Draft list of Work Packages

This is a draft list of work packages that have been identified during the discussions this week.
Some of the code already exists in people's contrib areas or on their laptops.

1. Make the changes to RawTrackerHit that are described in Summary
2. Deploy the code to create RawTrackerHits from SimTrackerHits. We need two versions of this code.
   • A fast version that is good enough for developing downstream code that must be fast but does not need high fidelity.
   • A high fidelity version that is allowed to be slow.
3. Hit collection management. These are collections that can be added to the event.
   • A structured collection of RawTrackerHits, suitable for finding clusters.
   • A structured container of clusters that link back to RawTrackerHits.
   • A structured container of crosses that link back to clusters.
4. Cluster finders for both strips and pixels.
5. Cross finders.
   • Definitely needed for forward detectors.
   • Should also work in the barrel since we will wish to evaluate stereo in the barrel.
   • Should also clusters to cross at other than right angles.
6. Geometry for forward trackers:
   • For existing devices that are described as the annulus of a disk.
     • Create code to cover the surface with a tiling of pixels or strips. But the representation as the annulus of a disk is not changed.
     • Hans calls this a "readout geometry". Norman calls it virtual segmentation.
   • Implement a more realistic geometry with overlapping planar wafers and associated support structure. Create a scheme to tile each individual wafer with strips or pixels.
7. Cheaters for creating clusters that have unbiased positions and true gaussian resolution, for both strips and pixels. Probably not necessary for crosses since I expect that crosses will be split into clusters when given to the fitters.
8. Port exsiting pattern recognition cheater code to use the new machinery. Where appropriate, extend it to work in the endcap.
9. Port code that feeds hits to fitters to work with the new machinery. Where appropriate, extend it to work in the endcap.
10. Understand what really works and does not work in the world of track fitters and decide which one we will use as the main tool for future development.
    • • Norman's TRF
      • Nick's weight matrix fitter.
      • Caroline's Kalman filter
      • Colorado code ( KFitter +SOD Tracker - not sure what is what here ).
11. Be ready to contribute to the trajectory class ( or an improved track class if that is what is chosen ).