Core Minutes 7/6/2004  

Present:    Joanne Bogart, David Chamont, Jim Chiang, Richard Dubois,  Riccardo Giannitrapani, Berrie Giebels, Navid Golpayegani, Traudl Hansl-Kozanecka, Heather Kelly, Michael Kuss, Julie McEnery, Pat Nolan, James Peachey, Tom Stephens, Tracy Usher

• ROOT file headers: (David) The project is well documented on the wiki page and the links it contains. The code is not yet in CVS, but could (and probably should) be committed.  Suggestions and comments are solicited.

• MRvcmt: (Riccardo) Latest additions include an option (off by default) for CVS parsing: within the package browser frame, files known to CVS are distinguished form those that aren't.

• FRED: (Riccardo) Ability to find a particular run/evnet from FRED is largely working; he just needs to handle exceptional situations, such as when the requested event doesn't actually exist, gracefully.

• System tests: (Julie) System test results for v4r3p1 are available. One significant difference between these results and those for the reference, v4r2, is in some of the TKRTRKENERGY plots.

Flight integration news for the week:

• LDF System/Unit test: (Heather) The CAL part is done. Still some work remaining to deal with TKR readout order.

• Calibration news: (Joanne) Support for ROOT calibration files is in progress: no ETA yet. Awaiting a hardy user for rdbGUI.

• Tracker alignment: (Tracy) Some time ago Hiro wrote a stand-alone program to do inter-tower alignment. This involves finding six parameters for a given tower, three position offsets and three (small) rotation angles. The current standalone approach operates in two steps. The first stage skims off events of interest (i.e. tracks passing through two towers). The second stage does the actual alignment: assuming one tower (the reference tower) to be perfectly aligned, alignment parameters for the second one (the alignment tower) are optimized using a MINUIT minimization algorithm. This algorithm "wiggles" the alignment tower around and looks at the effect this has on residuals formed from hits in the alignment tower to those predicted from tracks projected from the (stationary) reference tower. Hiro developed this program over the winter but, due to pressing I&T responsibilities, has recently handed it over to the TkrRecon group for continued development.

  The program works but, as a stand-alone, it has no access to standard tools and algorithms of tracker reconstruction, so code is duplicated; with no access to the xml geometry description, geometry is hard-coded.  Tracy and Leon have done some refactoring to make it possible to incorporate the project into the Gaudi framework.  Once this is complete, there are a couple other alternatives they'd like to try: one an analytic approach described in a CHEP03 paper, and another proposed by Leon which looks at the intersection of the two pieces of a track fit separately in each tower.  Exactly how they proceed depends on just when the system is needed by I&T; something usable exists already.(Richard) About 5 weeks from now.

• Pipeline: (Richard for Dan; Navid) Pipeline was to be delivered today or tomorrow to I&T and may in fact be ready today. Dan has finished the stored procedures.  The job job scheduler is/will be a Perl script which invokes them. (Navid) The web interface is ready [after Navid generously gave up his weekend to make it so].

And finally..

• Open JIRA issues: (Richard) Here is a convenient way to view them.  Of particular note is SYSTESTS-7, the discovery of increased memory usage in System Tests between v4r1 and v4r2. Julie suspects TkrDigi.
• Announcements: (Richard)
  • Our new tech writer has arrived. Heather will be at SLAC Thursday/Friday this week; should be a good opportunity for a documentation get-together.
  • GLAST budget (generally; not us) is still creeping upwards. This will be investigated next week.
  • Congratulation, Johann and Mirina! Johann will be back at work Monday.

previous

minutes index

next

J. Bogart Last Modified: 04-Aug-2004 15:40:48 -0700