Release Note for SAS 16.0
		     ===========================

		   Carlos Gabriel & Eduardo Ojero
	                 XMM-Newton SOC




* Contents

  . SAS releases covered by this note
  . URLs
  . Summary of important changes
  . New and withdrawn packages
  . Detailed list of changes between SAS 15.0 and SAS 16.0

* SAS releases covered by this note


The version of the SAS installed can be displayed with the command:

			      sasversion

This version (16.0.0): xmmsas_20170112_1337

Last public version  (15.0.0): xmmsas_20160112_1832


* URLs

SAS pages at the XMM-Newton SOC: http://www.cosmos.esa.int/web/xmm-newton/sas-news
   http://www.cosmos.esa.int/web/xmm-newton/sas

XMM-Newton calibration information:
   http://www.cosmos.esa.int/web/xmm-newton/calibration


*Summary of important changes



The main improvements of this release are:


- SAS 16 has been compiled with the GNU C/C++ compiler collection GCC 6.2.0, including gfortran 6.2,
used for compilation of the SAS fortran code instead of the so far used commercial NAG Fortran compiler.
This change implied as well adopting the newest C++ and Fortran coding standards.


- The new task evqpb creates for any given EPIC Full Frame science exposure an event list containing the associated
quiescent particle background (QPB). The task extracts it from the Filter Wheel Closed data (FWC).
This instrumental background can be used for correcting spectra or images.


- The new metatask edetect_stack performs standardized EPIC source detection on overlapping fields of
different observations (also Mosaic Mode ones). It is still in experimental stage, but it is expected to be the basis for a future catalogue of
stacked observations with XMM-Newton.


- The new task implotregions produces EPIC thumbnail images (PNG, GIF) showing selected source
and background extraction regions.


- The task epxrlcorr, performing corrections related to EPIC X-ray loading has been upgraded,
optionally providing
diagnostic output images. They can be used as visual tools to decide whether an observation
is affected by X-ray loading.


- The task ebkgreg, introduced in SAS 15, determining the most suitable position of a circle region
to get the background estimation of any
source in an EPIC image, has been upgraded to work also with MOS data.


- The metatask eslewchain, in charge of Slew Data reduction, includes now source detection.


- The task rgsrmfgen, calculating the RGS response matrices, offers now a new (optional)
RGS effective area correction, aiming to bring a quality improvement in the cross calibration between RGS1
and RGS2, first and second order.


- The new task rgsimageforbadpixfind is ancillary to the implementation of an alternative
for RGS bad column finding, using the MOS task embadpixfind. This method is recommended only in case the user suspects that
the profile of a bright emission line is distorted due to the false detection of a hot column.


- The esas collection of tasks for extended source analysis, is now fully coded in F90
(no F77 code included anymore). This
is an important step for making it fully compliant with the rest of the SAS S/W. expected by the next SAS release.




This version is released in two 64-bit binary versions for Linux (RHEL 6.8 and Ubuntu 16) and two for MacOs (Yosemite and El Capitan).
The versions released have been tested to run correctly on several other platforms (we provide some libraries where
necessary together with the distribution).

 As repeateadly announced, no 32-bit binary versions are released anymore.


As always, many other less important improvements and small bug fixes have taken place. 90 tasks have been
upgraded all in all, with a total number of 229 sub-version changes. A detailed list follows further below.



* New and withdrawn packages

NEW:
Package edetect_stack: version 0.1.1 [NEW]


Package evqpb: version 0.4.7 [NEW]


Package implotregions: version 0.4 [NEW]


Package rgsimageforbadpixfind: version 0.1.1 [NEW]





WITHDRAWN:
none

DELETED:
none



* Detailed list of changes between SAS 15.0.1 and SAS 16.0.0

A detailed list of all upgraded tasks wrt SAS 15.0.1 follows, the corresponding
ChangeLogs are further below:

ChangeLog

Version 1.93.1 - 2016-12-01 (RDS)
------------
	+ (src/BadPixelCorrector.cc,      - Ensure that the OFFSETn.ds
	       BadPixPSFCorrector.cc,       temporary files are deleted
	       DetmapFilter.cc)             at the end of the run
	                                        (SPR-7412)

Version 1.93 - 2016-04-20 (RDS)
------------
	+ (src/ConcreteArfDataServer.cc)  - Support boxes and RAW-Y range
	   src/RegionFunctions.cc)	    when extracting FAST mode data
	   					(SPR-7342)

Version 1.92 - 2015-12-04 (RDS)
------------
	+ (src/BadPixelCorrector.cc)      - Uses the NEXT_TO_OFFSET_COLUMN
					    rather than CLOSE_TO_BAD_COLUMN
					    to interpret pixels lying next to
					    offset columns.  (SPR-7330)

ChangeLog

Version 2.30 - 2016-09-05 (RDS)
----------------------
	src/asmooth_mod.f90	- changed to dfftw from sfftw in convolveByFft
				  to resolve problems with GCC 6

Version 2.29 - 2009-05-01 (MAS/LUX)
----------------------
- The new DEPENDency on fitsutils is added (1.0) and that on ssclib is upgraded to 4.8, to follow the change of the place of FITS-related libraries.
	

ChangeLog

Version 4.37 - 2016-09-06 (AI)
---------------------------------
	- (src/attcalc_mod.f90) Initialization to null of all pointers that are passed
	to C++ (Dope vectors). If we do not initizialize the pointers,
	we get seg fault at run-time with gcc-6.2.	

Version 4.36 - 2016-07-20 (AI)
---------------------------------
	+ (src/attcalc_mod.f90) compatibility changes with gfortran applied.
	The Bmatrix is passed to C++ with a wrong array descriptor elements (dtype and triplet).
	Only if we create the local variable matrix, we can pass the right values to C++. Not clear
	why this is happening, we have to keep an eye on it.
	The OAL_toEulerAngles subrotutine has to be called with the matrix transpose.
	Otherwise, the C++ function handle the matrix elements in the wrong order. We
	have a similar effect in  rgslib(euler321_mod.f90) module. This problem also appears in
	eexpmap package (the code is duplicated...)
	+ (src/Makefile) gfrotran preprocessor macros added.
	
Version 4.35 - 2016-07-15 (AI)
---------------------------------
	+ (src/attcalc_mod.f90) gfortran compatibilty changes.
	
Version 4.34 - 2015-08-25 (RDS)
---------------------------------
 + (src/attcalc_mod.f90): Changed the use of the psi euler angle such that
 			  it subtracts rather than adds. This is only valid if
			  the BORESIGHT CCF contains PSI with a changed sign
			  (SCR-7265)

ChangeLog

Version 1.3.3 - 2016-12-01 (AI)
------------
	+ (src/attfilter.cc) isnan extern macro removed.
	
Version 1.3.2 - 2010-04-01 (AI)
-------------
	+ (src/attfilter.cc) The data filter expression in "cut" method was created using
	the "toString" to convert start and stop time to strings variables. This "toString"
	method truncate the variable so we are losing time due to this conversion. We have
	changed "toString" method by double+stringstream definition.

ChangeLog

Version 1.40 - 2016-09-06 (AI)
--------------------------------
- (src/badpixfind_mod.f90) Initialization to null of all pointers that are passed
  to C++ (Dope vectors). If we do not initizialize the pointers,
  we get seg fault at run-time with gcc-6.2.	

Version 1.39 - 2014-04-08 (EC)
--------------------------------
- moved test/*.NEW to testprods/badpixfind_data
- removed unused binary data from test/

ChangeLog

Version 3.239.2 - 2016-11-03 (AI)
---------------
	+ (src/EffectiveAreaDataServer.h/cc) New AREACORR methods added.

Version 3.239.1 - 2016-10-18 (RDS)
---------------
	+ (cxxtest/fwc.cc)	   	   - Changed FWC test to EMOS1

Version 3.239 - 2016-10-06 (RDS)
---------------
	+ (src/FWCDataServer.cc/h)	   - Added a new atom to return
					     filter wheel closed data
					     (SCR-7361)

Version 3.238 - 2016-09-02 (RDS&AI)
---------------
	+ (f90/CalF90cxxGlue.cc) check of dope vector added to several methods
	+ (src/CalDataVector.h,CcdQuantumDataServer.cc,Redistribution.cc)
	gcc-6.2 compatibility changes
	+ (f90test) Initialization to null of all pointers that are passed
	to C++ (Dope vectors). If we do not initizialize the pointers,
	we get seg fault at run-time.	
	
Version 3.237 - 2016-06-30 (AI)
---------------
	+ (f90/CalF90cxxGlue.cc) gfortran compatibility
	chaged for Dope (array descriptor) structures.
	+ (f90test/caltest_epn.f90 _emos.f90 _rgs.f90) gfortran stdout write
	compatibility changes.
	
Version 3.236.3 - 2016-02-01 (RDS)
---------------
	+ (src/GeometryDataServer.cc)     - Remove debug causing problem in
					    psfgen test (SPR-7341)

ChangeLog

Version 1.112.2 - 2016-10-06 (RDS)
---------------
	+ src/Ccf.h: ADded new CCF 'FWC'

Version 1.112.1 -2016-09-09 (EO)
---------------
	+ src/PhysicalConstants.h: GNUC >= 6 since gcc 5.3.1 doesn ot use C++
	  2011 by default.

Version 1.112 - 2016-09-05 (RDS)
---------------
	+ src/PhysicalConstants.h: GCC 6 change from const -> constexpr
				
Version 1.111 - 2016-09-05 (RDS)
---------------
	Version for the delivery track - same as 1.110.1

Version 1.110.1 - 2016-08-26 (EO)
---------------
	+ src/caloaldefs.f90: For NAG Fortran the inclusion of the -fpp in all
	  compilations makes it more strict regarding formating. In this case
	  the file caloaldefs.f90 was full of empty lines between continued
	  statements (those ended with "&" signs), something that is
	  considered a syntax error only when compiled with -fpp. Once these
	  empty lines are suppressed, compilation works fine.

Version 1.110 - 2016-06-29 (AI)
---------------
	+ src/CalOalUtility.h: gfortran compatibility change
	for Dope (array descriptor) structures
	
Version 1.109.3 - 2014-11-13 (EO)
---------------
	+ src/xmm.h: stime.h replaced by stime_c.h after it has been renamed in
	  package utils.

ChangeLog

Version 0.9.1 - 2016-04-26 (RDS)
----------------------
- Updated Statistical method  documentation (SPR-7277)

Version 0.9 - 2015-11-27 (RDS)
----------------------
- Fixed problem with building documentation (SPR-7327)

ChangeLog

Version 0.4.2 - 2016-12-05 (AI)
-------------
	+ (src/catprep_aux.f90) The variables used to store the
	catalogue directory paths have been increased to 250 legnth.

Version 0.4.1 - 2016-10-10 (EO)
-------------
	+ Added missing test/testdata subdirectory to hold test fits files.

Version 0.4 - 2016-09-07 (AI)
-------------
- (src/cat_prepaux.f90) Parameters in catprep_read subrotuine  must
be declared as "inout" instead of "out", otherwise a seg-fault is
thrown out at run-time (gcc-6.2)	
	
Version 0.3.9 - 2014-03-30 (EC)
-------------
- Moved test/cat_data into testprods/test/catprep_data
- cleaned up some additional unused binary data

ChangeLog

Version 1.35.1 - 2009-09-09 (EO)
--------------
	+ rgs/rgslincoordgrcal.cc: GNUC >= 6 since g++ 5.3.1 does not use C++
	  2011 by default.

Version 1.35 - 2009-09-06 (AI)
------------
	+ (rgs/rgslincoordgrcal.cc) GCC 6 portability change from const -> constexpr

Version 1.34.1 - 2009-09-03 (AI)
------------
	+ Fixes to make ccftools gcc-4.3 compliant.

ChangeLog

Version 1.191.1 - 2016-12-02 (AI)
-------------	
	+ (src/common/dal_utilities.cc) There was a bug in the calculation of
	the number of elements in a string array. In Mac OSX (gcc 6.2) a seg.
	fault was thrown (don't undestand why Linux could cope with this bug...)

Version 1.191 - 2016-08-24 (RDS)
-------------	
	+ (src/dal_utilities.cc)     - Removed useless debug line

Version 1.190 - 2016-08-23 (RDS)
-------------	
	+ (src/ExtDal.cc) 	      - Changed array to arrayB to avoid
					problem with GCC 6
	   src/FitsReaderWriter.cc,
	   src/MemFitsReaderWriter.cc   Solved issues with logical expressions
	   				which were identified in GCC 6 testing

Version 1.189 - 2016-06-30 (AI)
-------------	
	+ (doc/reference/examples/stringcell.f90 setstringcell.f90)
	Test cases upgraded to be compatible with gfortran compiler.
	
Version 1.188 - 2016-06-29 (AI)
-------------	
	+ Version 1.188 identical to 1.2
	
Version 1.2 - 2016-06-29 (AI)
-------------	
	+ (f90/dal_implementation.cc) - Dope (array descriptor)
	compatibility changes with gfortran compiler.
	
Version 1.187 - 2014-04-15 (AI)
-------------	
	+ gcc4.8.2 portability changes
	

ChangeLog

Version 1.14.1   - 2016-04-25 (RDS)
-----------
	+ (src/DummyPixelIterator.cc/h, - INtroduced a fast conversion from
					  image pixel to DET coords
						(SPR-7357)

Version 1.14   - 2015-11-09 (RDS)
-----------
	+ DEPEND			- Added ssclib dependency

ChangeLog

Version 2.8 - 2016-05-21 (AI)
-----------
	+ (phasecalc.cc) Error handler exception properly coded in a
	try/catch structure

Version 2.7 - 2015-09-22 (AI)
-----------
	+ Bug related to the timesys convertion of UTC and TT/TDB times fixed.
	SOC-SPR-2853 fixed.
	

ChangeLog

Version 0.7.2 - 2016-10-28 (JVP)
-----------
	+ Documentation updated

Version 0.7.1 - 2016-07-20 (JVP)
-----------
	+ Documentation updated

Version 0.7 - 2016-06-22 (JVP)
-----------
	+ (src/ebkgreg.cc	- In addition to PN, MOS1/2 cameras are also covered by this task.
	   src/ProcSrcList.cc	- Background extraction region is CIRCLE, or ANNULUS only for MOS1/2 in
	   src/Bkgreg.cc	  SmallWindow mode (PrimePartialW2/3) and the source is in the
	   src/Bkgreg.h		  central CCD (CCDnr = 1). In this case the annulus has always the same size
	   src/Lookupradius.cc	  and it covers the external CCDs, but enclosing the central CCD into the inner
	   src/Lookupradius.h	  radius of the annulus. Thus no counts from the central CCD are taken
	   config/ebkgreg.par	  into account for the background estimate.
				- Include instrument label: PN, M1, M2 in the Source List columns
				- Three new columns included in Source List:
				- Background extraction region shape: CIRCLE or ANNULUS
					- *_RAD1 ( = Circle radius or inner radius for Annulus)
					  *_RAD2 ( = 0 for Circle or outer radius for Annulus)
				- New output parameters for RAD1 and RAD2 (arcsec and detector units)
				- Pixel size set depending on the EPIC camera (pixsizePN = 4.18, pixsizeM1/2 = 1.10)

Version 0.6.2 - 2015-12-21 (JVP)
-----------
	+ No background region is produced if the background extraction radius is less than 3 pixels

ChangeLog

Version 4.27.3 - 2016-12-03 (IT)
--------------
	+ Removed debugging comments from the code after testing v4.27.2.
	+ Updated documentation: Parameter withoffsets=yes (obsolete
	  eident offsets) yields a warning; no other effect anymore.

Version 4.27.2 - 2016-11-23 (IT)
--------------
	+ New algorithm to determine the subimages, in order to avoid
	  memory leaks which could lead to a large number of spurious
	  detections with the gfortran compiler. Tentatively removed the
	  obsolete support of eident offsets. Test version.

Version 4.27.1 - 2016-10-24 (IT)
--------------
	+ ebox_io.f90: fixed an old bug in the calculation of the degrees
	  of freedom of multiple observations, used for the equivalent
	  likelihoods per instrument and image. Has neither affected single
	  observations nor summary lines of the output table. SPR-7400.

Version 4.27 - 2016-09-30 (IT)
--------------
        + Parameter constraints: changed lower limit of parameter ecf
          from 0.001 to 0.0. Purpose: Fluxes can be set to zero (e.g. by
          edetect_stack), if no ecfs are defined.
	+ Number of detections in output list as informational output and
	  new fits keyword NDETECT. Minimum detection likelihood: LIKEMIN.
	  Pixel size: XBINSIZE, YBINSIZE.
	+ ebox_io.f90: Write summary line also if only one instrument is
	  active. Adjusted tests.
	+ edetect_io_box.f90: subimages via imagebuffersize: shift
	  reference point by -16 in x and in y (as in emldetect's
	  edetect_io_ml.f90) in order to make sure that subimages have
	  zero values in their edge rows and columns. "Chains" of spurious
	  detections between the subimage and the image edge may arise
	  otherwise.
	+ edetect_io_box.f90: Made determination of imbuff more concise.
        + DEPEND: updated to emldetect-7.0
	+ cf. SCR 7399: source detection on stacked observations


Version 4.26 - 2016-07-18 (AI)
--------------
	+ gfortran compatibiliy changes applied.
	
Version 4.25.2 - 2016-04-25 (IT)
--------------
        + Throw error "TooManyBands" if images for more than six energy
          bands and / or more than six pimin/pimax are provided. Documented
          error message and maximum number of bands and corrected typo in
          the documentation that resulted in a scrambled line on the
          "obsmode" parameter (SPR-7358 and SPR-7359).

Version 4.25.1 - 2015-10-01 (IT)
--------------
	+ Input data in detector coordinates: Use CDELT1L, CDELT2L
	  header keywords instead of CDELT1, CDELT2 in order to determine
	  the pixel size. For pn and MOS images in detector coordinates,
	  CDELT1/2 have different units.

ChangeLog

Version 3.15 - 2016-09-30 (IT)
------------
	+ DEPEND: updated to emldetect-7.0 and esplinemap-5.0

Version 3.14.2 - 2016-04-25 (IT)
--------------
	+ Throw error messages if eboxdetect & emldetect limits on number
	  of images, energy bands etc. are exceeded. Test for
	  SAS_CCF. Repeat the maximum number of energy bands in the
	  parameter section of the documentation. (cf. SPR-7360)

Version 3.14.1 - 2015-02-19 (IT)
--------------
	+ Documentation: Updated default values of 10 parameters
	  (cf. SPR-7275); added missing constraints of four parameters.

ChangeLog

Version 0.1.1 - 2016-10-20 (IT)
--------------------------
	+ test harness completed. Check whether tasks are run and output
	  source lists exist.
	+ minor code updates

Version 0.1 - 2016-10-07 (IT)
------------------------
        + Initial commit, comprising the perl files edetect_stack.pl and
	  stack_sourcelist.pl and a minimal test whether the edetect_stack
	  can be started. For details on the structure of the task,
	  consult the draft documentation.

ChangeLog

Version 4.11.1 - 2016-11-14 (AI)
------------	
	+ (src/eexpmap_mod.f90) Initialization to null (at declaration statement)
	of those pointer that will be used by the CAL to avoid seg fault in gfortran.

Version 4.11 - 2016-07-20 (AI)
------------
	+ (src/eexpmap_mod.f90) compatibility changes with gfortran applied.
	The Bmatrix is passed to C++ with a wrong array descriptor elements (dtype and triplet).
	Only if we create the local variable matrix, we can pass the right values to C++. Not clear
	why this is happening, we have to keep an eye on it.
	The OAL_toEulerAngles subrotutine has to be called with the matrix transpose.
	Otherwise, the C++ function handle the matrix elements in the wrong order. We
	have a similar effect in  rgslib(euler321_mod.f90) module.
	+ (src/Makefile) gfrotran preprocessor macros added.
	
Version 4.10.1 - 2015-12-16 (RDS)
------------
	+ test/expo_refima.fits.gz:

	Changed the template test exposure map back to give the same result as
	the calculation with the old boresight file (XMM_BORESIGHT_0015)

ChangeLog

Version 2.20 - 2016-08-26 (EO)
------------
	+ src/Makefile: Added code to proccess Fortran preprocessor
	directives.
	+ DEPEND: Updated to sas-1.573 that implements new policy for
	preprocessor compilations.

Version 2.19.1 - 2016-08-25 (EO)
--------------
	+ src/Makefile: Removed ifeq to set F90FLAGS depending on Fortran
	  compiler. This was left unchanged in 2.19.
	+ DEPEND: sas 1.572 is required to build properly this version.

Version 2.19 - 2016-08-24 (EO)
------------
	+ src/efftplot_mod.f90: Added preprocessor directive ifdef
	 NAGf90Fortran before use f90_unix_proc to allow to load this NAG
	 Frotran supplied module only when NAG Fortran compiler is used.
	 This change is effective only after SAS configure includes
	 preprocessor capabilities (after sas-1.571). Otherwise a similar
	 functionality must be available in src/Makefile.
	 For NAG Fortran compiler, the inclussion of preprocessor switch -fpp
	 unveiled a syntax error in original line 324, a blank line after
	 "window         ",&
	 which has been removed.

Version 2.18.2 - 2012-19-04 (NW)
------------

	The coordinates given in the header of the output of efftplot
	are now displayed to .00 for RA and .0 for Dec. even when the
	(arc)seconds are inferior to 10, as requested in SPR 6714.

ChangeLog

Version 0.5 - 2016-05-19 (RDS)
------------------------------
	+ src/Efluxer.cc               - Put units as ergs/s/cm2/keV in
					 the output file (SPR-7362)

Version 0.4.6 - 2013-04-02 (AI)
------------------------------
	+ Response manipulation functionality extracted to a new class (RMF.h)
	+ src/Efluxer.cc updated to use the new RMF class
	

ChangeLog

Version 2.7 - 2016-08-26 (EO)
-----------
	+ src/Makefile: After changing the preprocessor policy for all Fortran
	  compilations, added code here to allow for inclussion of
	  NAG Fortran specific modules.
	+ DEPEND: Updated to sas-1.573.


Version 2.6 - 2016-08-25 (EO)
-----------
	+ src/ekstest_mod.f90: Added an "ifdef NAGf90Fortran - endif"
	  preprocessor directive around "use f90_unix_proc" to allow its
	  loading only when NAG Fortran compiler is used.
	+ DEPEND: Requires sas 1.572.


Version 2.5.1 - 2014-10-24 (RDS)
----------------------------
	Documentation update (lccorr -> epiclccorr) - SPR-7233

ChangeLog

Version 8.8 - 2016-09-08 (AI)
-------------
	+ (src/emenergy_mod.f90) Initialization to null of all pointers that
	are passed to C++ (Dope vectors). If we do not initizialize the
	pointers, we get seg fault at run-time with gcc-6.2.	

Version 8.7 - 2015-05-13 (AI)
-------------
	+ Documentation updated. SOC-SPR-7226 fixed
	

ChangeLog

Version 8.7 - 2016-09-07 (AI)
-----------
- (src/emevents_mod.f90) More pointers initialized to null...
	
Version 8.6 - 2016-09-06 (AI)
-----------
- (src/emevents_mod.f90) Initialization to null of all pointers that are passed
  to C++ (Dope vectors). If we do not initizialize the pointers,
  we get seg fault at run-time with gcc-6.2.	

Version 8.5 - 2016-07-18 (AI)
-----------
- gfortran compatibility change
	
Version 8.4 - 2006-01-27 (JB)
-----------
- Set y_size_window to 1025 in compressed timing mode.

ChangeLog

Version 5.10 - 2016-09-06 (AI)
-------------
- (src/emframes_mod.f90) Initialization to null of all pointers that are passed
  to C++ (Dope vectors). If we do not initizialize the pointers,
  we get seg fault at run-time with gcc-6.2.	

Version 5.9 - 2016-07-18 (AI)
-------------
	+ (src/emframes_mod.f90) gfortran compatibility change applied.

Version 5.8.1 - 2010-03-10 (JB)
-------------
- Fix SPR 6600: problem with OBSID 0611180101 which occurred in timing mode
  when most frames were lost to counting mode.

ChangeLog

Version 7.0.3 - 2016-11-08 (IT)
-------------
	+ fftpsf.f90: Sloppy indexing caused occasional segmentation
	  faults, in particular if emldetect is compiled with gfortran.
	  Also make sure that arrays are initialized with 0 for gfortran
	  compatibiliy. Fixes SPR-7405.

Version 7.0.2 - 2016-10-26 (IT)
-------------
	+ typo - sorry

Version 7.0.1 - 2016-10-24 (IT)
-------------
	+ eml_io.f90: fixed an old bug in the calculation of the degrees
	  of freedom of multiple observations, used for the equivalent
	  likelihoods per instrument and image. Has neither affected single
	  observations nor summary lines of the output table. SPR-7400.
	+ documentation updated

Version 7.0 - 2016-09-30 (IT)
------------
        + new parameter "withrawrows" for processing of stacked images
	  (expert-only use):
	  Include additional rows in the output source list per image in
	  which the source was detected, in case the source position is
	  within the detection mask of the observation / image
	  set. Additional rows are identified by a modified ID_INST,
	  composed of the observation ID followed by the instrument
	  number. ID_INST becomes double precision displayed as a long
	  integer in this case and the usual integer in the default
	  case. Store EXP_IDs in new keywords EXP_ID_INn with n giving the
	  first digits of the modified ID_INSTs.  Consequences: New global
	  variable "exp_id", argument to get_files.  Flag "rawflag" in
	  subroutine mlout.
        + Parameter constraints: changed lower limit of parameter ecf from
          0.001 to 0.0. Purpose: Fluxes can be set to zero
          (e.g. by edetect_stack), if no ecfs are defined.
	+ Header keywords HRPNDEF, HRM1DEF, HRM2DEF added: will be read by
	  edetect_stack.
	+ emldata.f90: increased max_events and max_sou
	+ edetect_io_ml.f90: increased max_ins and max_ima
	+ emldetect_mod.f90: throw error message, if number of photons
	  exceeds max_events (-> make it clearer to users)
	+ edetect_io.f90: same for max_ins
	+ eml_io.f90: number of detections in output source list as new
	  informational output at high verbosity level and header keyword
	  NDETECT. Minimum detection likelihood: LIKEMIN.  Pixel size:
	  XBINSIZE, YBINSIZE (to be consistent with eboxdetect headers).
	+ eml_io.f90: check whether coordinate columns are NULL.
	  Motivation: allow to use srcmatch lists as input source lists in
	  which X_IMA Y_IMA are NULL and RA, DEC non-NULL.
	  Caveat: emldetect sorts the input list according to its SIGMA or
	  LIKE or DET_ML or WCORR column. srcmatch input lists remain
	  unsorted thus.
	+ eml_io.f90: Show a warning if none of these four likelihood
	  columns is found in the input list.
	+ emldata.f90: changed dimension of hrdef and xiddef from
	  (...,max_ins) to (...,3): pn, m1, m2
	+ emlfit.f90: Runtime improvements for large arrays (loops).
	+ fftpsf.f90: More concise and thus faster version of the
	  re-binning. Source parameters may change slightly compared to
	  the previous version because of the limited number precision.
	+ edetect_io_ml.f90: Made determination of imbuff more concise.
	+ config/emldetect.par: sort parameters for better readability;
	  separate single values from lists, mandatory from obsolete.
	+ Cleaned code in comments sections for better readability:
	  Removed old code versions of before ~2006 and debugging
	  "write"s. They can be still accessed in the archive.
	+ cf. SCR 7399: source detection on stacked observations


Version 6.3 - 2016-09-08 (AI)
--------------
	+ (src/edetect_io_ml.f90) Initialization of pointers using
	"nullify" statement instead of using "=> null()" initialization
	at pointer declaration. NAGWare 5.2 gives wrong array sizes for ChipX,ChipY,
	theta and phi pointers if we use the "=> null()" statemtent, but it gives
	correct array sizes if we initialize the pointers with "nullify()" statement.
	With gcc-6.2 both initizalization gives correct results.

Version 6.2 - 2016-09-07 (AI)
--------------
	+ (src/eml_io.f90) Initialization to null of all pointers that are passed
	to C++ (Dope vectors). If we do not initizialize the pointers,
	we get seg fault at run-time with gcc-6.2.	
	
Version 6.1 - 2016-07-18 (AI)
--------------
	+ (src/eml_io.f90) gfortran compatibility changes applied.

Version 6.0.5 - 2016-04-25 (IT)
--------------
	+ Throw error "TooManyBands" if images for more than six energy
	  bands and / or more than six pimin/pimax are provided. Documented
	  error message and maximum number of bands (cf. SPR-7358 on
	  eboxdetect).

Version 6.0.4 - 2016-01-25 (IT)
------------
        + Documentation: Expanded description of fluxes and ECFs
          (SPR-7306). Re-structured the description section.

ChangeLog

Version 1.17 - 2016-03-16 (JVP)
--------------
	+ Update documentation typo: \task{DAL}-supported by \task{dal}-supported ( SPR-7262  )

Version 1.16 - 2011-04-26 (RDS)
--------------
	+ src/emosaic_mod.f90	

- Made compatible with image and mask files. It uses the PRIMARY array if
found, otherwise it uses the MASK array or throws an error.
					(SCR-314)

ChangeLog

Version 0.4.1 - 2016-03-28 (AI)
-----------
	+ Typo in the parameter imagesets of emldetect call fixed.

Version 0.4 - 2015-12-04 (AI)
-----------
	+  Inconsistent order of input images sets and pi ranges fixed in the the call
	to emldetect and eboxdetect. (SOC-SPR-7290) closed
	

ChangeLog

Version 2.31.1 - 2016-12-05 (AI)
------------
	+ (test/testEnergyCombinef90 & test/runemutils.f90) pointers, that are
	passed to C++ interface, initialized to null to avoid seg. faults with
	gfortran.
	
Version 2.31 - 2016-07-18 (AI)
------------
	+ (src/CalF90cxxGlue.cc) gfortran compatibility changes regarding
	array descriptor (dope vector) definitions.
	+ (scripts/compare_realcols_mod.f90) "public" statement added to
	compare_realcols subrotutine
	
Version 2.30 - 2015-11-18 (AI)
------------
	+  New compatibility with RGS images to detect hot pixels and columns
	

ChangeLog

Version 1.20 - 2016-09-16 (MJF)
-------------------------------
- (src/epatplot_mod.f90): warning if empty EVENTS extension

Version 1.19 - 2013-11-14 (MJF)
-------------------------------
- (src/epatplot_mod.f90): plot also real-valued SRCPOSX and SRCPOSY for
   EPIC-pn if new parameter "withsrcxy=Y" (default "Y") and SRCPOSX and
   SRCPOSY keywords found in event file header
- (config/epatplot.*): as above
- (doc/epatplot_description): as above

ChangeLog

Version 8.76 - 2016-07-15 (MJF)
-------------------------------
- (src/epchain): added parameter "withxrlimage" (task "epxrlcorr", default "N")
   that controls creation of a diagnostic X-ray loading image image (SCR-7114)
- (config/epchain.*): as above
- (doc/epchain_description.tex): as above
- (DEPEND): needs `epxrlcorr-1.0'

Version 8.75 - 2015-04-01 (MJF)
-------------------------------
- (src/epchain): allow input list of offset maps (instead of the ones
  identified by the OAL, or sometimes not identified ;-)
  This is useful in the case when the offset maps were dumped before or after
  the observation with different ObsID (e.g., early mission data, slews, ...)
  Moreover, this new feature enables calibration work without changing contents
  of ODF/SDF data or CCF files. The new interface is used in task "epreject".
  This is controlled via new boolean parameter "withoffsetlist" (default "N")
  that takes precendence over "withoffsetmap" if set to "Y": it opens the file
  specified in new parameter "odilist" and reads the ODI name to be used
  from there. The input list is an ASCII with 2-digit CCD number followed
  by a blank and then the path to the offset map for that CCD (there can be
  more than one offset map in that file so that the input file can be used
  for the whole observation processing; only the first entry for a particular
  CCD is used; the order does not matter). (SCR-7284)
- (config/epchain.*): as above
- (doc/epchain_description.tex): as above
- (DEPEND): needs `epreject-5.16'

ChangeLog

Version 6.50.1 - 2016-10-27 (RDS)
----------------------------------
- (src/epevents_mod.f90) Initialization to zero of ipat4 array
			 (SPR-7401)

Version 6.50 - 2016-00-06 (AI)
----------------------------------
- (src/epevents_mod.f90) Initialization to null of all pointers that are passed
  to C++ (Dope vectors). If we do not initizialize the pointers,
  we get seg fault at run-time with gcc-6.2.	

Version 6.49.10 - 2016-01-24 (MJF)
----------------------------------
- (src/epevents_mod.f90): complete change to avoid I*1 overflow
  to allow compilation with NAG-6.0 compiler on Mac-OSX

ChangeLog

Version 0.14 - 2016-09-16 (MJF)
-------------------------------
- (src/epexposure_mod.f90): take care of empty input EVENTS extension

Version 0.13 - 2014-05-26 (MJF)
-------------------------------
- (src/epexposure_mod.f90): special warning "offsetCorrectionNotHomogeneous"
  if applied offset corrections for the CCDs are not the same
  (SOC/SAS-SCR-7199)
- (DEPEND): needs "epreject-5.15"

ChangeLog

Version 8.109.1 - 2016-10-09 (MJF)
----------------------------------
- (src/epframes_mod.f90): disable dummy HK extensions at compile time

Version 8.109 - 2016-09-06 (AI)
---------------
- (src/epframes_mod.f90) Initialization to null of all pointers that are passed
  to C++ (Dope vectors). If we do not initizialize the pointers,
  we get seg fault at run-time with gcc-6.2.	

Version 8.108.1 - 2016-08-26 (EO)
---------------
	+ src/epframes_mod.f90: Due to the introduction in the SAS configure
	  of the "-cpp" option for Fortran code preprocessing, the compiler is
	  more strict with syntax rules.
	  The line 6066 is a comment line but contains a "\" at the end of the
	  line. When this code is fed to the gfortran compiler with the -cpp
	  option, the preprocessor interprets it as a continuation sign, hence,
	  the following line, which is an "if ... then" statement, is not taken as
	  such. The consequence is that the "endif" at 6660 is interpreted
	  as orphan and a syntax error is issued. By removing the "\",
	  everything is fine.

Version 8.108 - 2016-07-20 (MJF)
--------------------------------
- (src/epframes_mod.f90): remove obsolete statement "use f90_unix_env"
- (src/epframes_mod.f90): added dummy HK extensions output for interface
  development of intra-exposure corrections based on HK (SCR-7312)

Version 8.107 - 2016-07-19 (MJF)
--------------------------------
- (src/epframes_mod.f90): check E_Cx_IVDDA (related to NCR-138)
- (src/epframes_mod.f90): gfortran compatibility change related to environment
  variable SAS_JUMP_TOLERANCE: getenv() -> get_environment_variable()
- (src/epframes_mod.f90): compare CAL and enviroment JUMP_TOLERANCE values

Version 8.106 - 2015-10-14 (RDS)
--------------------------------
- (doc/epframes_description.tex): improved description of JUMP tolerance
					(SPR-7056)

ChangeLog

Version 1.21.1 - 2016-10-27 (AI)
------------
	+ (src/FrameList.cc) set random number generator between -1 and 1.
	
Version 1.21 - 2016-10-06 (AI)
------------
	+ (src/FrameList.cc) New method getRandomTime() added. This method
	returns a timestamp with a particular frame (used in evqpb task).
	Related to SCR-7361.
	
Version 1.20 - 2015-12-09 (AI)
------------
	+ Parameter "extendedsource" set you "yes" in the arfgen background
	call.
	

ChangeLog

Version 2.21 - 2016-08-26 (RDS)
-------------- 		
	+ DEPEND - updated

Version 2.20 - 2015-12-09 (AI)
-------------- 		
	+ mos doc updated (SOC-SPR-7100)
	

ChangeLog

Version 0.9 - 2016-09-06 (AI)
-----------------------------
- ((src/epprintinfo_mod.f90) Initialization to null of all pointers that are passed
  to C++ (Dope vectors). If we do not initizialize the pointers,
  we get seg fault at run-time with gcc-6.2.	
	
Version 0.8 - 2016-07-18 (AI)
-----------------------------
	+ (src/epprintinfo_mod.f90) index variable in loop write
	statements properly defined.
	
Version 0.7 - 2016-06-30 (AI)
-----------------------------
	+ (src/epprintinfo_mod.f90) gfortran compatibility changes.
	
Version 0.6 - 2013-07-04 (MJF)
------------------------------
- (src/epprintinfo_mod.f90): call CAL_camCoord1ToChipRealCoord() in routine
   SourcePositionOnDetector() for real-valued CCD coordinates
- (DEPEND): needs "cal-3.216"

ChangeLog

Version 5.17 - 2016-05-06 (MJF)
-------------------------------
- (src/epreject_mod.f90): test on existence and completeness of offset map
  list introduced in previous version. Write also keyword OFSTFILE to EXPSOURE
  extension as the used offset map may originate from a different ODF.

Version 5.16 - 2015-03-24 (MJF)
-------------------------------
- (src/epreject_mod.f90): allow input list of offset maps (instead of the ones
  identified by the OAL, or sometimes not identified ;-)
  This is useful in the case when the offset maps were dumped before or after
  the observation with different ObsID (e.g., early mission data, slews, ...)
  Moreover, this new feature enables calibration work without changing contents
  of ODF/SDF data or CCF files.
  This is controlled via new boolean parameter "withoffsetlist" (default "N")
  that takes precedence over "withoffsetmap" if set to "Y": it opens the file
  specified in new parameter "odilist" and reads the ODI name to be used
  from there. The input list is an ASCII with 2-digit CCD number followed
  by a blank and then the path to the offset map for that CCD (there can be
  more than one offset map in that file so that the input file can be used
  for the whole observation processing; only the first entry for a particular
  CCD is used; the order does not matter). (SOC/SAS-SCR-7284)
- (config/epreject.*): as above
- (doc/epreject_description.tex): as above

ChangeLog

Version 1.0 - 2016-07-14 (RDS)
-----------

	+ src/Epxrlcorr.cc		- Optionally produces a diagnostic
					  output image (SCR-7114)

Version 0.1.7 - 2013-11-14 (RDS)
-----------

	+ src/Epxrlcorr.cc		- Refuses to run iof keyword
					  REJ_CORR is set

ChangeLog

Version 0.9.39 - 2016-10-18 (BP)
	+ Removed tasks cheese-bands and conv-region (both redundant).
	+ Re-added task cheese (removed by mistake, SPR:7402).
	+ Cleaned up config files.
	+ Adjusted hardcoded array to accommodate new QPB files (subsn).
	
Version 0.9.38 - 2016-10-18 (BP)
	+ Selection expression changes in perl task cheese.
	+ QDP output cleanup in adapt.
	+ Removal of esas-mosaic, esprep\_mosaic, esprep\_mosaic\_time perl.

Version 0.9.37 - 2016-10-17 (BP)
	+ Removed parameter chisqfl from mos\_back, pn\_back.
	+ Removed sls code from perl subdir to avoid paralled devel confusion.

Version 0.9.36 - 2016-09-08 (AI)
	+ Tag added to conv_reg_mod.f90 source code to enable (NAGWare) or
	disable (gfortran) to use of the f90_unix_proc module needed in NAGWare
	for system call.

Version 0.9.35 - 2016-07-21 (BP)
	+ Proof of concept "old esas" still compiles in gfortran.
	+ Cleaned up failing laTex.
	+ removed f90_unix_proc from several tasks.
	+ Cleaned up antique F77 code.

Version 0.9.34 - 2016-02-18 (BP)
	+ Corrected inclusion of task merge\_comp\_xmm.

Version 0.9.33 - 2016-01-10 (BP)
	+ Removed CCD4 post-CCD3-loss if statements (selection expressions) from mos-spectra.
	+ Fixed bug in conv\_reg (erroneous single quote found in select expr).
	+ Deprecated tasks esprep\_mosaic and esprep\_mosaic_time. The functionality of these are now done in the PPS with the subpointings in mosaic mode.
	+ Fixed bug in merge\_comp\_xmm for rectangular matrices.

ChangeLog

Version 1.10 - 2016-09-08 (AI)
----------------------
- (src/econvolverprep_mod.f90) Initialization to null of all
pointers that are passed to C++ (Dope vectors). If we do
not initizialize the pointers, we get seg fault at run-time with gcc-6.2.

Version 1.9 - 2009-02-02 (IMS/JBO)
----------------------
- Originally developed by IMS/JBO on 2006-10-19.
- Upgraded ssclib dependency to 4.0.
- All logical(bool) types changed to logical.
- Fixed bug in which numBands was being obtained from some public declaration in a used module.

ChangeLog

Version 3.12.2 - 2016-04-22 (IT)
------------
	Documentation: Typo corrected.

Version 3.12.1 - 2016-01-25 (IT)
------------
	Documentation: Description of parameter mlmin. Parameter limits
	updated (SPR-7339).

ChangeLog

Version 1.18 - 2016-07-07 (RDS)
----------------------
	+ src/edet2sky_mod.f90		- Checked if raw coordinate entry
					  is on or off chip

Version 1.17 - 2015-08-24 (RDS)
----------------------
	+ Support for detector coord images in edet2sky

ChangeLog

Version 1.11 - 2016-09-09 (RDS)
-------------------------------
	+ src/eslewsearch.pl            - Set extension name of combined
	  				  lists to SRCLIST (SPR-7398)

Version 1.10 - 2016-08-11 (RDS)
-------------------------------
	+ src/eslewchain.pl             - Converted to use taskmain
	  src/eslewsearch.pl		  (SCR-7370)
	  config/eslewchain.par
	  doc/eslewchain_description.tex

Version 1.9 - 2016-06-30 (RDS)
-------------------------------
	+ src/eslewsearch               - New task to perform a source search
					  on slew subimages (SCR-7366)

Version 1.8 - 2016-02-17 (RDS)
-------------------------------
	+ doc/eslewchain_description.tex - Documentation update

Version 1.7.1 - 2016-01-12 (RDS)
-------------------------------
	+ src/eslewchain		- Use ecoordconv rather than
					  eregionanalyse to convert from X/Y
					  to RA/DEC (SPR-7334)

ChangeLog

Version 1.44.2 - 2016-03-11 (RDS)
-----------
	+ doc/eupper_description.tex          - Better description of upper
	  doc/eregionanalyse_description.tex	limit calculation

Version 1.44.1 - 2016-01-12 (RDS)
-----------
	+ src/ecoordconv.cc             - Added option, pos2eqpos, which
	  config/ecoordconv.par		  just performs the x/y->ra/dec
					  conversion, avoiding problems where
					  the help point is >90 degs away from
					  the image centre (SPR-7334)

ChangeLog

Version 2.6 - 2016-09-08 (AI)
	+ Tag added to espfilt_mod.f90 source code to enable (NAGWare) or
	disable (gfortran) to use of the f90_unix_proc module needed in NAGWare
	for system call.
	
Version 2.5 - 2016-07-20 (BHP)
	+ Fixed laTeX bug causing make doc failure.
	+ Compliant with gfortran (removed use f90_unix_proc)

Version 2.4 - 2015-11-17 (BHP)
	+ Fixed laTeX bug causing make doc failure.

ChangeLog

Version 5.0 - 2016-10-07 (IT)
--------------
	+ parameter nsplinenodes: maximum increased to 40 (for use with
	  large images)
	+ new fitmethod smooth (main part of the code: GL for eRosita).
	  Introduced parameters snrmin and smoothsigma, module esp_smooth_mod.
	  Modified exdb.f, restor.f, dbpari.f90, include/cdbio.cmn,
	  include/cdbio_f90.cmn. Introduced wrinfo.f90.
	+ ootmap_mod.f90: nullify CAL pointers for compliance with
	  gfortran 6.2
	+ moved many unused variables into comments (still some left ...)
	+ cf. SCR 7399: source detection on stacked observations


Version 4.5 - 2014-07-24 (IT)
--------------
	+ cp_head_mod.f90: copy full dataset and array headers of the
	  input image to the background maps (instead of a list of selected
	  keywords). In particular: make physical coordinates available.
	+ documentation: Updated parameter descriptions.

ChangeLog

Version 1.7 - 2016-09-07 (AI)
-----------
- (src/evigweight_mod.f90) Initialization to null of all pointers that are passed
  to C++ (Dope vectors). If we do not initizialize the pointers,
  we get seg fault at run-time with gcc-6.2.		

Version 1.6 - 2003-02-28 (JB)
-----------
- Changed test harness to follow new QUANTUMEF in testccf 1.102.

ChangeLog

Version 0.4.7 - 2017-01-09 (AI)
----------------------------
	+ FILTER keyword set to "Colsed" in all output extensions.

Version 0.4.6 - 2017-01-01 (AI)
----------------------------
	+ Parameter "calctlmax" from attcalc added to the task.

Version 0.4.5 - 2016-12-27 (AI)
----------------------------
	+ Error message if instrument mode is not Full
	Frame Window changed.
	
Version 0.4.4 - 2016-12-20 (AI)
----------------------------
	+ (config/evqpb.info) VISIBLE tag set to true.
	+ New Column REVNO added to the output file.
	+ doc updated

Version 0.4.3 - 2016-11-23 (AI)
----------------------------
	+ New column "EWEIGHT" added with the exposure weight per CCD.

Version 0.4.2 - 2016-10-27 (AI)
----------------------------
	+ Dead CCDs removed from FWC data computations.
	
Version 0.4.1 - 2016-10-21 (AI)
----------------------------
	+ Set instrument in cal calls to retrieve correct ccds numbers.
	+ Remove PAT_ID, PAT_SEQ and TIME_RAW for EPIC PN event files.
	
Version 0.4 - 2016-10-07 (AI)
----------------------------
	+  keyword XPWEIXX updated to a
	percentage.

Version 0.3 - 2016-10-06 (AI)
----------------------------
	+  Added keyword XPWEIXX: Exposure
	Weight per CCD
	
Version 0.2 - 2016-10-06 (AI)
----------------------------
- First stable version. SCR-7361 closed.


Version 0.1 - 2016-09-26 (AI)
----------------------------
- Directory structure created by pkgmaker.

ChangeLog

Version 1.52.10 - 2017-01-12 (AI)
------------
	+ src/SasTask.h               Resolved problem with name() returning blank string

Version 1.52.9 - 2009-08-31 (AI)
------------
	+ Modifications is several source files to make gui gcc 4.3 compliant.
	

ChangeLog

Version 0.4 - 2015-09-15 (JVP)
----------------------
- Change circle color notes

Version 0.3 - 2015-09-06 (JVP)
----------------------
- Update dependencies file DEPEND

Version 0.2 - 2015-09-05 (JVP)
----------------------
- Re-uploaded because of minor git problem

Version 0.1 - 2015-09-02 (JVP)
----------------------
- Original version.

ChangeLog

Version 1.8 - 2016-09-08 (AI)
-----------
	+ (tools/testCistream.cc) gcc-6.2 compatibility changes.

Version 1.7 - 2011-05-05 (EO)
-----------
	+ src/CommentBuf.h: Included missing , which produced compilation errors
	  of the formm "'EOF' was not declared in this scope", when compiler with gcc 4.6.0
	  without '-fpermissive'.

ChangeLog

Version 1.21 - 2016-07-15 (AI)
-------------------------------
	+ (src/lcplot_mod.f90) "label" variable renamed (this word is
	not allowed in gfortran).

Version 1.20.1 - 2013-03-05 (NW)
        -------------------------------

	A check for NaN values is now included and a warning is called if
	all of the data points have values=NaN. The PSF FRAC. INCLUDED
	length has also been reduced to make the output plot clearer. The
	documentation has been updated to include the new warning.
	

ChangeLog

Version 0.6 - 2016-07-18 (AI)
-------------
- (src/ingestsrcnames_aux.f90) grfrotran refractoring: Usage of an optimal paramater as part of an allocatable variable definition is not possible in gfortran.
- (src/makethumbs_aux.f90) gfrotran refractoring: leading blanks spaces removed properly.
- (test/testmakethumbssimple) test comparasion on (X,Y) pixels values rounded to the fifth decimal place to avoid specific platform compiler representations in perl.
	
Version 0.5 - 2012-10-10 (MAS)
-------------
- Added a subroutine allocateSrcPrm() in ingestsrcnames_aux.f90 to make it more robust, particularly for the types of the column other than Int32.
	

ChangeLog

Version 2.78.1 - 2016-11-29 (AI)
------------	
	+ (src/omutility_mod.f90) Increase to 6 digits the integer casting to char
	and adjust string to the left in the output region file.

Version 2.78 - 2016-08-16 (AI)
------------	
	+ (src/omutility_mod.f90) gfortran compatibility changes
	(casting integers to char)
	
Version 2.77 - 2016-07-21 (AI)
------------	
	+ (src/omskybackground_mod.f90) casting to a integer kind 9
	to avoid multiplication overflow in AllocateSingle subroutine
	+ (src/omutility_mod.f90) gfortran compatibility changes applied
	(write output format statement)
	
Version 2.76.1 - 2014-05-22 (AI)
------------	
	+(src/omphotometry_mod.f90) use modules reordering to avoid error
	with NAG fortran compiler 6.2
	

ChangeLog

Version 3.133.1 - 2016-09-09 (EO)
---------------
	+ (src/AttituteLocator.h AuxHKaccess.h): GNUC >= 6 since
	 g++ 5.3.1 does not include C++ 2011 by default.

Version 3.133 - 2016-09-06 (AI)
---------------	
	+ (src/AttituteLocator.h OalServer.h AuxHKaccess.h) macro preprocessor
	added to constexpr statement to keep gcc backward compatibility.

Version 3.132 - 2016-09-05 (AI)
---------------	
	+ (src/AttituteLocator.h OalServer.h AuxHKaccess.h) gcc-6.X
	portability changes. (constexpr added to "static const" statements)
	+ (f90test/oalexample.f90 oaltest.f90) pointers initialized to null().
	(mandatory fo gcc-6.X)

Version 3.131 - 2016-07-14 (AI)
---------------	
	+ (f90/oal.f90) CONTIGUOUS statement added in the OAL_toEulerAngles
	subroutine

Version 3.130 - 2016-07-06 (RDS)
---------------	
	+ (f90/oal.f90) Remove contiguous option to get it to build

Version 3.129 - 2016-07-01 (AI)
---------------	
	+ (f90/Makefile) Conditional compiler option added

Version 3.128 - 2016-06-30 (AI)
---------------	
	+ (f90/OalF90cxxGlue.f90) gfortran compatibility changes for
	Dope (array descriptor) structures.
	+ (f9/Makefile and f90/oal.f90) fortran CONTIGUOUS statement is
	not present in NAGfor 5.2. To solve this dependency,
	we have added a preprocesor macro.
	
Version 3.127 - 2016-03-11 (RDS)
---------------	
	+ DEPEND		- added testodf into DEPEND (SPR-7349)

Version 3.126.1 - 2015-12-17 (RDS)
---------------	
	+ test/myboresight.ccf - changed to ALGOID=1 for CAL compatibility

ChangeLog

Version 1.6 - 2016-09-07 (RDS)
	src/coordinates.f90        accept directory names containing a "."
	src/ommastercatalogue_mod.f90   open dataset with MODIFY and line
	lengths
Version 1.5.1 - 2014-04-02 (EC)
	moved test/inp to testprods/omcat_data

ChangeLog

Version 5.36.1 - 2016-11-18 (RDS)
-------------------------------			
	+ (src/omdetectdetection_mod.f90) set default value to
	                                  .true. in the function
				          backgroundGradientOkay1

Version 5.36 - 2016-09-21 (AI)
-------------------------------			
	+ (src/omdetect_mod.f90) (nx,ny) variables cast to integer kind 9
	to avoid multiplication overflow in ImagingDetection subroutine

Version 5.35 - 2016-07-18 (AI)
-------------------------------			
	+ (rc/parameters_mod.f90) gfortran compatibility changes applied.

Version 5.34.1 - 2016-01-12 (AI)
-------------------------------			
	+ (src/sourceanalysis_mod.f90)  use modules reordering to avoid error
	with NAG fortran compiler 6.2
	

ChangeLog

Version 1.10 - 2016-09-07 (AI)
-------------
	+ (src/omgprep_mod.f90) Initialization to null of all pointers
	that are passed to C++ (Dope vectors). If we do not initizialize
	the pointers, we get seg fault at run-time with gcc-6.2.	

Version 1.9.1 - 2016-01-12 (AI)
-------------
	+ (src/utility_mod.f90)  Memory allocation error using NAG fortran
	6.2 fixed.
	

ChangeLog

Version 0.4 - 2016-07-15 (AI)
----------------------------
	+ (src/omgrismresp_mod.f90) "getenv" system call replaced with
	"get_environment_variable" F90 standard procedure.
	+ (src/omgrismresp_mod.f90) "f90_unix_env" nagfor module removed.
	+ (src/Makefile) Macro statements removed

Version 0.3 - 2016-07-14 (AI)
----------------------------
	+ (src/Makefile) Makefile fixed
	
Version 0.2 - 2016-06-29 (AI)
----------------------------
	+ (src/omgrismresp_mod.f90) Added a conditional def macro for
	the NAGWare f90_unix_env module. Makefile
	updated to add the corresponding compiler options.
	
Version 0.1.1 - 2014-11-27
----------------------------
	Modified DEPEND

ChangeLog

Version 1.10 - 2016-07-19 (AI)
-----------	
	+ (src/DopeVector.h) gfortran array descritor definition added
	
Version 1.9 - 2014-05-22 (AI)
-----------	
	+ (src/OmSourceParameters.h, src/OmgSourceParameters.h)
	 c++ struct refractoring to avoid seg faults in 64bit platforms
	+ (src/omgsourceparameter_mod.f90) Derived type modified acoordingly
	to match c++ struct variable order.
	

ChangeLog

Version 1.0.5 - 2016-04-21 (AI)
------------------------
	+ Error when there is not any repeated SRCID in final output file fixed.
	
Version 1.0.4 - 2016-03-30 (AI)
------------------------
	+ Added to SRCMOS column all SRCID OBSMOS sources megered into output file.

ChangeLog

Version 2.47 - 2016-07-18 (AI)
------------------------------	
	+ (src/omprep_mod.f90) gfortran compatibility changes applied.

Version 2.46.1 - 2014-09-09 (RDS)
------------------------------	
	+(src/checkpath.cc): checked for the LAST occurrence of the
	"_OM" string when finding the exposure number to avoid problems with
	directories which contain this string - SPR-7078
------------------------------	

ChangeLog

Version 1.4 - 2016-07-18 (AI)
-------------
	+ (src/addflags_mod.f90) gfortran comapatibility changes applied.
	+ (src/Makefile) macro statements added for gfortran compatibility.
	
Version 1.3 - 2014-05-22 (AI)
-------------
	+ (src/OmSourceParameters.h)  c++ struct refractoring to avoid seg
        faults in 64bit platforms
	

ChangeLog

Version 1.26 - 2016-07-19 (AI)
-------------	
	+ (src/DopeVector.h) gfortran array descritor definition added
	+ (src/omsource_old.f90) subroutines added as public
	setESourcePhotometryMethod and setweightingmethod added as public

Version 1.25.1 - 2016-01-12 (AI)
-------------	
	+ (src/omsource_old.f90) use modules reordering to avoid error
	        with NAG fortran compiler 6.2

ChangeLog

Version 2.22.2 - 2016-04-08 (AI)
--------------	
	+ (src/omimagealign_mod.f90) OM windows assigment fix to take into account
	windows repetitions in different files
	
Version 2.22.1 - 2016-04-05 (AI)
--------------	
	+ (src/omimagealign_mod.f90) Loop to assign OM windows to files modified.

ChangeLog

Version 1.4 - 2016-06-30 (AI)
-----------------------------
	+ (src/omthlcplot_mod.f90) gfortran compatibility changes.
	"label" variable name is not allowed in gfortran.
	
Version 1.3 - 2002-09-12 (NW)
-------------------------------

        Changed the keyword warning statement in the documentation to
 	include the values of the HDUCLASS attributes, to close SPR-2973
 	(Wrong reference to data set attributes in doc).  Also changed the
 	figure in the documentation to .eps (was .ps) to avoid warnings
 	during the distribution check.

ChangeLog

Version 2.16.3 - 2016-09-09 (EO)
--------------
	+ src/TimeFormatter.h: GNUC >= 6 since g++ 5.3.1 does not use C++ 2011
	  by default.

Version 2.16.2 - 2016-08-28 (EO)
	+ src/TimeFormatter.h: Added missing #if __GNUC__ >= 5 to use
	  constexpr,  otherwise keep const.

Version 2.16.1 - 2016-08-27 (EO)
--------------
	+ src/TimeFormatter.h: const double JDMIN -> constexpr double JDMIN

Version 2.16 - 2016-06-30 (AI)
--------------
	+ (src/param_impl.cc) gfortran array descriptor compatibility changes.
	+ tools directory (empty) removed from Makefile.

Version 2.15.3 - 2016-03-28 (EO)
	+ tools: Missing Makefile made clobber target fail. Added dummy
	  Makefile.

Version 2.15.2 - 2009-12-16 (EO)
--------------
	+ (SPR-6573 by H. Chen) Included a missing  between lines 578
	  and 579.

ChangeLog

Version 0.9 - 2016-02-09 (AI)
-----------	
	+ DISTRUBUTION changed to world

Version 0.8 - 2015-02-10 (AI)
-----------	
	+ New condition added to take into account attitudes without P, C or O
	identifiers.
	

ChangeLog

Version 4.8 - 2016-09-20 (AI)
-------------
	+ (src/rgsangles_mod.f90) Statement to check
	NaN values in angles2channels subroutine improved.

Version 4.7.8 - 2014-09-29 (AI)
-------------
	+ Default parameter values for "withheliocentriccorr" and
        "withsunanglecorr" changed to "yes". SOC-SPR-7240 closed
	

ChangeLog

Version 2.19.1 - 2017-01-09 (AI)
--------------
	+ Doc updated to include the parameter "embadpixalgo". This parameter
	enables the call to "embadpixfind" task algorithm to filter
	hot pixels and columns.
	
Version 2.19 - 2015-11-18 (AI)
--------------
        + embadpixfind optional usage to filter hot pixel and columns
	+ New parameter to chose the badpixel filter algorithm (SOC-SPR-6862)
	

ChangeLog

Version 1.18.3 - 2016-10-31 (AI)
--------------
	+ ON_NODE_INTERFACE flag is automatically disabled in case RGS is
	working	in singlereadout mode. SOC-SPR-6757 closed
	
Version 1.18.2 - 2013-11-05 (AI)
--------------
	+ Call to ExposureMapper (rgslib) to calculate the exposure map taking into account the
	sun aspect angle and heliocentric correction, in case they were enabled in rgsangles.
	

ChangeLog

Version 3.18 - 2016-09-07 (AI)
--------------
	+ Initialization to null of all pointers that are passed
	to C++ (Dope vectors). If we do not initizialize the pointers,
	we get seg fault at run-time with gcc-6.2.	

Version 3.17.2 - 2007-09-13 (AI)
--------------
	+ Single Readout Node change: Keep the same array size for ccdnode and
	rawx variables when duplicate pixles are found.
	

ChangeLog

Version 0.1.1 - 2017-01-09 (AI)
-----------
	+ First stable version.
	+ DISTRIBUTION changed to "world"
	
Version 0.1 - 2015-11-18 (AI)
-----------
	+ First draft version

ChangeLog

Version 0.55 -  2016-09-16 (AI)
-------------
	+ (src/rgslccorr.cc) Output keyword SRCNUM now contains the correct
	source identifier value.

Version 0.54 -  2015-05-12 (AI)
-------------
	+ Brackground rate and error column added to the background corrected light curve.
	+ Obs id, exp id, ra_obj and dec_obj keywords added to the output
	file to make it compatible with elcplot.
	+ error when filtering with single CCDs fixed.
	

ChangeLog

Version 1.6 - 2016-09-07 (AI)
------------
	+ (src/rgslinepos_mod.f90) Initialization to null of all pointers
	that are passed to C++ (Dope vectors). If we do not initizialize
	the pointers, we get seg fault at run-time with gcc-6.2.

Version 1.5 - 2016-07-21 (AI)
------------
	+ (src/rgslinepos_mod.f90) gfortran compatibility changes applied.

Version 1.4 - 2009-05-08 (CG)
------------
	+ DEPEND updated - ssclib 4.8

ChangeLog

Version 1.18.2 - 2016-12-20 (RDS)
------------
	+ src/rgsoffsetcalc_mod.f90. Pointer to time array nullified.

Version 1.18.1 - 2004-06-10 (CG&AI)
------------
	+ test/testodfrgsoffsetcalc.cc. The way to create a
	  vector changed to prevent problems on MAC.

ChangeLog

Version 1.34.7 - 2017-01-09 (AI)
------------
	+ Doc updated to include the choice parameter "badpixalgo".
	This parameter enables the option "embadpixalgo" that
	calls to "embadpixfind" task algorithm to filter
	hot pixels and columns.

Version 1.34.6 - 2016-11-18 (AI)
------------
	+ Version number in Changelog fixed.

Version 1.35.5 - 2016-11-18 (AI)
------------
	+ Parameter "order" forwarded to rgsrmfgen call.

Version 1.34.4 - 2016-11-03 (AI)
------------
	+ New rgsrmfgen parameter (witheffectiveareacorrection) added.
	
Version 1.34.3 - 2016-01-13 (AI)
------------
	+ The test Boresight ccf file was corrupted because of a
	git upload problem. It has been replaced by a correct one.
	

ChangeLog

Version 1.15.6 - 2016-11-28 (AI)
-----------
	+ (src/calib.cc) Wrong initialization of correction vector removed.	

Version 1.15.5 - 2016-11-23 (AI)
-----------
	+ (src/calib.cc) response vector passed by reference.
	
Version 1.15.4 - 2016-11-21 (AI)
-----------
	+ (src/calib.cc) "static" statement removed from effareacorr
	variables.
	+ (src/calib.cc) Initialization of effarecorr variables to 0

Version 1.15.3 - 2016-11-03 (AI)
-----------		
	+ Implementation of the new effective area correction.
	+ New parameter (witheffectiveareacorrection) added.
	
Version 1.15.2 - 2016-01-29 (AI)
-----------		
	+ Doc updated.
	

ChangeLog

Version 2.8.2 - 2016-11-03 (AI)
------------
	+ DATAMAX quality check removed if exposure map is in lambda
	space.
	+ Documentation updated.

Version 2.8.1 - 2014-09-10 (AI)
------------
	+ Bug using rebin parameter in lambda space fixed. SOC-SPR-6923 closed
	

ChangeLog

Version 2.5.1 - 2016-09-09 (EO)
-------------
	+ src/DummyResponse.h: GNUC >=6 required since g++ 5.3.1 does not
	  use C++ 2011 by default.

Version 2.5 - 2016-09-07 (RDS)
------------
	+ (src/PiledUpResponseCorrector.cc/h - changes to the pile-up
	                                       correction method.
					    . Only use events within region
					    . Use events, not charges for
					      normalisation and Include
					      one-photon frames
					    . Weight events with no. of good
					      events in frame before creating RMF

Version 2.4 - 2016-09-06 (RDS)
------------
        + (src/DummyResponse.h)             - gcc-6.X portability changes.
	                                              (constexpr added to "static
						      const" statements)

Version 2.3 - 2016-07-15 (RDS)
------------
	+ (config/rmfgen.par)             - Add parameter acceptchanrange
	+ (src/Rmfgen.cc)		    Exit with error if spectrum has
					    non-standard channel range
						(SPR-7363)

Version 2.2.1 - 2014-09-17 (RDS)
------------
        + (config/rmfgen.par)             - Change the parameter name from
                                            applymirroradjustment to
                                            applyxcaladjustment
					    Add parameter withdetbins
						(SPR-7219)

ChangeLog

Version 1.580.1 - 2017-01-12 (EO)
---------------
	+ configure_install: Fix perlsed when SAS_PERL is selected.

Version 1.580 - 2017-01-10 (EO)
-------------
	+ configure_install: Function identify_perl enhanced to handle
	  hardcoded "/usr/local/bin/perl" she-bangs, perl scripts with
	  non-printable hidden characters and additional blank space between
	  she-bang "#!" and the perl binary location.

Version 1.579 - 2016-12-16 (EO)
-------------
	+ configure_install:
	  - Fixed identify_perl and change_perl functions
	    to improve detection and change of the perl shebang.
	  - Better identification of errors.

Version 1.578 - 2016-11-17 (EO)
-------------
	+ configure.in:
	  - Removed "-Wc,-fno-common" link option to be passed to
	  GNU g++ linker on Mac OS X. This is not used in Linux and prevented the
	  use of named common blocks from different objects that will form part of
	  a shared library.

Version 1.577 - 2016-10-28 (EO)
-------------
	+ configure.in: Added checking on whether GFORTRAN version is greater
	  or equal to 5 or not. If not issue WARNING notice on some Fortran
	  2003 standards will not be supported, e.g. IEEE_ARITHMETIC module
	  used in omcat.

Version 1.576 - 2016-10-13 (EO)
-------------
	+ configure.in: Reinstated missing F90VERSION variable.
	+ sas.environment.in: Replaced GFORTRAN_VERSION by GFORTRANVERSION as
	  this is now the new name of the old variable.
	+ packagedetails.html.in: Replaced GFORTRAN_VERSION by GFORTRANVERSION
	  as this is now the new name of the old variable.

Version 1.575 - 2016-10-11 (EO)
-------------
	+ configure.in: Fixed definition of Makefile variable F90.

Version 1.574 - 2016-10-10 (EO)
-------------
	+ configure.in:
	  - No Fortran compiler by default. Both USENAGFORTRAN and USEGFORTRAN
	    are set to "no" by default. Therefore, we must use either
	    --with-nagfor   -> NAG Fortran compiler
	    --with-gfortran -> GNU Fortran compiler
	  - f95 is not tested anymore to avoid confusion of old f95 from NAG
	    with soft link to gfortran.
	  - NAG Fortran compiler is searched as nagfor.
	  - NAG Fortran compiler license availability is tested when the NAG
	    Fortran compiler is set to use.
	  - Removed -Wall for gfortran.
	+ test/tesconfiguremake: The configure now runs with the same options
	  used in the reference build.
	+ AKA: 16.0.0 => Start of release mode for SAS 16.0.0.
	

Version 1.573 - 2016-08-26 (EO)
-------------
	+ configure.in: Given the consequences on source code syntax
	  produced by using -fpp (NAG) or -cpp (GNU) preprocessor switches
	  on all Fortran compilations, we have decided to remove them from
	  SASF90FLAGS and SASFFLAGS.
	  Just for reference of what we have found so far here are two cases
	  which we have observed:
	  - Blank lines can not be used between continuation lines (ended
	  by &).
	  - Standard Fortran comment lines (started with !) can not be ended
	  with character "\" because this is the preprocessor character used
	  to continue preprocessor lines (those started with #).
	  Therefore, Fortran preprocessor directives MUST be processed by
	  means of code inlcuded ONLY in the individual source Makefiles
	  (src/Makefile).
	
	  A good example is the following, that is used to allow to process
	  the use of the "f90_unix_proc" module only by the NAG Fortran compiler:
	
	  ifeq ($(USENAGFORTRAN),yes)
	  	F90FLAGS += -fpp -DNAGf90Fortran
	  else
	  	F90FLAGS += -cpp -DGFORTRAN
	  endif
	
	  This code must be placed after the statement

	  include $(SAS_DIR)Make.include


Version 1.572 - 2016-08-25 (EO)
	+ configure.in: Changes to this file introduced in previous version
	  were not made effective due to forgeting to run the update-configure
	  script.

Version 1.571 - 2016-08-24 (EO)
-------------
	+ configure.in: Added preprocessor switches to both NAG and GNU
	  Fortran compiler compilation flags, SASF90FLAGS, to allow for
	  processing conditional compilation preprocessor directives
	  embeded in the code, as for example
	      #ifdef NAGf90Fortran
	      use f90_unix_proc
	      #endif
	  which will load NAG supplied module f90_unix_proc only when NAG
	  Fortran compiler is used (when the macro NAGf90Fortran is included
	  in F90FLAGS).
	  The preprocessor switch is also added to SASFFLAGS.
	  Given that NAG Fortran compiler is still the default Fortran
	  compiler, the default compilation flags SASF90FLAGS and SASFFLAGS
	  include now the macro -DNAGf90Fortran.
	  This change makes innecessary any possible changes included in local
	  Makefile.

Version 1.570 - 2016-08-23 (EO)
-------------
	+ configure.in: Added -DGFORTRAN to SASFFLAGS to avoid missing this
	  macro when building .f Fortran programs, e.g. saltec.
	  Aligned the SASF90FLAGS and SASFFLAGS used for Mac OS X with these
	  for Linux. This operation implied to add -Wall, -fno-range-check and
	  -DGFORTRAN to SASF90FLAGS and -Wall, -fno-range-check and -DGFORTRAN
	  to SASFFLAGS.

Version 1.569 - 2016-08-23 (EO)
-------------
	+ bin/f90depend: Looking for Fortran compiler dependencies, exclude
	  any NAG Fortran supplied modules (f90_*) when NAG Fortran compiler
	  is in use. This situation occurs when the NAG Fortran "lib"
	  subdirectory (which is part of the NAG Fortran installation) is
	  included in the list of include directories passed to the script as
	  first argument ($(F90INCFLAGS)). This variable is defined in the SAS Makefile
	  and includes the NAGWare "lib" subdir only when USENAGFORTRAN is set to "yes".
	  Script documentation updated.

Version 1.568 - 2016-07-12 (RDS)
-------------
	+ bin/sastaskhdr_nag, bin/sastaskhdr_gfortran:	
	  - Split the sastaskhdr script into two parts so that the
	    module names are set as:
	    NAGf90Fortran: $1_mod_MP_$1
	    gFortran: __$1_mod_MOD_$1

	+ Make.include.in:
	    uses one of the two sastaskhdr scripts depending on the compiler
	    in use

Version 1.567 - 2016-06-29 (EO)
-------------
	+ configure.in:
	  - Added -DGFORTRAN to SASCXXFLAGS when -with-gfortran is used (pending change
	    from previous changes to configure.in).
	  - Added -fno-range-check to SASFFLAGS to support gfortran usage.

Version 1.566 - 2016-04-06 (EO)
-------------
	+ configure.in:
	  - Added GFORTRAN macro to CXXFLAGS and SASCXXFLAGS
	  to include specific defintions when the GNU Fortran compiler
	  is used (--with-gfortran).
	  - Cleaning up of obsolete code.
	  - Removed g77 and libf2c.
	  - For Mac OS X : Added -bind_at_load option to build shared
	  libraries.
	+ configure_install: Added missing "\" before $platform in setsas.csh.

Version 1.565 - 2016-02-09 (EO)
-------------
	+ AKA = 16.0.0-devel
	+ configure.in: 16.0.0-devel

Version 1.564 - 2015-12-11 (EO)
-------------
	+ configure.in: Cleaned from obsolete code.
	+ Make.include.in: Changed sed field separators from / to ? to fix the
	  problem of replacing -sasperl- by $$PT, given that $$PT
	  is a path and includes / as directory separators.

ChangeLog

Version 1.87.1 - 2016-11-30 (RDS)
------------
	+ (lib/saslib/md5global.h)  Define UINT4 properly for 64-bit machines
					(SPR-7410)

Version 1.87 - 2012-01-17 (AI)
------------
	+ (lib/saslib/md5global.h)  Fix the MD5 error in 64-bit platforms.
	 In 64-bit, longs are 64-bit, not 32-bit. The definition of UINT4
	has to change depending on the platform.

ChangeLog

Version 4.74.1 - 2016-11-30 (AI)
--------------
	+ (src/selcore/NumericFunctors.cc) Removed an isnan extern macro that was giving an
	error in MAC OSX (gcc 6.2) compilation.
	
Version 4.74 - 2016-30-08 (AI)
--------------
	+ (src/selcore/Evaluator.cc) Conditional expression in input attibute of
	lexinit method call removed (gcc-6.1 throws a run time error) and
	now the contional expression is evaluated outside the method call.
	
Version 4.73 - 2015-13-03 (AI)
--------------
	+ src/selcore/VectorFunctors.cc: The inner product of the median spacecraft attitude and the
	current spacecraft attitude vectors can end up with a value greater than 1.0 (error propagation).
	If this is the case, the ConeFilterLogic is wrong and the attitude is discared from the GTI returned by attfilter.
	The result of the inner product is checked and if it is greater than one, the value is reseted to 1.0.
	

ChangeLog

Version 1.19.1 - 2016-12-07 (AI)
------------
	+ (src/srcdisplay.cc) Temporary string object is created to hold the
	const char pointer.

Version 1.19 - 2014-01-22 (RDS)
------------
        + src/srcdisplay.cc		- added option to use the positional
	+ src/SourceList.cc/h		- error from the source list to
	+ Documentation			- determine the source radius
					  (SPR-7170)

ChangeLog

Version 3.20 - 2016-09-30 (IT)
------------
	+ srcmatch_io.f90: handle eboxdetect source lists
	+ srcmatch_mod.f90: no html file will be written if parameter
	  htmloutput is empty (used by edetect_stack)
	+ srcmatch_mod.f90: negative values of maxerr are interpreted as
	  absolute offsets in arcsec. This feature is meant to stay
	  experimental and therefore not fully included in the documentation.
	+ srcmatch_def.f90, srcmatch_io.f90: increased internal parameter
	  max_nsets from 9 to 60 for use within edetect_stack and adjusted
	  column names (intended for usage within edetect_stack only)
	+ srcmatch_io.f90: informational output of the number of sources
	  in the final list at higher verbosity levels.
	+ cf. SCR 7399: source detection on stacked observations


Version 3.19 - 2016-07-14 (AI)
------------
	+ (src/srcmatch_mod.f90) gfortran compatibility changes.
	+ (src/Makefile) Preprocessor statement added
	
Version 3.18.1 - 2015-08-19 (IT)
------------
	+ srcmatch_mod.f90, subroutine srcmatch: increased maximum length
	  of output file names to 1024 (consistent with other
	  source-detection tasks).

ChangeLog

Version 4.32 - 2016-07-21 (AI)
------------
	+ (src/ssc_misc.f90) gfortran compatibility changes applied (string to integer conversion.
	+ (test/testregridders1.f90) "oldimage" array initialization
	(error in gfortran)
	
Version 4.31 - 2016-07-21 (AI)
------------
	+ (test/testarrayutils.f90) gfortran compatibility changes applied.
	
Version 4.30 - 2015-07-28 (RDS)
------------
	+ (src/wcs_aux.f90) Added routine, readWcsSkyKeys, to read the
		WCS structure from the REFxxx keywords.
		Also removed hard error when CUNITs not set and
		transform is not RA---TAN
		(SCR-7300)

ChangeLog

Version 1.95 - 2016-04-30 (AI)
------------
	+ (src/main.cc task_c.cc) gfortan compatibility changes
	
Version 1.94 - 2016-03-18 (EO)
------------
	+ DEPEND: Added dependency on sasperl to fix a failure to find
	  SAS::Error perl module while running the test runperltask.
	  This dependency appeared when building a reduced set of packages
	  where sasperl package is not included.

Version 1.93 - 2015-12-04 (EO)
------------
	+ src/main.pl: The sha-bang for perl (/usr/local/bin/perl) is now the variable
	  -sasperl-, which, like other variables in this file which are
	  eclosed by "-", it is replaced at build time by the proper value.
	   -sasperl- is replaced by the PERL determined by the SAS configure.
	   PERL can be either equal to a SAS_PERL value defined by the user or
	   determined among several possible perl installations:
	   /usr/bin/perl, /usr/local/bin/perl, /sw/bin/perl and /opt/local/bin/perl,
	   the last two only valid for Darwin.

ChangeLog

Version 1.122.1 - 2016-10-18 (RDS)
---------------
	+  constituents/ccf.cif: Set reference to EMOS1/2/PN_FWC_0001.CCF

Version 1.122 - 2015-12-10 (RDS)
---------------
	+  constituents/XMM_BORESIGHT_0015_SCISIM.CCF: Updated 0015 to ALGOID=1
	+  constituents/ccf.cif: Set reference to XMM_BORESIGHT_0015_SCISIM.CCF
	+  constituents/scisimccf.cif: Set reference to " "

ChangeLog

Version 1.47 - 2016-09-05 (JVP)
----------------------------
- Created implotregions_data/ with  m1_image_full.ftz,
src1_srcregfile.ftz and src1_bkgregfile.ftz

Version 1.46 - 2014-04-08 (EC)
----------------------------
- Created badpixfind_data with badpixfind/test/*.NEW

ChangeLog

Version 1.136.3 - 2016-10-27 (EO)
---------------
	+ leapsecdata/leapsecs.dat: Added 1 sec to TAI-UTC:
	  TAI-UTC = 37s (From 2017-01-01 0h UTC).
	  IERS (Int. Earth Rotation and Ref. Syst. Service ) Bulletin C 52,
	  Observatoire de Paris, 6 July 2016
	  (services.iers@obspm.fr/http://hpiers.obspm.fr/eop-pc).

Version 1.136.2 - 2016-09-09 (EO)
---------------
	+ src/STime.h: Builds on venus(IFCA, Santander) made with GNU C/C++
	  5.3.1 demonstrated that even with this version, the default C++ 2011
	  is not in effect. As for 4.8.5, 5.3.1 requires option -std=c++11 to
	  understand C++ 2011. Therefore, GNUC variable must be >= 6 to apply
	  constexpr.

Version 1.136.1 - 2016-08-23 (EO)
---------------
	+ src/STime.h: Amended to allow also for GNU C++ version greater or
	  equal to 5 (__GNU__ >= 5).

Version 1.136 - 2016-08-22 (EO)
-------------
	+ src/STime.h: In C++ 2011, the float and double are not
	of integral or enumeration type, therefore the initializer of a class must use
	constexpr instead of const. C++11 is the default (-std=c++11 or
	-std=gnu++11) since GNU C/C++ version 5. Thus, to keep code compatible
	with earlier versions of GNU C++, the constexpr is selected only if
	__GNUC__ macro is greater than 5.

Version 1.135 - 2016-07-21 (AI)
-------------
	+ Roll-back to major number 1.XXX. Code same as version 2.0.

Version 2.0 - 2016-07-21 (AI)
-------------
	+ (test/testf90cxx.cc) gfortran compatibility regarding array
	descriptors applied.
	
Version 1.134 - 2016-06-30 (AI)
-------------
	+ (src/f90cxx.h): Dependecy to libgfortran.h removed. Define GFC_DTYPE_TYPE_SHIFT
	statement copied from libgfortran.h to f90cxx.h
	
Version 1.133 - 2016-06-29 (AI)
-------------
	+ Version 1.133 identical to 1.2

Version 1.2 - 2016-06-29 (AI)
-------------
	+ src/f90cxx.h: New Dope (Array descriptor) definitions compatible
	with gfortran compiler.
	
Version 1.132 - 2016-03-22 (EO)
-------------
	+ src/cfortran.h: Until now we have used cfortran.h version 4.1
	  1990-1998, from Burkhard Burow (burow@desy.de), to deal with
	  the interface between C (C++) and Fortran. That version supported
	  NAG Fortran but was not prepared to work with GNU Fortran.
	  We have looked for an updated version of this file that could be
	  able to work with GNU Fortran. That updated version is version 4.4,
	  which is an improved version of the original version 4.1 but
	  has been maintained by HEASARC Kevin McCarty et al, at least until 2008.
	+ src/f90cxx.h: Removed obsolete code that was enabled only if SUNCC
	  and SUNF90 macros were defined previously. These definitions were
	  used to support using Sun C/C++ and Sun F90 Fortran compilers.
	  The definition of NAGF90 and GNUCC, within the file itself was absurd.
	The only solution to switch on/off specific pieces of code is to
	  do it through the configure.
	  After checking in the SAS configure, the only macro defined when NAG Fortran
	  is used is NAGf90Fortran, also used to select specific code in cfortran.h.
	  There is no similar counterpart for GNU Fortran.
	  To be able to enable the definitions of f90_io_init and f90_io_finish, only
	  when NAG Fortran is used, we have set then within a block of preprocessor
	  #ifdef-#endif block. The remaining definitions are left enabled always,
	  either for NAG Fortran or GNU Fortran.
	  Further testing will tell us how we have to modify this and possibly other
	  files in this package (and possibly others), to be able to work with both
	  compilers
	+ test/testf90cxx.cc: f90init and f90finish enabled only if NAGf90Fortran
	  macro defined.

Version 1.131 - 2015-11-23 (AI)
---------------	
	+ Refractoring and renaming of ranlux.f F77 code to rluxgo.f  to solve a weird
	problem in Mac OS X 10.11 (Yosemite) that was unable to generate random numbers
	producing a segmentation fault.
	

ChangeLog

Version 1.4.4 - 2016-11-17 (EO)
-------------
	+ External perl modules updated as follows:
	  - Log-Log4perl-1.31         -> 1.48
	  - XML-NamespaceSupport-1.11 -> Unchanged
	  - XML-SAX-Base-1.08         -> New module.Required by XML-SAX-0.99.
	  - XML-SAX-0.96              -> 0.99
	  - XML-Simple-2.18           -> 2.22
	  - XML-LibXML-1.70           -> 2.0128

Version 1.4.3 - 2016-08-18 (AI)
------------		
	+ (src/odfParamCreator.cc) ECFs values for default energy bands added
	to XML parameter file.
	+ (src/run_edetectchain.pl) New ECFs interface to XML input file.
	+ (config/epn_confFile.xml, emos1_confFile.xml and emis2_confFile.xml)
	new ecf parameter added to edetect_chain task