==================================================
Minutes of the Tracking Meeting, November 15, 2007
==================================================
taken by U. Husemann


Aart Heijboer: SVX Alignment
============================

- Compared residuals between Gen6 and Gen7 tracking on approximately the
  same data (jet samples with lots of tracks)

- r-phi alignment (fixing SVX layer 5 and SVX beam line):
  o average residuals identical

- r-phi alignment (using only COT tracks):
  o 25 um offset for sinking detector clearly visible
  o patterns similar, but shifts up to 10 um (not exactly same data?)

- z alignment (L1 and L5 fixed)
  o Fairly large by very stable residuals
  o Future: better z alignment may be required for forward b-tagging
    (would require major changes to alignment)

- Conclusion: no need for new SVX alignment for Gen7
- First look at post-shutdown data expected next week
- Working on a code clean-up before handing over code (and responsibilities) 
  to Andreas Schmidt (Karlsruhe)


Antonio Boveia: Backward tracking
=================================

- Improving mass resolution:
  o By attaching more COT hits: 
    only 50% of hits are correctly attached (if all COT hits are 
    dropped and re-assigned), under investigation
  o By seeding from larger radius (layer 4 instead of layer 2)
  o Result: similar resolution as IO tracks

- Adding stereo segments:
  o Old: attach stereo segments to axial segment before passing on to silicon
  o New: move stereo hit assigment after silicon tracking: more efficient

- Optimization:
  o Many small tweaks to remove unnecessary operations in Backward tracking
    (with careful validation after each step)
  o Ran profiler on ProductionExe: many places identified where optimization 
    is possible, but nothing touched outside Backward tracking

- Identified bugs in Silicon part of OI code, fixed in private code
  o z hit assignment
  o alternate phi hit


Discussion on next Pre-Release
==============================

- Status:
  o Matt's changes to FWD tracking are ready to go
  o A couple of days needed to address remaining Backward tracking issues
  o Ming would like to incorporate a muon update by Pasha Murat 
- Plan: try a pre-release by late Monday next week