Commit summary from repository starlink
------------------------------------

  (Tim Jenness) smurf: Simplift SMO model logic and use correct quality bits
   100.0% applications/smurf/libsmf/
  
  (Tim Jenness) subpar: Remove debug print that was added in commit 408aa362
   100.0% libraries/pcs/subpar/
  
  (Ed Chapin) smurf: extra status check and fix quality indexing smf_fix_steps
   100.0% applications/smurf/libsmf/
  
  (Ed Chapin) smurf: annul bad status and continues of tswcs construction failed
   100.0% applications/smurf/libsmf/
  
  (Ed Chapin) smurf: smf_find_science returns darks if no science obs available
   100.0% applications/smurf/libsmf/
  
  (Ed Chapin) smurf: clearer error message smf_get_goodrange
   100.0% applications/smurf/libsmf/
  
  (Ed Chapin) smurf: smf_templateFit1 can optionally just fit gain (not offset)
   100.0% applications/smurf/libsmf/
  
  (Tim Jenness) smurf: Fix dim_t vs size_t problem seen in 32-bit builds
   100.0% applications/smurf/libsmf/
 

Commits from repository starlink
-----------------------------

  commit bcb1c52106e51468d46cec0ce957a495acd2b076
  Author: Tim Jenness <[log in to unmask]>
  Date:   Tue Jun 1 16:25:08 2010 -1000
  
      smurf: Simplift SMO model logic and use correct quality bits
  
   applications/smurf/libsmf/smf_calcmodel_smo.c |   86 ++++++++++++++---------
   1 files changed, 53 insertions(+), 33 deletions(-)
  
  commit d6478a7e0ef025bc443337960b34216e59b37ef0
  Author: Tim Jenness <[log in to unmask]>
  Date:   Tue Jun 1 16:13:22 2010 -1000
  
      subpar: Remove debug print that was added in commit 408aa362
  
   libraries/pcs/subpar/sup_crint.f |    1 -
   1 files changed, 0 insertions(+), 1 deletions(-)
  
  commit aa837b5be45a00ec6943892d1120279e589305df
  Author: Ed Chapin <[log in to unmask]>
  Date:   Tue Jun 1 17:25:40 2010 -0700
  
      smurf: extra status check and fix quality indexing smf_fix_steps
      
      The status check is needed in case smf_get_goodrange failed (like
      for a short smfData that is pure padding or apodization).
      The tstride for the quality index was also incorrect, but we didn't
      notice because we normally run this routine on bolo-ordered data.
      For time-ordered data with padding it would almost certainly fail
      (i.e. in an sc2clean call).
  
   applications/smurf/libsmf/smf_fix_steps.c |    6 ++++--
   1 files changed, 4 insertions(+), 2 deletions(-)
  
  commit 210b7e8ad0204b890589c1457fdaca76e3116642
  Author: Ed Chapin <[log in to unmask]>
  Date:   Tue Jun 1 17:23:59 2010 -0700
  
      smurf: annul bad status and continues of tswcs construction failed
      
      This situation may occur in smf_concat_smfGroup if certain metadata
      are broken in some way. Since we usually don't need tswcs for
      anything just annul bad status and continue.
  
   applications/smurf/libsmf/smf_concat_smfGroup.c |   11 +++++++++++
   1 files changed, 11 insertions(+), 0 deletions(-)
  
  commit d669efd23a09ab7ffe3df00c4874d77a66b6f8a9
  Author: Ed Chapin <[log in to unmask]>
  Date:   Tue Jun 1 17:22:47 2010 -0700
  
      smurf: smf_find_science returns darks if no science obs available
      
      This modified behaviour was introduced to make it easy to reduce
      dark data using SMURF tasks as though they were normal science
      data.
  
   applications/smurf/libsmf/smf_find_science.c |   47 +++++++++++++++++++-----
   1 files changed, 37 insertions(+), 10 deletions(-)
  
  commit c13bd38853a34c9d44c66f90008266ddf831389e
  Author: Ed Chapin <[log in to unmask]>
  Date:   Tue Jun 1 17:15:22 2010 -0700
  
      smurf: clearer error message smf_get_goodrange
  
   applications/smurf/libsmf/smf_get_goodrange.c |    2 +-
   1 files changed, 1 insertions(+), 1 deletions(-)
  
  commit 8d0877feeaa7a154fb471a921ff3fad0dcab487d
  Author: Ed Chapin <[log in to unmask]>
  Date:   Tue Jun 1 15:01:19 2010 -0700
  
      smurf: smf_templateFit1 can optionally just fit gain (not offset)
      
      I've added this feature since we potentially have many model
      components with DC level as a free parameter (mean level in
      the map, COM, DKS...). This level is completely unconstrained
      and is also totally degenerate between models. So, you can turn
      off offset fitting in the linear fitter... for now I've just done
      this for DKS but we may want to do this in other places.
  
   applications/smurf/libsmf/smf_calcmodel_com.c   |    2 +-
   applications/smurf/libsmf/smf_clean_dksquid.c   |    6 +-
   applications/smurf/libsmf/smf_flat_fitpoly.c    |   16 +++--
   applications/smurf/libsmf/smf_templateFit1.cgen |   66 ++++++++++++++-------
   4 files changed, 59 insertions(+), 31 deletions(-)
  
  commit 2d5db0c05e4e16c7614546865e4c4044429d252a
  Author: Tim Jenness <[log in to unmask]>
  Date:   Tue Jun 1 09:23:00 2010 -1000
  
      smurf: Fix dim_t vs size_t problem seen in 32-bit builds
      
      Reported by Trystyn Berg.
  
   applications/smurf/libsmf/smf.h.source           |    2 +-
   applications/smurf/libsmf/smf_qualstats_model.c  |    8 +++++---
   applications/smurf/libsmf/smf_qualstats_report.c |    8 ++++----
   3 files changed, 10 insertions(+), 8 deletions(-)