[d2n-analysis-talk] Added Diana's Farm-replay files to CVS and d2n.jlab.org

[Brad Sawatzky]

Diana's farm-replay files have been added to CVS and to
/usr/local/d2n_analysis/d2n/replay/ on d2n.jlab.org.  The change notes
are at the bottom of this email.

I also took the liberty of updating the BB farm replay script to use
Diana's trimmed odef file from Dec 3.  These are the changed/added
files:
  replay_farm_BB.C
  replay_farm_BB.odef


In other news:

As Diana and Matt had worked out, the compile-from-scratch problems
people have been having are due to changes uploaded to the central Hall
A analyzer CVS trunk.  All the stuff in the d2n repositories appears to
be current.

The proper fix here is probably to (create and) use an analyzer branch
for our work.  I'll set it up and send out instructions.

The quick fix is to grab the analyzer tarball:
  /usr/local/d2n_analysis/analyzer-d2n.tar.bz2
and use it rather than pulling the analyzer source from CVS.

I'm going to poke at the 64-bit build issues now.

-- Brad

----- Forwarded message from Diana Parno <dparno at cmu.edu> -----

From: Diana Parno <dparno at cmu.edu>
Subject: Farm-replay files for the CVS repository
To: Brad Sawatzky <brads at jlab.org>

Hi Brad,

I've attached a few files for the d2n_analysis/d2n/replay directory of
the CVS repository.

replay_farm_L.C, replay_farm_BB.C, and FarmReplayCore.C are new files
and are slight revisions of the normal replay macros for farm running.
Unlike our ordinary scripts (replay_det_L.C, replay_det_BB.C, and
ReplayCore.C), they are designed to run from a single data file at a
time, rather than analyzing all extant data files for a run. Our farm
replay system is currently designed on the one-job-per-data-file model
(rather than one-job-per-run), so we end up needing slightly different
macros.

def.h and rootlogon.C are modifications of files that already exist in
the replay. def.h now #includes FarmReplayCore.C and rootlogon.C loads
the two replay_farm macros into the analyzer environment when it
starts. They both maintain the functionality of their originals --
that is, you shouldn't have to modify them in order to run the
original (non-farm) replay macros.

I'd like to get these added to CVS so that Dave and Matt can
(eventually and if desired) set up their own farm replay systems more
easily, and so that I won't have to spend a lot of effort on my farm
replay directory every time I'd like to incorporate some updated code
in CVS. I think these changes are of fairly general interest.

Have you had a chance to update the CVS repository for the new
analyzer files and momentum code? Or to look into running the analyzer
on a 64-bit system?

Thanks,
Diana

-- 
Brad Sawatzky, PhD <brads at jlab.org>  -<>-  Jefferson Lab / Hall C / C111
Ph: 757-269-5947  -<>-  Fax: 757-269-5235  -<>- Pager: brads-page at jlab.org
The most exciting phrase to hear in science, the one that heralds new
  discoveries, is not "Eureka!" but "That's funny..."   -- Isaac Asimov