Attendance:

  Kent Blackburn,  Peter Shawhan,  Masha Barnes,  Greg Mendell,  Warren
  Anderson, Joe Romano,  John Whelan,  Siong Heng,  Patrick Brady,  Alan
  Wiseman, Jolien Creighton,  Duncan Brown .........


Redefining the working group:
  
  Patrick raised the issue of expanding the notion of the MPI group to have
  slightly broader context,  i.e. to become a users group.   General agreement
  with the idea.   Kent wanted to insure the integrety of the working group
  format with Patrick as coordinator.   Nevertheless,  all communication will 
  start to go through the LAL web pages and lal-discuss.    


Signal handling in LAM and LALWrapper:

  Kent is worried about the conflict between signal handler in LAM and that in
  LALWrapper.   Question:  What does LAM signal handler do when a signal is
  raised and what would LAM do if there is a a signal?   If something untoward
  happens in LAL analysis code,  signal handler will catch it.


Metadata in Frame format output from LDAS:

  Greg is outputing frequency series into frame format.  He saw that the
  sample rate is being set in the frame different than what is used in
  frequency series.  What did he mean by the sample rate field.   If LDAS is
  outputting the frame object with frequency series,  etc.   Does not have
  information about data acquisition rate to produce.   Need to know the start
  time.   f0,  df,  number of points ........ 

  The srate entry for the frequency series seems redundant or useless.   We
  should have the wrapperAPI pass the right information.  Set this to -1 in
  the event monitor.   Check with somebody on frame side to if this is a
  sensible thing to do.  Is Benoit still at Caltech?  In the case of
  time-frequency map,  set both to -1:  i.e. fshift and sample rate.  

  Greg would like to have the time duration of a frequency series.   Should we
  raise this with the frame people?