E895 Weekly Meeting (held at DAVIS)
				June 3, 1997


	Attended by Daniel Cebra (DC), Jim Draper (JD), Nathan Stone (NS)
		Juan Romero (JR), Jerry Chance (JC), Mark Gilkes (MG)
		Jessica Kintner (JK), William Caskey (WC), 
		Paul Brady(FPB), Mike Heffner (MH)

DC	I have received some faxes from D. Best concerning the vertex
	fitting as approached from ALEPH and H0.  The high energy
	mathematics say that least squares fitting goes as n^2 or n^3 but 
	I only have n.
	Dieters Proposal No vertex refitting in Pass1
	- No refitting using original hits
	- No refitting V0's
	Dieter argues that with the Helix par. you can still refit using
	mathematical algorithms.

JD	Is it better to have a common vertex?

DC	The problem is compute time dominated in HE physics with low mult
	but high numbers of events.

NS	I would like to have the hits table written out if possible.

DC	Different collaborations do keep the hits for refitting NA49 for 
	one.

JD	Question of Storage?

DC	Question of Cost, they are a bigger collaboration and can afford
	the increased size.

NS	Lets measure it without the fat in the trks and hits table and get
	and find out how much space we are talking about.

DC	We ran two separate versions in the EOS dst production.

NS	We should avoid this at all cost.

JD	What about the magnetic field, does the const approximation 
	effect the vertexing.

NS	Only Dieter and Gulshan looked into this and then only for 
	momentum resolution.

DC	The mag field is not used, the helices are wound back to the
	the vertex and treated as though they are vectors.

JD	Is it really faster to do vectors?  Instead of treating the trks
	as helices.

DC	The mathematics are soluble for the vector approach and we are
	talking about accuracy of 10's of microns 

JD	Can you get the wrong answer?

DC 	Perhaps with bad events.

SP	Geant study on it?

MG	Fast and slow simulators should be ready and are running on the
	VMS.

NS	Is the code in CVS?

MG	I don't know.

NS	I have been corresponding with Chris P. at SUNYSB about the VMS. 
	Files are in the CVS util  but there are some problems.

JK	Working getting the TAS3 code on my machine at St. Mary's.

NS	Working on checking CVS -vs- PDSF.

WC	Hitfinder_2d working fine in mode 0 and 6 on HP's at Davis.

Everyone	We should use bench mark file at /home/rncus2/d2/data
	on sseos to make reports on code.

MG	Looking at watch feature on gdb to watch memory locations to solve
	some of the memory problems.

NS	Problem with BADPADS and memory overwriting. Need the watch
	feature to solve this.

JR	Music update, we are coding up the MUSIC code to run in tas3.

JC	Calibration procedure established for music but not done yet.
	Nothing done on TPC MUSIC matching either.

DC	This is probably not needed anyway.  The gain on the TPC so high
	large fragments are off the TPC scale.

___________________________________________________________________________

Minuteman: Bill Caskey