Attendance:
        
  Peter Shawhan,  Linqin Wen, Albert Lazzarini, Kent Blackburn, Masha Barnes,
  John Whelan, Joe Romano, Patrick Brady, Jolien Creighton, Alan Wiseman,
  Rauha Rahkola,  Greg Mendell,  Ed Maros.....


Action items:  
  1) Patrick will try to do this ...... try the inspiral MDC tests again.
  Get back to Phillip on this.  Duncan thinks the bug might be in simulation
  part.   

  2) Jolien,  Patrick,  Peter will get the calibration into LAL, produce and
  document a piece of code to generate the response function in ILWD and
  format suitable for ingestion into the database.

  3) Patrick will e-mail developers to find out what they use,  if anything.  


Status of LDAS:  
  
  code freeze this friday.  Testing next week.  Release on Dec. 14
  target date.  Will not be able to do disk cache API.  May or may not be able
  to do decimation in frameAPI.  Optimistic to get it in.  Kent will e-mail
  lal-discuss.  Things that used to work breaking.  Greg asked about the slice
  action for the frequency series.  Stochastic group is reasonably happy with
  progress;  LDAS end they are testing proc frames.  


LAL and LALWrapper releases:  
  
  target dates still stand.  Code freeze starts on Dec 9;
  release on Dec 18.   Mapping between units and channels.  There is a file at
  the sites which has that,  but nobody knows where it is.   Rolf or Alex -->
  Peter can do that.  It will only be things in the frames.  Human readable
  only;  no consistent.  John will check if it has the info;  deal again
  later.


Calibration records:  
  
  Patrick asked about the channel called AS_Q in frames,  and if that is in
  fact AS_Q?  Peter explained that the calibration is based on what is
  recorded in frames, so it should not matter.    One arm at a time?   Not a
  problem because absolute calibration of each arm is known and DARM can be
  reconstructed.   It is possible that both arms differ by as much as 20% in
  calibration;  does this need to be accounted for in the search codes.
  Unclear about the time-delays.

  Albert pointed out that this should be handled with the DetChar group.
  Patrick will touch base with Keith and propose that the calibration team be
  augmented by one member from each upper limit group to understand
  the calibration and fully document it.  
  
  LDAS can get the calibration frequency series either from a file or
  database.    Kent has used this in a shared object.  It works.  Peter will
  point me to file with the response;  will take care of getting into the
  database.  

  Shared objects use different things.  Stochastic uses transfer function,
  others use response.  Need to decide on one or the other.  Datacond cannot
  invert a frequency series at this time. 
  
  
Shared objects status:
  
  Stochastic:  code in LALWrapper and LAL will be ready to give the integrand
  of optimal statistic.  To go CC stat and estimates of upper limits.  They
  still need some post-processing stuff to be put in place.   Modulo issues of
  units and format of response.   

  Stochastic-bar:  ran a lot of tests.  Six out of eight tests worked. Not
  sure of status.  No developers on line.   

  Known Pulsar:  things are ok.  A little concerned about the calibration.
  Just resample AS_Q.   Will think about the calibration.  Not ready to
  detect pulsars,  but for first step of producing short Fourier transforms.   
  When LDAS can read sft's back into code then he can run whole thing in LDAS.    

  FCT:  at this moment,  she is adding a new subroutine to make it more
  efficient.   So that they search only physically meaningful regime.  FCT
  hierarchical search code.  Still need to move two subroutines from flat to
  hierarchical;  this week.  


Other issues:
  
  Job submission during the engineering run:  People should think ahead to job
  submission.   Automated or semi-automated.  There is an example which was
  worked out by Peter and Julien.  Peter will circulate the final script that
  was used.