E895 Meeting Minutes                 23 September, 1997

Attending:  Gulshan Rai, Nathan Stone, Dieter Best, Heng Liu,
            Sergei Panitkin, Bill Caskey, Mike Heffner, Daniel Cebra,
            Jessica Kintner

            phone in: Roy Lacey

------------------------------------------------------------------------

News & Status Report:

GR  -   I have appointed a 4 member group (NS,DB,ML,CP) to finalize
        our pass1 table structure.
    -   We have an account on mass storage which can be used by the
        group, but I only want to see our official pass1 output there.
    -   On the distortion correction, Jim Thomas agrees with what I have.
        and he found that reconstructed momentum changes about 10% with
        distortion correction on.
    -   Is MUSIC code fairly stable.

NS  -   We need input from Jerry.
    -   sys directory with xdf can be compiled on HP. The trick is to
        add -extend flag to makedef files.
    -   Propose we compare the output from the summary table on the
        benchmark files on various platforms.

SP  -   12 pentium boxes have arrived. By the end of the day we can
        set them up in the PDSF room.

HL  -   pass1 jobs with 2d+jtf are running on pdsf for 6 GeV.

DB  -   compared 1d with 2d for 250 events at 2 GeV,
                        1d+btf          2d+btf
        timing          30 sec/event    1st pass 30 sec/event,
                                        2nd pass 10 times slower
        performance  1) hit-removal-efficiency is about the same
                     2) hits/track is about the same
                     3) track multiplicity is larger for 2d+btf
                     4) log-log plot of dEdx about the same, 2d may be
                        slight better.

    -   showed the study with/without distortion correction with 1d+btf
        for 6 GeV data.
        hit-removal-efficiency increases slightly with distortion on, number
        of tracks increases and hits/track drops with distortion ( are we
        finding more shorter tracks ?), momentum distribution doesn't
        change much, vertex resolution along z improves with distortion.

GR  -   SUNY should also run BTF to get v0 and do the comparison.

DC  -   showed the flow analysis for EOS 1.15 GeV data and E895 2,4 GeV
        data, using the fourier method.



                                Prepared by Heng Liu