sPHENIX discussion of software

[pinkenburg <pinkenburg AT bnl.gov>]

Hi folks,

in preparation for tomorrow I think the following is the approach which 
preserves a maximum of our physics:

1) run the full calorimeter reconstruction for all events (1.4s/evt)

2) run every third event through the pass1 tracking (10s/evt). This 
should be sufficient to extract the distortion correction, using fewer 
events for this needs further evaluation.

3) run a trigger using the calorimeter output where the output from the 
pass1 tracking can also be used and basically get a list of events we 
want to run the full reconstruction for.

4) run pass1 and pass2 for the selected events. We probably can take a 
shortcut for tracking selected events since we have the distortion 
correction in hand for this event (and if it is an event which ran 
through pass1 we only have to run the fast pass2 on those)

Running 1) and 2) needs about 1/3rd of our requested resources which is 
what we have guaranteed (on the floor and on order). What is appealing 
here is that we will get the distortion correction for all events and 
given the fixed interval we can preserve the output so we don't throw 
this investment away and have to rerun it in the offline pass.

As for the MB events it depends how the pass2 fits into the budget, in 
principle the above would preserve 1/3rd of our MB statistics plus 
whatever our triggers pick out (as long as they have a halfway decent 
rejection).

Chris

--
*************************************************************

Christopher H. Pinkenburg       ;    pinkenburg AT bnl.gov
                                ;    http://www.phenix.bnl.gov/~pinkenbu

Brookhaven National Laboratory  ;    phone: (631) 344-5692
Physics Department Bldg 510 C   ;    fax:   (631) 344-3253
Upton, NY 11973-5000

*************************************************************