Attendance:

  Julien Sylvestre,  Kent Blackburn,  Masha Barnes,  Peter Shawhan,  Mary Lei,
  Isaac Salzsman,  Patrick Brady,  Duncan Brown,  Jolien Creighton, Greg
  Mendell,  ........ 

Action items:

  1. Ed Maros will put up and document the latest patches for LAM.


Status of LDAS release:  

  They have some problems reading 16 second frames.  They tested with the
  latest version of LAM and it seems to stabilize a number of issues in LDAS.
  There is a new patch file for this.  See action items.
  
  Current release of LDAS allows querying database to construct quality
  channel.  Can now get spectrum out of the database too.  It is tested and
  works.  The n-tuple format has been tested at a simple level.   They dont
  have a sound check to do,  so they need to have detailed checks made.  


Getting response function to DSO's:

  There is some confusion.   Need new functionality to get the response
  function out of the database.  Interpolation is on the list to be
  implemented.  The dbspectrum may provide a mechanism to get the response
  out;  there is a problem that this requires exactly the right "start time".
  Is it something that the job control library may be able to handle?  A
  possible long/short term solution is to put the response functions in files
  which are located in a given location like HPSS and/or a web site.  Action
  is needed to ask about best way to do this.


Job pacing and control:

  Peter has released a new ligotools package.  It  contains a library with the
  basic functionality.   More capabilities such as sending stuff from local
  system could be added,  but try it out first.  For datapipeline jobs:  some
  or all products are handled asynchronously by LDAS --> the output will show
  up at some point.  This causes some problems for automated or human
  processing.  There is no way around this at this time without LDAS.  Kent
  has wanted this changed for a long time:  why can it not be done?   Ought to
  be possible to make a change to do this.  This is necessary for a pacing
  system that is reliable.  Right now it just waits 10 seconds and then tells
  you what's there.  

  Mainly is designed to monitor the live segments ........ issues a
  getmetadata at certain interval (2 minutes) looks at segment data retrieved.


Pulsar MDC and CVS archives:
 
  Greg does not think it is essential that the CVS archives be closed.   AEI
  group has not checked their latest code in;  this is causing problems with
  the timing so he is just planning to test building DFT's