conference results.

• We will prepare a single document that covers results and methodology for all subgroups on a twice-yearly basis.
• The document should be available about 2 weeks after the relevant conference, e.g., so it can be referenced in proceedings contributions.
  
• The first such document will be prepared for the PDG 2004 update. [ ed: To be useful, this would need to be finalized early enough to be referenced in the RPP. Since the results will be duplicated in the RPP, this would primarily be a reference for methodology. ]
• David and Yoshi will coordinate writing assignments.
  

• See notes prepared by DK
• Input files, programs, and scripts used to generate each average should be archived in CVS
• CVS repository already exists in HFAG AFS space. Only COMBOS archived there now.
• DK will provide guidelines and define top-level strucutre of the CVS repository

• See notes prepared by DK
• At a minimum, new algorithms are needed in COMBOS to allow some recent results to be included in the oscillation averages.
• A complete redesign of COMBOS + BFRACTIONS + ... to use object-oriented methodology, etc, is well motivated but would be a significant software engineering project. We could not identify the necessary manpower at the meeting.
• DK will draft a set of physics requirements and uses cases for our averaging software as the next logical step in this discussion.

• Andreas Hoecker described difficulties of quality control in rush before a conference.
• We will include an explicit QC phase in our next schedule. Checks will be performed internally and completed before the conference (rather than waiting for the writeup).