[Frost] Revised: Eloss error

[Michael Dugger]

My previous email should have read:

-----------------------------------
Hi,

Brian V. found a small error in the geometry contained in the targets.F
file of Eloss. This has been fixed and I have asked Brian M. to check the
new version into CVS and rootbeer.

The error:
For events that come from the Carbon or CH2 targets, there is an
approximation that was not valid. This only affected a small number of
events and only for the Carbon and CH2 targets that had angles large
enough that the path did not go through the front or back of the target
but, instead, went through the Carbon or CH2 cylinder wall.

Because the thickness of the Carbon and CH2 targets are less than the
vertex resolution, events that are assumed to come from these targets are
treated as a special case, where the z-vertex is set to originate at
the z-center of these targets. The pathlength was approximated as
0.5*targetThickness/z-direction-cosine. For large enough angles, this 
simply does not hold and a more correct expression has replaced the 
previous approximation.

I am thankful that Brian V. found the error, and I apologize for
any inconvenience this may cause members of the FROST group.

Sincerely,
Michael