Hi All,
At yesterday's meeting, and at a previous one, I brought up a point
about backwards compatibility of the tracking code that we plan to
develop. I was confused by some of the responses but I think I now
understand my confusion.
Is the following a correct summary of the main issues?
1) There is a large body of simulated events.
a) The PFA folks intend to use these events for their development
work. They neither need nor want to generate new events.
b) Even for "tracking only signals", these events can be used for
estimating backgrounds for many of the channels that we
have not yet looked at.
For these jobs the cylindrical approximation to the vertex and
tracking detectors is sufficient.
2) There is a separate goal: improve the simulation and reconstruction
code for the vertex and tracking detectors so that they are made of
planar sensors. This step is needed to understand questions that
can only be answered once realistic pattern recognition codes exist.
My comments were made within the context of 2: for purposes of doing 2,
there is no reason for the new reconstruction code to work on the
previously generated events.
This leads to two questions:
A) Are there improvements to the tracking code that are required in
order to get the best benefit from job 1?
B) Can we design job 2 in such a way that we get the
improvements required for 1 at minimal additional cost?
Is this a correct summary of the situation?
Rob
|