Документ взят из кэша поисковой машины. Адрес оригинального документа : http://xmm.vilspa.esa.es/sas/5.4.1/doc/cal/ChangeLog Дата изменения: Mon Dec 9 17:46:01 2002 Дата индексирования: Sat Dec 22 07:05:44 2007 Кодировка: Поисковые слова: http astrokuban.info astrokuban

ChangeLog for Calibration Access Layer (CAL)
============================================

Open SOC-SPRs/SCRs:
*none*

Version 3.152.2 - 2002-12-09 (UL)
---------------
+ fixed bug in EpnRedist::spectrum() - caused incorrect mode information
to be passed to calpnalgo for TI/BU mode
(SPR 2381: bug in I/F to calpnalgo redist code)
+ minimum exposure duration now 300s; setting exposureDuration to
value less than that causes state warning
(SPR 2382: minimum exposure duration too large)

Version 3.152.1 - 2002-12-03 (UL)
---------------
+ fixed bug which caused reading of EMOS/EPN_QUANTUMEF CCF to terminate
in error for last science mode (FastCompressed/FastBurst) [SPR 2375]
+ state variable 'exposureDuration' can now be set/obtained from f90
[SPR 2376]

Version 3.152 - 2002-11-20 (UL)
-------------
+ closed SPR 2303: CAL_psfValidityRanges with ACCURACY_HIGH
+ fixed problem with CAL_pnAdditionalGainCorrect() in C++/f90 layer
+ fixed problem with missing virtual declaration in EnergyCorrector.h
+ implemented ALGOID=1 of EPN long-term CTE model (time-dependence modelled
as 3. order polynomial; all parameters mode dependent); goes with
structural change in EPN_CTI CCF; all changes fully backward compatible
+ implemented access to mode-dependent EMOS?_REDIST CCF with new MODE_ID
column; changes backward compatible

Version 3.151 - 2002-08-23 (UL)
-------------
+ frame time parameter of ModeParam atom now obtained from TIMECORR CCF
+ additional parameter passed to CTE code in calpnalgo

Version 3.150 - 2002-08-19 (UL)
-------------
+ more transparent fixes to enhance C++ standard compliance

Version 3.149 - 2002-08-16 (UL)
-------------
+ added various f90 routines to retrieve pattern fraction data from CCF
+ fixed wrong name of CAL_psfEnboxedEnergy() in c++/f90 I/F layer
+ gcc-2.96/3.1 portability fixes

Version 3.148 - 2002-08-02 (UL)
-------------
+ fixed problem in CAL_releaseMemory I/F (3D vectors could not be released)

Version 3.147 - 2002-07-31 (UL)
-------------
+ implemented ALGOID=2 MOS CTI correction

Version 3.146 - 2002-07-26 (UL)
-------------
+ read additional gain/CTI parameters from CCF and pass to CTI/gain
correction codes in calpnalgo>=2.39 [requires EPN_CTI_>=0009.CCF]
+ implemented new routines
CAL_pnAdditionalGainCtiCorrect
CAL_pnGainTimingCorrect
CAL_pnGainBurstCorrect
these replace corresponding ones in calpnalgo which become obsolete;
needed parameters are likewise read from EPN_CTI_>=0009.CCF; updated
doc w.r.t to new routines
+ closed SPR 2246: Useof CCDID from EMOSn_QUANTUMEFF in calview
[QE Calviewable was not sensitive to theta/phi values in state]

Version 3.145 - 2002-07-22 (JBa)
-------------
+ changes in source files in order to comply with ANSI C++

Version 3.144.1 - 2002-07-01 (UL)
---------------
+ fixed minor problem in one test harness related to 3.144 changes

Version 3.144 - 2002-06-28 (UL)
-------------
+ replaced all occurences of 'strstream' with standard-conformant
'stringstream'

Version 3.143 - 2002-06-26 (UL)
-------------
+ bug fix in misalignmentFigureDistribution() related to recent addition
of ICO argument
+ CAMCOORD1->CHIPCOORD conversion will now yield CCD with higest elevation
in focal plane (only relevant for EMOS for overlapping chip areas)
(closes SPR 2326)

Version 3.142 - 2002-06-14 (UL+JDR)
-------------
+ changes to alignment of first moment in dispersion figure distribution
(relevant for rgsrmfgen only)
+ added new BackgroundDataServer atom with support for RGS HTR mode
+ added inChannelOffset argument to
GratingDataServer::misalignmentFigureDistribution()

Version 3.141 - 2002-06-12 (UL)
-------------
+ fixed bug in calculation of errors of converted OM count rates
+ fixed minor problem caused by not capturing invalid atan2() arguments
in CAMCOORD2->TELCOORD conversion
+ added CAL_getMiscallaneousDataCcd()

Version 3.140 - 2002-05-24 (GV)
-------------
+ Merge changes made on the RT at 3.138.{3,4}
+ (Ccf.cc) remove some print statements that had been left in the
constructor of CCfAccessLog. Modify operator<<() for
CcfAccessLog, so that it also re-issues warnings about expired
CCF constituents. This operator is used in taskmain before the
task is exited, so these warnings are again prominently shown. Is
this an example of a 'critical' warning? ;-)

+ (Ccf.{cc,h}) Change the class CcfAccessLog so that it knows
about possible expired CCF constituents. In Ccf.cc rename the
FileLocator fileLocator to myfileLocator to avoid a name clash
with extern fileLocator (from SasFileLocator.h). + Add the
configuration file ccf.conf to the config subdirectory. + Write
test harness for the above: testccfconf

Version 3.139 - 2002-04-18 (UL)
-------------
+ *-vs-theta calview plots go out to theta=1200" now
+ brought in line with utils-1.88
+ some code aesthetics

Version 3.138 - 2002-03-14 (UL)
-------------
+ empirical correction to RGS effective area (driven by table
RGS?_QUANTUMEF_????.CCF:RGA_EFFAREACORR) now done through interpolation
of correction data

Version 3.137 - 2002-03-06 (UL)
-------------
+ added CAL_toDirectionCosineMatrix()/CAL_toEulerAngles()
(meant to replace CAL_getBoresightMatrix()/CAL_euler321toMatrix()
which will be phased out of the API at some point)
+ extended and revised documentation of boresight related routines
+ adapted f90 test harnesses w.r.t above additions

Version 3.136.1 - 2002-02-19 (UL)
---------------
+ relaxed interval for allowed chip/camera temperatures in CAL state
to now [0, 300] K for all instruments (was only EMOS in 3.132)
+ updated error labels in doc

Version 3.136 - 2002-02-15 (UL)
-------------
+ CAMCOORD2<->TELCOORD conversions now properly take values of
OPTICS_X/OPTICS_Y in XMM_MISCDATA CCF into account
+ more precise definition of TELCOORD/CAMCOORD2 in doc

Version 3.135.8 - 2002-02-15 (UL)
---------------
+ bug fix in CrossPsf construction - distributions were wrongly
centered
+ changes to CrossPsf Viewable (now show 90% contour + first moment
of distribution vs beta)
+ minor bug fix in EMOS redistribution code which prevented pattern-0-only
spectra from being accepted

Version 3.135.7 - 2002-02-12 (UL)
---------------
+ return values of new CAL_getEffectiveArea() are again close to
pre-3.135.6 values (used to be significantly different for large
off-axis angles); now depends on caloalutils-1.64

Version 3.135.6 - 2002-02-08 (UL)
---------------
+ bug fix in EmosRedist::at()
+ significantly improved performance of CAL_getEffectiveArea()

Version 3.135.5 - 2002-02-05 (UL)
---------------
+ fixed minor compilation problem under Linux

Version 3.135.4 - 2002-01-30 (UL)
---------------
+ RgsCcdQuantumDataServer: added support for virtual H2O layer on RGS CCD
to model observed absorption feature around O-edge in QE cuves
+ brought in line with testccf-1.96

Version 3.135.3 - 2002-01-29 (UL)
---------------
+ fixed minor problem in RgsCcdQuantumDataServer inadvertently
introduced during code reorganization in cal-3.134
+ fixed problem in GratingDataServer inadvertently introduced
with cal-1.135.1 (caused failure in case of old-style RGS QUANTUMEF
CCF w/o reflection efficiency correction data)
+ fixed coding bug in PsfDataServer helper class (caused crash under
Linux)

Version 3.135.2 - 2002-01-28 (UL)
---------------
+ fixed minor problem with caching of PsfDataServer atom in case accuracy
level was changed autonomously and CAL is used in metatask context
+ closed SPR 2261: problem with error propagation in CAL_omPhotoNatural

Version 3.135.1 - 2002-01-25 (UL)
---------------
+ new algorithm to interpret CCF data RGS?_QUANTUMEF.CCF:RGA_EFFCORR
(fudge data are interpolated before being applied to reflection
efficiencies curves)

Version 3.135 - 2002-01-18 (UL)
---------------
+ downgraded CTI mode-mismatch warning to message (Verbose)
+ extended list of possible cmode values passed to calpnalgo
+ harmonized CcdQuantumDataServer/Redistribution classes w.r.t
event grade selections

Version 3.134 - 2002-01-15 (UL)
-------------
+ new class PatternFractionDataServer which gives access to new
tables in CCF::QuantumEf
+ extensive internal re-organization of CcdQuantumDataServer code
in conjunction with addition of new PatternFractionDataServer class
+ cal{mos|pn}nalgo I/F: added pattern and patternFraction argument
+ failure to find EXTENDED-accuracy PSF in CCF will change accuracy
level to HIGH also in the CAL state
+ downgraded couple of CAL warnings to messages (Msg::Noisy)
+ temporarily disabled
o cxxtest/emosredist
o f90test/calexample_emos[redist]
until updated QuantumEf CCF constituent (with pattern fractions)
is in testccf
+ check that bad pixel locations read from CCF are within valid bounds -
otherwise end in error (used to give assertion failure)

Version 3.133 - 2001-12-21 (UL) - the Christmas 2001 CAL!
-------------
+ fall back upon HIGH-accuracy PSF model if CCF table for EXTENDED
model is not found (used to fall back to MEDIUM)
+ fixed problem with size of last pixel row in case RGS OCB 3x3

Version 3.132 - 2001-12-20 (UL)
-------------
+ fixed another problem with the pixel size of the outermost RGS pixel
columns for OCB>1; work now correctly for all OCBs (tested 1x1, 2x2,
3x3, 4x4, 5x5)
+ added new CalViewable which plots outer perimeter of all CCD pixels in
CAMCOORD1 space (can be used to check RGS pixel sizes for OCB>1)
+ relaxed interval for allowed chip/camera temperatures in CAL state
to now [0, 300] K
+ in new test harness rgspixel: d is now .5 as it ought to be

Version 3.131 - 2001-12-18 (UL)
-------------
+ internal meliorations in CalDataVector class (implemented ref counting);
fixes a problem with instantiation of XmmPixelList from real-valued
coordinates; overhaul of C++/F90 glue in view of changes; added
test harness to test correct RGS pixel size in OCB 3x3
+ removed re-normalization of Lorentzian CrossPsf component

Version 3.130 - 2001-12-13 (UL)
-------------
+ changes to CrossPsf normalization; resulting probability spectrum
is no longer re-normalized to given cross dispersion window to
properly account for light loss outside the window; also arbitrary
combination of Gaussian+Lorentzians is now correctly handled
+ bug fix to CalDataVector::operator=
+ fixed minor problem with peaking EE-curve in calview

Version 3.129 - 2001-12-05 (UL)
-------------
+ altered reading of *CTI*::CTI-EXTENDED to make it compliant with
emos-cti-0.7 changes (CTI_X/CTI_Y are now vector columns of length 4)
related to new ALGOID-1 EMOS CTI correction
+ added CAL_getBadPixelCode()/CAL_getAduconvCode() for access to
attributes *BADPIX*CCF:BADPIX%BPT_CODE/*ADUCONV*CCF:OFFSET_GAIN%BPT_CODE;
updated test harnesses + doc accordingly

Version 3.128 - 2001-11-26 (UL)
-------------
+ new King-model PSF (activated on accuracy level EXTENDED) fully
fully implemented (PSF map + EE + inverse EE)
+ EE(PSF) viewable now plots EE vs r [arcmin]
+ fixed bug in CalDataVector.h (caused crash in HTR offset correction)
+ mode recognition in setting state: make use of SUBMODE attribute if
present
+ no exception handling in CalView backend to allow GUI to catch
and present errors

Version 3.127 - 2001-11-21 (UL)
-------------
+ RgsEnergyCorrector:
- bug fix in HTR offset correction: was testing against wrong mode
- added HTR CTI I/F
+ more changes related to CCD QE construction with EXAFS data (RGS only)
+ bug fix in GeometryDataServer: lead to wrong sizes of outermost pixels
columns of RGS in OCB 3x3/5x5

Version 3.126 - 2001-11-20 (UL)
-------------
+ fixed bug with instantiation of new EXTENDED-level PsfDataServer
+ several bug fixes related to new EXAFS scheme affecting
RgsCcdQuantumDataServer

Version 3.125 - 2001-11-06 (UL)
-------------
+ added virtual HTR-specific offset correction method to EnergyCorrector
+ new experimental Psf model (from CNR; analytical King model) selected
with new accuracy level 'EXTENDED'
+ fixed couple of problems with applying empirical correction curves to
various RGS entities (see 3.123 changes)

Version 3.124 - 2001-10-30 (UL)
-------------
+ added dedicated RgsEnergyCorrector::offsetCorrect() for HTR mode
+ fixed problem of SPR 2214: SCR74: list/document errors
+ fixed problem of SPR 2220: raw rate input parameter description for
om_phottonatural

Version 3.123 - 2001-10-29 (UL)
-------------
+ brought in line with CAL HB 2.2 changes
o construction of RGS QE curves amended by fine structure in
RGS?_EXAFS CCF
o self-vignetting + reflection efficiencies constructed with
empirical correction curves (driven by tables in RGS?_QUANTUMEF)
o multiplicative empirical correction curve to overall RGS effective
area (driven by table in RGS?_QUANTUMEF
all changes are fully backwards compatible; missing CCF data will
give pre-3.123 results
+ mode determination logic:
o in line with caloalutils-1.57 changes; now more reliable for
engineering modes
o can identify mode from 'SUBMODE' attribute
+ added 'std::' prefix to all standard C++ lib constructs in header files
and 'using namespace std;' in all .cc

Version 3.122 - 2001-10-18 (UL)
-------------
+ fixed problem of SPR 2215: coding error in CAL_omPhotoMagnitude
+ requesting filter transmission outside FOV will now yield
warning + transmission=0 instead of error

Version 3.121 - 2001-10-17 (UL)
-------------
+ CAMCOORD1/2/SACCCORD/ROWCOORD coordinates now in double precision to
resolve accuracy problem with sub-pixel positions; affects C++/F90 APIs
+ implemented CHIPCOORD->PIXCOORD1 conversion also for non-binodal
readout modes
+ added to cal.f90 routine to release int8 1-D array

Version 3.120 - 2001-09-27 (UL)
-------------
+ bug fix: if a CCF constituent was specified with path component (in CIF
or as replacement on command line) it was nevertheless searched along
SAS_CCFPATH
+ modiciations to new experimental MOS CTI correction (ALGOID=1)

Version 3.119.1 - 2001-09-26 (UL)
---------------
+ another harmless optimizer related change

Version 3.119 - 2001-09-25 (UL)
-------------
+ fixed SPR 1996: different OCB factors in X/Y for RGS HTR mode not taken\
into account
+ couple of trivial changes to make optimizer happy
+ new experimental MOS CTI correction (ALGOID=1)

Version 3.118 - 2001-09-14 (UL)
-------------
+ hot/discarded pixel data for EPN CTI correction are now read from
CCF::EPN_BADPIX instead of dedicated tables in CCF::EPN_CTI

Version 3.117 - 2001-08-15 (UL)
-------------
+ in RgsEnergyCorrector::shapeToGrade: turned error into warning
+ reading of selectlib-regions from CCF (LinCoord, FilterTrans,
QuantumEf) will now transparently replace "polygon" by "polygon2"
filters to include points on the boundary (see selectlib doc for
more details on new polygon2 filter);
requires selectlib-4.44 + utils-1.76

Version 3.116 - 2001-07-27 (UL)
-------------
+ fixed minor doc problem
+ fixed problem of SPR 2140: mode dependence of bad pixels not taken into
account

Version 3.115 - 2001-07-18 (UL)
-------------
+ brought in line with caloalutils-1.54 (one of the tests depends
on value of Xmm::missionStartTime which has changed 1.54)
+ also allow FILTER="UNDEFINED" in definition of state

Version 3.114 - 2001-06-01 (UL)
-------------
+ implemented SPR 2054: CAL_rowCoordToSacCoord produces wrong result
+ implemented SCR 61: Obtaining the integrated PSF function for OM
+ fixed couple of broken links in doc

Version 3.113.3 - 2001-05-15 (UL)
---------------
+ commented out diagnostic output statements in PsfDataServer

Version 3.113.2 - 2001-05-11 (UL)
---------------
+ fixed SPR 2027: Normalization of dispersed PSF distribution wrong

Version 3.113.1 - 2001-05-07 (UL)
---------------
+ updated I/F description of new
GratingDataServer::broadeningDistribution()
+ CHIPCOORD->CAMCOORD1 now also works for negative pixel numbers

Version 3.113 - 2001-05-04 (UL)
---------------
+ bug fix in assignment operator of Spectrum class
+ added GratingDataServer::broadeningDistribution()
+ fix of minor bug in GeometryDataServer::lsfDefocusDistribution()
+ access tables in CCF::OM_ASTROMET/CCF::LARGESCALESENS related to
magnifier filter as "FILTER-MAGNI" instead of "FILTER-MAGNIFIER"
(this fixes SOC-SPR 2002: wrong table name in OM_ASTROMET)

Version 3.112 - 2001-05-01 (UL)
-------------
+ brought in line with calmosalgo-2.6
(fixes SPR 2007: Update to EPIC-MOS redistribution code)
+ implemented RGS LSF defocus distribution for usage by rgsrmfgen
+ corresponding Calviewable

Version 3.111 - 2001-04-23 (UL)
-------------
+ changed return type of CAL_getMOSdarkFrameMap to real32 (was int16)
+ CIF is now always opened in low-mem mode
+ additional offset value in RGA defocus-length (Eq. 22, CAL-HB 2.1)
from RGS?_LINESPREADFUNC:FIGURE%DXOFFS
+ implemented SPR 1990: Extra parameter for observation mode in PN
redistribution code (=> needs calpnalgo-2.18)
+ bug fix in EpnRedist::spectrum(): caused erroneous message about
null-spectrum being returned
+ new state variable 'auxiliaryParameter'

Version 3.110 - 2001-04-10 (UL)
-------------
+ brought EPN test harness in line with calpnalgo-2.16

Version 3.109 - 2001-04-09 (UL)
-------------
+ fixed SPR 1984: pnCtiCorrect: wrong type of frame counter argument
+ MODE_N/A column in EMOS?_BADPIX can be of type fixed or variable length
vector

Version 3.108 - 2001-04-04 (UL)
-------------
+ fixed SPR 1973: MOSdarkFrameMap is incorrect
+ fixed minor documenation problem (CAL_omGetPSFmap)
+ removed forgotten debug statement

Version 3.107 - 2001-04-02 (UL)
-------------
+ brought GeometryDataServer::rowlandDefocusLength() in line with
CAL HB 2.1
+ brought RgsEnergyCorrector::offsetCorrect() in line with
CAL HB 2.1 (offset is time-dependent - triggered by ALGOID-1 in
RGS?_ADUCONV)
+ fixed SPR 1977: Wrong link to cal.f90
+ removed obsolete debug statements in OmPsfDataServer

Version 3.106 - 2001-03-29 (UL)
-------------
+ minor transparent change in GratingDataServer
+ bug fix in GeometryDataServer introduced with method onCcd() in
3.97.1 - caused problem in RGS (ALGO-1) CTI correction when
OCB!=1 && CCD!=1 (fixes SPR 1971: wrong results from RGS CTI correction)

Version 3.105 - 2001-03-23 (UL)
-------------
+ better error message if filter transmission is requested for point
outside FOV
+ moved BadPixel/BadPixelMap class to caloalutils
+ ->

Version 3.104 - 2001-03-15 (UL)
-------------
+ changed "error"s to "fatal"s in same cases to signal that execution
must really stop (the procs catch some errors and carry on
which can lead to bizarre problems further downstream ...)
+ implemented I/F to new EMOS redistribution code in calmosalgo
package (updated DEPEND); usage of this I/F is triggered by
ALGOID=1 in EMOS?_REDIST

Version 3.103 - 2001-03-12 (UL)
-------------
+ added version of GratingDataServer::scatteringDistribution()
which takes number of channels in output spectrum explictely
+ added 'fail-safe' mode to Redistribution object
+ fixed small problem in new f90/sky2pix test (causes crash under
True64)
+ fixed SPR 1949: Wrong verbosity level

Version 3.102 - 2001-03-09 (UL)
-------------
+ extended capabilities of Boresight class: can now project equatorial
directions onto TELCOORD frame (+f90 I/F)
+ fix of minor bug in GratingDataServer::scatteringDistribution()
+ EncircledEnergyDataServer now works for off-axis angles

Version 3.101 - 2001-02-26 (UL)
-------------
+ implemented workaround for linking problem under True64/gcc

Version 3.100 - 2001-02-20 (UL)
-------------
+ small improvement in mode exception vector reading of CCF::BadPixel
+ added GratingDataServer::bowingFigureDistribution()
+ added XmmCartesianPointList::rowlandDefocusLength()
+ closed SCR 37: update CAL

Version 3.99 - 2001-02-19 (UL)
------------
+ I/F roll-back in GratingDataServer::scatteringDistribution():
(betaM, alpha, E)->(alpha, order, E)
+ added forgotten dependency on caloalutils-1.40
+ implemented mode dependence of CCF::BadPix (fully backwards compatible)

Version 3.98 - 2001-02-15 (UL)
--------------
+ brought various OM quantities in line with updated descriptions
in issue 2.0 of CAL HB
* new color-color transformation algorithm with possibility to
specify different branches in Herzsprung-Russel diagram
* new count rate conversion (with optional empirical linearity
correction)
* CCF based OM PSF
* new function CAL_omGetApertureRadius
+ new Viewable showing OM PSF
+ GratingDataServer::scatteringDistribution():
* revised I/F: can now be directed to suppress small- or large angle
contribution; inputs betaM instead of order
* bug fixes (due to error in CAL HB)
+ scattering distribution viewable now plots small+large angle constituents
together with total spectrum
+ brought doc in line with above changes

Version 3.97.1 [for SAS 5.03] - 2001-02-01 (UL)
--------------
+ new CAL atom: EncircledEnergyDataServer which gives access to
encircled energy data from CCF::XEncirEn constituent
+ added Viewable associated with new atom
+ RgsRedist::spectrum(): return null-spectrum and give warning
message if input energy is outside energy axis bounds
+ added GeometryDataServer::onChip()/isOnAnyChip()

Version 3.97 - 2001-01-29 (UL)
------------
+ new CalViewable: RGS X-PSF vs beta
+ changed integration boundaries in
PsfDataServer::dispersionFigureDistribution() back to what they were
in cal<=3.91

Version 3.96 - 2001-01-25 (UL)
------------
+ fixed minor problem with CalViewable plot title introduced in 3.95
+ fixed SPR 1904: bug in RGS CTI correction (readout from node D was
treated incorrectly)

Version 3.95 - 2001-01-12 (UL)
------------
+ GratingDataServer::alpha() does now include rotation of RGA about
Y-axis
+ increased precision of alpha in GratingDataServer CalViewables

Version 3.94 - 2001-01-11 (UL)
------------
+ brought RGS redistribution model in line with latest update in CAL HB

Version 3.93 - 2001-01-10 (UL)
------------
+ CrossPsf Viewable does now take theta/phi into account to compute source
position in X-dispersion direction
+ added SelfVignetting-vs-beta plot as SubViewable of GratingDataServer
+ print grating incidence angle in all Grating Viewables plots

Version 3.92 - 2000-12-20 (UL) - the new Millenium CAL!
------------
+ bug fix in GratingDataServer::misalignmentFigureDistribution()
[spectrum was not aligned on first moment]
+ brought in line with testccf-1.90
+ introduced concept of ALGOID selector to prepare for support
of different calibration algorithms; this is fully backwards
compatible
+ pre-3.89 RGS CTI correction re-instantiated; old/new algorithm
is now selected with ALGOID=0/1 in CCF
+ PsfDataServer::dispersionFigureDistribution():
* removed constraint: 0<=inChannelOffset<1
* more precise I/F description
+ fixed SPR 1880: Tries accessing RGS cal files

Version 3.91 - 2000-12-04 (UL)
------------
+ bug fix in GratingDataServer::alpha(): bug caused effective area maximum
at phi=180 (it must be a phi=0)
+ PN gain correction: better message in case of calpnalgo failure

Version 3.90 - 2000-11-23 (UL)
------------
+ fixed problem in RGS test harness introduced with v3.89
+ added GratingDataServer::reflectionEfficiency(alpha, order)
+ absorbed releasetrack change 3.88.9->3.88.10
[fixed SPR 1834: Task aborts with Bus Error]
+ added GratingDataServer::misalignmentDistribution()
+ added GratingDataServer::scatteringDistribution()
+ added PsfDataServer::dispersionFigureDistribution()
+ PN gain correction: better message in case of calpnalgo failure
+ updated implementation of CrossPsf::probability() in line with
latest CAL HB changes; new implementation uses extended
CCF::CrossPsf but is backwards compatible, i.e., old CCFs can
still be read

Version 3.89 - 2000-11-10 (UL)
------------
+ trivial changes to eliminate warning messages from optimizer
+ new RGS CTI correction

Version 3.88.9 - 2000-11-09 (UL)
--------------
+ fixed SPR 1827: Does not recognize PrimeFullWindow for MOS 1

Version 3.88.8 - 2000-11-08 (UL)
--------------
+ RealCcf has a new friend ShowCcf
+ gain correction for EMOS in modes FastUncompressed/FastCompressed
for CCD 1 takes additional multiplicative factor (FAST_G in
CCF::AduConv) into account

Version 3.88.7 - 2000-11-07 (UL)
--------------
+ fixed SPR 1819: Does not identify PRI PART W6 MOS mode
+ minor change to description of ROWCOORD in the doc

Version 3.88.6 - 2000-11-01 (UL)
--------------
+ another minor change to the description of TELCOORD in the doc
+ CAL_omColorTransform() now provides a means to inquire the transformation
validity ranges
+ Ccf class has a new friend ShowCcf
+ added reading of absorption data in EPN::REDIST if present; otherwise
absorption lengths are calculated from Henke data

Version 3.88.5 - 2000-10-31 (UL)
--------------
+ better definition of TELCOORD frame in doc (Please note: The
definition as such has NOT been changed, just the wording to
describe it)
+ fixed problem in routine which computes the grating incidence angle
as a function of input TELCOORD-(theta, phi); as a result of this
fix the effective area maximum for EMOS1/2 is now correctly at
phi=180 deg (was at 0 deg)

Version 3.88.4 - 2000-10-27 (UL)
--------------
+ added config/cal.info

Version 3.88.3 - 2000-10-20 (UL)
--------------
+ fixed SPR 1667: MOS RFS mode: Outer CCDs are not in RFS

Version 3.88.2 - 2000-10-19 (UL)
--------------
+ bug fix in PsfDataServer::enboxedEnergy() - caused wrong (unnormalized)
integral value to be returned

Version 3.88.1 - 2000-10-17 (UL)
--------------
+ extended I/F of PsfDataServer()
+ supressed confusing warning message in TimeCorrector::correct()
+ minor modification in doc

Version 3.88 - 2000-10-06 (UL)
------------
+ re-activated `quantumef' test
+ use handler()

Version 3.87 - 2000-10-06 (GV)
------------
+ more of the same.

Version 3.86 - 2000-10-05 (GV)
------------
+ disable task quantumeff in cxxtest, as this requires taskmain.
+ setHandler no longer available, use handler instead (CalView.cc)

Version 3.85 - 2000-10-04 (UL)
------------
+ new time-dependent EMOS CTI correction algorithm; is backwards
compatible, i.e., pre-issue-5 CTI CCF constituents will give the
same result as with old algorithm
+ QE curves use again CubicHermiteInterpolant
+ construction of CalViewable list is now robust against missing
constituents
+ made one of the tests a proper task
+ fixed failing test in 3.84
+ internal changes related to CCF access

Version 3.84 - 2000-10-02 (UL)
------------
+ QE curves now implemented using LinearInterpolant instead of
CubicHermiteInterpolant
+ prototype of CcfAccessLog object (accessible through CalServer)
+ search for CCF constituents along SAS_CCFPATH
+ fixed SPR 1757: Wrong hierarchy in HTML doc
+ internal changes in documentation

Version 3.83 - 2000-09-29 (UL)
------------
+ internal change to bring in line with caloalutils-1.34
+ moved test 'xmm' to caloalutils
+ setting of filter in CalServer::setState() is optional
+ internal changes in documentation

Version 3.82 - 2000-09-26 (UL)
------------
+ bug fix in PsfDataServer
+ more precise definition of TELCOORD frame in doc
+ RGS CTI correction: now uses cubic-Hermite (instead of cubic-spline)
interpolation for the parallel coefficients
+ CAL_setState() now also tries to set the 'filterId' variable
from "FILTER" attribute in data set

Version 3.81 - 2000-09-22 (UL)
------------
+ PSF implementation now in line with CAL HB
(three accuracy levels implemented in terms of
single-gaussian/pre-computed PSF maps/triple-Gaussian)
+ added dedicated PSF test harness
+ updated documentation (new PSF routines)
+ replaced interim CAL_omGetPlateScale() in f90 layer
+ new CalServer::setState() methods
+ fixed problem with negative theta angles in effective-area related
classes
+ bug fix in several CalViewables (deg->rad conversion of phi)

Version 3.80 - 2000-08-29 (UL)
------------
+ fixed SPR 1736: Cannot get QE for chipy=1
for pixels outside the FOV a constant QE=0 curve is now returned instead
of bailing out with error
+ internal changes to MirrorDataServer enforced by structural changes
to CCF::XareaEf as of issue 6
+ fixed F/L bug in internal computation of RGA alpha-incidence angle from
mirror FOV angles (theta/phi)
+ fixed bug in GratingDataServer (caused segfault under certain
circumstances)
+ added EpnRedist() class for usage by EPN response generator;
implementation in terms of redistribution model code in package
calpnalgo
+ added two new CalViewables: RGA obscuration/intercept efficiency
vs. theta/phi
+ changed theta range in CalViewables to [0, 900"] to match instrument
FOV

Version 3.79 - 2000-08-03 (UL)
------------
+ implemented QE computation for RGS; updated test harness + corresponding
CalViewable; uses I/F designed for EMOS

Version 3.78 - 2000-08-01 (UL)
------------
+ added DarkFrameDataServer atom + f90 routines to access EMOS
offset vectors/darkframe images; documentation updated

Version 3.77 - 2000-07-31 (UL)
------------
+ fixed bug in OmGeoLinearizer class (caused polynomial coefficients
to be read always from last table row irrespective of the filter)
+ brought in line with calpnalgo-2.10

Version 3.76 - 2000-07-27 (UL)
------------
+ fixed SPR 1612: algorithm to CAL_omLEDTemplate not implemented
+ implemented GratingDataServer::misalignmentFigureError()

Version 3.75.8 - 2000-07-04 (UL)
--------------
+ depends on selectlib-4.35

Version 3.75.7 - 2000-06-27 (UL)
--------------
+ fixed SPR 1436: test harness fails on Linux
+ bug fix in CcdQuantumDataServer (caused spatial variation of QE
to be incorrectly retrieved)

Version 3.75.6 - 2000-06-20 (UL)
--------------
+ fixed minor problem with broken doc link

Version 3.75.5 - 2000-06-15 (UL)
--------------
+ another compilation problem under gcc-2.95.1
+ fixd SPR 1569: make depend fails without make clobber

Version 3.75.4 - 2000-06-15 (UL)
--------------
+ fixed compilation problem under gcc-2.95.1

Version 3.75.3 - 2000-06-14 (UL)
--------------
+ fixed SPR 1488: RGA misalignment not considered in ROWCOORD
transformation
(bug in Rowland coordinate transformation caused RGA misalignment
correction to go in wrong direction: positive RGA +Y axis angles lead
errorneouly to greater BETA values)
+ fixed SPR 1472: lincoord viewable not alway clear
(problem was OCB value in viewable)

Version 3.75.2 - 2000-06-05 (UL)
--------------
+ removed paramter default values of GratingDataServer methods

Version 3.75.1 - 2000-05-30 (UL)
------------
+ MirrorDataServer: intrinsic RGS effective area available for rgsrmfgen
+ GratingDataServer: alternative form of self-vignetting efficiency
available for rgsrmfgen
+ alternative form of EMOS redistribution I/F from Fortran-90
(tasks can specify energy axis)
+ default EMOS redistribution method now evaluates model on energy
axis from CCF::REDIST::BINNEDPI_EBOUNDS

Version 3.75 - 2000-05-24 (UL)
------------
+ camera mode determination logic is now CCF::ModeParam driven
(needs caloalutils-1.28)
+ fixed SPR 1347: CAT_getQuantum Efficiency: EPN/PATTERNLIB
+ depends on testccf-1.83 (reconciled test harnesses)

Version 3.74 - 2000-05-23 (UL)
------------
+ several bug fixes in new EMOS redistribution code
(verified correctness against rmfgen code of Steve Sembay)
+ made SUBDATE attribute in CCF constituents optional

Version 3.73 - 2000-05-22 (UL)
------------
+ bug fix in EnergyCorrector (caused RGS gain correction to fail)
+ fall back upon CCF numbers if given RGS offset override table is
empty

Version 3.72 - 2000-05-22 (UL)
------------
+ fixed compilation problem under True64
+ bug fix in CanonicalRedist constructor

Version 3.71 - 2000-05-19 (UL)
------------
+ fixed SPR 972 : CAL_getRedistribution and CAL_gainCorrect are incoherent
(new EMOS redistribution model implemented)
+ more functionality in CcdQuantumDataServer
+ test harnesses resynced with testccf-1.81
+ abscissa in Redist CalViewable is now in eV
+ brought in line with dal-1.119

Version 3.70 - 2000-05-16 (UL)
------------
+ reverted f90 test harnesses to programs

Version 3.69 - 2000-05-09 (UL)
------------
+ bug fix in RGS vignetting curve construction
+ new MOS gain model implemented
+ fixed minor problem in GeometryDataServer I/F
+ f90 test harnesses are proper tasks
+ CAL_pnReEmissionThreshold is now CCF-driven
+ depends on testccf-1.79

Version 3.68 - 2000-05-05 (UL)
------------
+ bug fix in reading of grating efficiency data
+ implemented access to vignetting information from f90
+ closed SPR 1279: Misusing macros in documentation
+ closed SPR 1327: Misspelling and inconsistency of var names
+ closed SPR 1264: Too sparse error message

Version 3.67 - 2000-05-04 (UL)
------------
+ really fixed problem in reading of RGS CTI coefficients (was supposed
to be fixed in 3.66)

Version 3.66 - 2000-05-02 (UL)
------------
+ fixed small problem with reading of RGS CTI coefficients from CCF
+ set default energy in CalView state to 100eV

Version 3.65 - 2000-04-27 (UL)
------------
+ added GratingProperties viewables
+ added AeVs{E/theta/phi} viewables
+ added vignettingVs{E/theta/phi} viewables
+ moved contents of ccftools directory to package ccftools
+ VALDATE/DATE in CCF constituents can contain invalid dates
+ minor documentation fix

Version 3.64 - 2000-04-25 (UL)
------------
+ fixed problem with RGS CTI correction (scale factor was not considered)
+ additional mode parameters access methods
+ temporarily disabled instantiation of low-accuracy MirrorDataServer
(numerical problems with 2-D polynomial fit from IDL)

Version 3.63 - 2000-04-13 (UL)
------------
+ fixed problem in Makefile which caused a stale symlink
+ brought in-line with calpnalgo-2.6 (new error code from
gain correction)
+ bug fix in GratingDataServer::~GratingDataServer()
(caused assert failure when releasing cal atom)
+ added retrieval of effective area data through f90
+ fixed SPR 1259: CAL_getEffectiveArea argument order different in
cal.f90 and cal.ps
+ fixed SPR 1250: Invalid absolute links

Version 3.62 - 2000-04-12 (UL)
------------
+ disabled consistency check of CTI coefficient values for MOS
+ minor fixes in RgsCrossPsf/EnergyCorrector
+ StateVariable.h moved to caloalutils package

Version 3.61 - 2000-03-29 (UL)
------------
+ reconciled with testccf-1.71/slatec-1.31

Version 3.60 - 2000-03-27 (UL)
------------
+ new CTI correction for EMOS (constant charge loss per transfer model)
+ test harnesses reconciled with testccf-1.70

Version 3.59 - 2000-03-24 (UL)
------------
+ dropped time dependence of CAL_pn{Gain, Cti}Correct
+ implemented CAL_rgsEffectiveArea + associated CalViewable
+ implemented Viewable Vignetting curve

Version 3.58 - 2000-03-22 (UL)
------------
+ minor bug fix in CalViewables

Version 3.57 - 2000-03-21 (UL)
------------
+ workaround for Dal bug implemented (to make test harness go through)

Version 3.56 - 2000-03-20 (UL)
------------
+ implemented CAL_getEffectiveArea
+ brought in line with calpnalgo-2.5

Version 3.55 - 2000-03-09 (UL)
------------
+ fixed problem with CalDataVector::clone()
+ added XmmPixelList::insideFOV() to determine whether a pixel in
within the FOV
+ FOV info is obtained from LinCoord::FOV table (optional)
+ added CAL_getFOVmap + corresponding test harness

Version 3.54 - 2000-03-09 (UL)
------------
+ FilterDataServer/CcdQuantumDataServer return now weighted integrals
of filter/QE curves
+ fixed compilation problem under gcc-2.95.2
+ added CalDataVector::clone() (= deep copy constructor)
+ fixed another problem with RgsEnergyCorrector::offsetCorrect()
+ "CoordTransformer" -> "GeometryDataServer"

Version 3.53 - 2000-03-03 (UL)
------------
+ RgsCrossPsf::probablity: center channel is _negative_ of input source
off-axis angle
+ reconciled implementation of OM routines with latest descriptions in
CAL HB

Version 3.52 - 2000-03-01 (UL)
------------
+ brought in line with utils-1.66
+ RgsEnergyCorrector::offsetCorrect(): bug fix + more sensible handling of
illegal event shapes

Version 3.51 - 2000-02-29 (UL)
------------
+ fixed typo in documentation of CAL_setState
+ RGS offset correction I/F:
o grade -> shape
o changed type of Ein from real32 to int16
changes are in line with CAL HB

Version 3.50 - 2000-02-23 (UL)
------------
+ fixed (another) bug in Rgs::Redistribution()
+ added f90 I/F to retrieve Boresight angles + conversion to
matrix
+ implemented RGS offset correction according to CAL HB incl.
optional usage of offset file produced by rgsoffsetcalc
+ altered I/F of CAL_getFilterTransmission/CAL_getQuantumEfficiency +
make use of underlying C++ classes
+ brought in line with dal-1.104 (changed compound name syntax)

Version 3.49 - 2000-02-17 (UL)
------------
+ prototype of CcdQuantumDataServer atom + simple test harness added
+ added CalViewable for QE data
+ defintion of cross-dispersion angle in ROWCOORD system has changed:
is is now defined w.r.t the center of the mirror (now longer w.r.t
to the G-point)
+ now depends on selectlib

Version 3.48 - 2000-02-16 (UL)
------------
+ another bug fix in f90-wrapper of RGS redistribution function
+ test harness for RGS tests now peak position and width at two
different energies
+ fixed SPR 1162: [PATCH] caltest_om: The 'Cut&Paste' strikes back with
DEBUG=no
+ fixed compilation problem under Linux in CoordTransformer.cc
+ depends now on selectlib
+ FilterDataServer now accepts new structure of CCF::FilterTransX
constituents (REGION column)

Version 3.47 - 2000-02-11 (UL)
------------
+ dependent on testccf-1.63
+ EPN gain/CTI test harness also tests the low-accuracy algorithm

Version 3.46 - 2000-02-09 (UL)
------------
+ bug fix in AbsCoef Viewable
+ fixed more compilation problems under gcc-2.95
+ bug fix in f90-wrapper to RGS redistribution function
+ better test harness for RGS redistribution from F90

Version 3.45 - 2000-02-04 (GV/UL)
------------
+ Redistribution.cc: the RGS::noise spectrum was too narrow, due
to a misinterpretation of Slatec::derf.

Version 3.44 - 2000-02-03 (UL)
------------
+ now depends on slatec-1.26
+ made presence of keywords EXTSEQU/EXTSEQID in CCF override constituents
optional

Version 3.43 - 2000-02-02 (UL)
------------
+ low-accuracy EPN gain correction implemented
+ low-accuracy EPN CTI correction implemented
+ prototype of FilterDataServer
+ AbsCoef CalCiewable with SubViewables to select different elements
+ new CalViewable for the FilterDataServer
+ test harness for FilterDataServer
+ added RgsEnergyCorrector::qSplitRatios()

Version 3.42 - 2000-01-31 (UL)
------------
+ fixed compilation problem with gcc-2.95
+ adapted to new AbsCoef structure in CCF (one table per element
each with own energy scale)

Version 3.41 - 2000-01-13 (UL)
------------
+ fixed convolution + gain problem in Rgs::Redistribution()
+ implemented CAL_omDistorition() according to Cal HB
+ implemented CAL_omColorTransform() according to Cal HB
+ implemented CAL_omPhotoMagnitude() according to Cal HB
+ implemented CAL_omPixelSenseVariation/omLargeSenseVariation according
to Cal HB
+ implemented CAL_omPhotoNatural according to CAL HB
+ implemented CAL_omGetFrameParameters according to CAL HB

Version 3.40 - 2000-01-05 (UL)
------------
+ retrival of atoms produces now proper error messages in case of missing
data in CCF (related to SPR 1056 : CAL_getModeParameters core dumps in
Timing mode for MOS)
+ fixed SPR 1052 : CAL_gainCorrect does not work in Timing mode for EMOS

Version 3.39 - 2000-01-04 (UL)
------------
+ fixed wrong normalization of RGS redistribution function (norm is now
tested in test harness)

Version 3.38 - 1999-12-23 (UL) - the Christmas CAL!
------------
+ added Int32Vector instantiation

Version 3.37 - 1999-12-17 (UL)
------------
+ RGS redistribution model fudged

Version 3.36 - 1999-12-16 (UL)
------------
+ added state variable 'temperatureTracking'
+ added support for calmosalgo

Version 3.35.2 - 1999-12-10 (UL)
--------------
+ brought-in line with calpnalgo-2.2
+ fixed SPR 1017 : [PATCH] cal still broken with TRACE=yes

Version 3.35.1 - 1999-11-30 (UL)
--------------
+ bug fix related to new compound syntax for attributes in CCF/MiscData

Version 3.35 - 1999-11-30 (UL)
------------
+ data items in MiscData can now point to attributes in different
constituents (new colunm "FIND_IT_IN")
+ tasks can access attributes in CCF constituents with compound
syntax CONSTITNAME:[BLOCKNAME]:ATTRNAME
+ normalization of RGS cross-dispersion PSF tested in test harness
+ bug fix in fine time computation of EPN in timing mode
+ dropped support for old SAS_CCFDIR environment variable
+ ROWCOORD now in RGA frame (takes into account rotation of
RGA w.r.t S/C frame)
+ fixed SPR 979 : buggy SAS_DEBUG statements break cal build [PATCH]

Version 3.34 - 1999-11-24 (UL)
------------
+ added CalInt32Vector to CalDataVector.h
+ added TimeCorrector atom class
(EPN: implementation + test harness; EMOS, RGS: leave input
unchanged)
+ updated I/F to calpnalgo-2.0
+ renames libcalview -> libcalviewbase to resolve name conflict
with 'calview' task
+ modified CAL_crossPsf I/F

Version 3.33 - 1999-11-15 (UL)
------------
+ minor documentation update
+ bug fix in f90 wrapper to pixCoord1->chipCoord conversion
+ updated RGS test harness
+ bug fix in XmmCartesianPointList::toRowCoord()
+ bug fix in RgsCrossPsf::probability()

Version 3.32 - 1999-11-10 (UL)
------------
+ avail new combilib features in sas-1.169
+ fixed error in documentation (TELCOORD frame is related to CAMCOORD2
_not_ CAMCOORD1)
+ boresight calibration related changes:
o new Boresight class with I/F to explicitely set the angles
o f90: + CAL_getBoresightMatrix returns pure matrix that
no longer incorporates CC2->SACCOORD transformation
+ new explicit I/F to retrieve CC2->SACCOORD transformation
+ all sources + test harnesses brought in line with testccf-1.42
(updated column/attribute/table names)
+ brought in line with caloalutils-1.14

Version 3.31.1 - 1999-10-28 (UL)
--------------
+ only calpnalgo is merged into the combilib (merging in slatec +
caloalutils does not work)

Version 3.31 - 1999-10-27 (UL)
------------
+ all libraries that CAL needs now in CAL combilib
+ reintroduced eBins()/eBounds distinction in Redistribution classes
(this fixes: SPR 911 : Reinstate CAL_getEbins())
+ return now correct number of bins for EMOS redistribution

Version 3.30 - 1999-10-26 (UL)
------------
+ calpnalgo lib now part of the cal combilib

Version 3.29 - 1999-10-21 (UL)
------------
+ added HKParmInt Viewable
+ renamed doc/sources to doc/calsources

Version 3.28 - 1999-10-19 (UL)
------------
+ added HKvalRangesServer atom

Version 3.27 - 1999-10-12 (UL)
------------
+ use INTERFACE / combilib mechanism
+ fixed SPR 795: broken links in documentation
+ fixed SPR 616: examples
(all PGPLOT calls in example tasks have been commented out; should
be replaced with proper CalViewables)

Version 3.26 - 1999-10-11 (UL)
------------
+ fixed SPR 888: CATEGORY = `SCISIMCCF' not accepted by option --ccffiles
(no longer check value of CATEGORY against "XMMCCF"; print actual value
as message if verbosity is >= VERBOSE
+ fixed SPR 889: Local replacement of a CCF does not work
(dirname of $SAS_CCFDIR was always prepended to constituent name
even if it came with an abosolute path)
+ fixed SPR 853 : non-descriptive NotYetImplemented warnings
+ boundary check of state variable `observationStartTime':
o lower bound check against Xmm::missionReferenceTime
o no check on upper bound (may XMM live forever ;)

Version 3.25 - 1999-09-30 (UL)
------------
+ fixed bug in CHIPCOORD->CAMCOORD1 conversion routines which caused
PN chip gaps to come out wrongly; the gaps are now checked for
in a dedicated test harness

Version 3.24 - 1999-09-27 (UL)
------------
+ bug fix in CoordTransformer
+ brought in line with caloalutils-1.7

Version 3.23 - 1999-09-17 (UL)
------------
+ fixed wrap-around problem of RGS-response/noise convolution
+ pnReEmissionThreshold expects energy in ADU
+ automatic flush of memory buffered atoms if new CCF constituent
is inserted (only relevant for calview)
+ added AbsCoefPlot viewable
+ RGS cross-dispersion PSF + corresponding Viewable implemented
+ removed DOCS=cal.html from cal/doc/Makefile

Version 3.22 - 1999-09-15 (UL)
------------
+ bug fix in implementation of RGS redistribution model

Version 3.21 - 1999-09-14 (UL)
------------
+ added DISTRIBUTION file
+ various changes following fix of DAL SPR 589 : incorrect treatment
of FITS B-type
+ added few utility classes to src/CalCommon.h
+ fixed SPR 692: calexample_emos and calexample_epn segfaults on Linux
(probably utils-1.43->1.44 has fixed this)
+ added BinnedAxis class family to src/CalCommon.h
+ GainCorrector -> EnergyCorrector
+ MiscData -> MiscDataServer
+ implemented fractional CHIPCOORD pixels (available from C++ I/F only)
+ implemented retrieval of RGS event thresholds
+ implemented access to absorption coefficients in CCF (Henke data)
+ new RGS CCD redistribution model implemented
+ new RGS CTI correction model implemented
+ EPN CTI correction takes now E_PER_ADU factor into account

Version 3.20 - 1999-09-13 (UL)
------------
+ brought in line with dal-1.70

Version 3.19 - 1999-09-09 (GV)
------------
+ added slatec to USEDLIBS in ccftools/Makefile

Version 3.18 - 1999-08-26 (UL)
------------
+ more diagnostic messages
+ udated documentation

Version 3.17 - 1999-08-20 (UL)
------------
+ fixed LaTeX problem in documentation

Version 3.16 - 1999-08-19 (UL)
------------
+ construction of boresight misalignment matrix overhauled; includes
now the transformation from CAMCOORD2->SACCCORD as agreed with
the author of the attcalc task (Andy Read)
+ fixed SPR 635: "pure virtual method called" is not an error messsage
+ removed obsolete state variables `status' and `verbosityLevel'

Version 3.15 - 1999-08-12 (UL)
------------
+ bug fix in operator<<(ostream &, CalState &);
+ new method Ccf::cifName()
+ changes to calview I/F
+ coordinate conversion routines: TELCOORD->CAMCOORD2 and
CAMCOORD1->CHIPCOORD completed

Version 3.14 - 1999-08-05 (UL)
------------
+ bug fix in toChipCoord conversion routine (OCB was not taken
into account)
+ fixed SPR 719: [PATCH] Unresolved symbol: CartesianVector::operator-= in
libcal; was actually rectified by fix in caloalutils-1.3

Version 3.13 - 1999-07-27 (GV)
------------
+ change dnumber of bins in calhack.f90

Version 3.12 - 1999-07-09 (GV)
------------
+ added calhack.f90, f90-only implementation of some urgent RGS
functions. Must be replaced with proper implementation.
+ DEPEND: slatec >= 1.13 is now required, together with sas >=
1.141, where the slatec library is added by Make.include

Version 3.11 - 1999-07-09 (GV)
------------
+ change DEPEND
+ comment out rgslincoordgrcal in Makefile in ccftools as this does not link.
Version 3.10 - 1999-07-09 (UL)
------------
+ better error/warning messages in CtiCorrector
+ coordinate conversion routines (CCRs) re-written (functionality is now
in new Cal atom CoordTransformer)
+ CCRs take thermal expansion effects into account
+ CCRs are in line with new CAMCOORD1 data in LinCoord sets
+ added functionality to convert to new ROWCOORD system (Rowland
coordinates beta/delta) for RGS
+ new RgsGainCorrector class which implements RGS-specific offset
correction (with SER-dependency)
+ new state variables
o ocb (on-chip-binning factor for RGS; is taken into account by CCRs)
o temperatureCamera (temperature of the camera structure which drives
the thermal expansion of the inter-CCD gaps)
+ ccftools: two new tools to generate RGS LinCoord data from ground
alignment data

Version 3.9 - 1999-06-30 (UL)
-----------
+ bug fix in PIXCOORD -> CAMCOORD1/2 transformation (incorrect
treatment of last pixel for EPN)

Version 3.8 - 1999-06-25 (UL)
-----------
+ bug fix interface directory
+ documentation updated (section on coordinate system revised
and extented; has now pictures with camera geometries)

Version 3.7 - 1999-06-23 (UL)
-----------
+ contents of subdirectory caloalutils moved into separate package
which the Cal now depends on; this has the advantage that tasks
which just use the Oal do not depend on the Cal any longer

Version 3.6 - 1999-06-22 (UL)
-----------
+ new class CtiCorrector which performs CTI corrections for EMOS/RGS/EPN;
the latter uses the code in new package calpnalgo
+ several internal changes
+ updates to GainCorrector related to EPN (uses now code in calpnalgo)
+ several changes in f90/C++ interface part as a consequence of above
changes

Version 3.5 - 1999-06-15 (UL)
-----------
+ minor bug fix in KeplerOrbitSimulator
+ added computation of velocity vectors to KeplerOrbitSimulator

Version 3.4 - 1999-06-14 (UL)
-----------
+ removed dependency on calpnalgo for the time being

Version 3.3 - 1999-06-11 (UL)
-----------
+ fixed SPR 610 : perl modules: no longer needed and/or in the wrong place
+ added KeplerOrbitSimulator class to caloalutils

Version 3.2 - 1999-06-08 (UL)
-----------
+ minor bug fixes related to offset correction for PN
+ fixed portability problem on Dec Alpha (SPR 681: CalF90cxxGlue.cc:
portability bug; assumption sizeof(int)==sizeof(ptr)
+ CAL_setRandom() added

Version 3.1 - 1999-06-07 (UL)
-----------
+ documentation updated; now in line with latest code changes in v3.0
(this fixes SPR 686: Doc is obsolete)

Version 3.0 - 1999-06-03 (UL)
-----------
+ tidying up of F90 API:
+ energy conversion routines (offset, gain, CTI) harmonized across
instruments
+ removed obsolete parameters in F90 API (vector sizes)
+ naming of routines is more coherent
+ a couple of obsolete routines dropped from API (e.g. CAL_error)
+ overloaded version of routines which accept Dal column pointers
now obsolete and dropped (vector-processing routines accept
now pointers to allocated vector as well as uninitialized pointers)
+ new: CAL_integerToReal() - with (optional) randomization
CAL_realToInteger()
+ RAWXY->DETXY does now make use of pixel randomization (optional)
+ pnReEmissionThreshold added to F90 API
+ internal changes

Version 2.21 - 1999-05-12 (UL)
------------
+ brought in-line with error-1.18

Version 2.20 - 1999-04-27 (UL)
------------
+ more fixes to make the build under Dec succeed
+ no longer use fatal()
+ changed all error/warning/message tags
+ redistribution channel range now from 0-4095 for EMOS+EPN

Version 2.19 - 1999-04-22 (UL)
------------
+ in setting the state: unit of attribute "FRMTIME" is taken into
account
+ mode-recognition logic for EMOS slightly changed

Version 2.18 - 1999-04-19 (UL)
------------
+ src/CalState:
+ minor bug fix in operator<<
+ state can now be read from DataSet or Block (also available from F90)
+ caloalutils/Xmm.cc:
+ fixed minor bug in recognition of RGS modes

Version 2.17 - 1999-04-15 (UL)
------------
+ added newlines to some source files to avoid warning from C++

Version 2.16 - 1999-04-15 (UL)
------------
+ minor changes in test harnesses enforced by error-1.16

Version 2.15 - 1999-04-14 (UL)
------------
+ minor bug fix in f90/C++ layer
+ newlines removed from error/warning messages

Version 2.14 - 1999-04-14 (UL)
------------
+ changes enforced by error-1.16

Version 2.13 - 1999-04-12 (UL)
------------
+ caloalutils/Xmm.h: minor improvements
+ calview: renamed lib calview.a to calviewbase.a to resolve name conflict
with the calview task
+ src/StateVariable.h: fixed minor problem resulting from a name conflict
with a global function in a Qt header file (caused compilation
problems for tasks using Cal and Qt together)

Version 2.12 - 1999-04-01 (UL)
------------
+ f90: minor bug fix in mode conversion routines (caused caltest_misc
to fail)

Version 2.11 - 1999-04-01 (UL - this is no April-fools-day version ;)
------------
+ first working version of CalView backend (three Viewables are
implemented: CcdPlot, RmfCutPlot, BadPixPlot
+ fixed SPR 543: Improvements to CAL_setStateFromScienceFile
+ caloalutils: new mode and filter names
+ new implementation of Cal state: setting variables to illegal values
now caught + other improvements; it is now possible to inquire the
legal range/list of legal values of variables
+ resolved problem with vector constructor (unsigned -> size_t) which
caused compilation problem on Dec Alpha

Version 2.10 - 1999-03-17 (UL)
------------
+ new EMOS RMF parameterization from Dave
+ default name for CIF changed to "ccf.cif"
+ documentation has now table of contents to simplify searches
for particular routines

Version 2.9 - 1999-03-12 (UL)
-----------
+ documentation updated with description of CAMCOORD systems

Version 2.8 - 1999-03-10 (UL)
-----------
+ upgrades pertaining to low memory Dal
+ ERM part of xmmPayLoad
+ f90/cal.f90: 'LOW/MEDIUM/HIGH' -> 'ACCURACY_LOW'/'ACCURACY_MEDIUM'/
'ACCURACY_HIGH'
+ caloalutils/Xmm.h: o nNodes() for RGS was giving the wrong answer
o maxX/maxY were swapped
+ several changes to test harnesses
+ detector coordinate systems introduced
+ works with TRACER disabled

Version 2.7 - 1999-02-24 (UL)
-----------
+ caloalutils: global object xmmPayLoad

Version 2.6 - 1999-02-23 (UL)
-----------
+ moved common OAL/CAL functionality to separate new directory;
will be moved to a separate package later

Version 2.5 - 1999-02-22 (UL)
-----------
+ f90/caloaldefs.f90: new routine toTimeTag to convert
"yyyy-mm-ddThh:mm:ss" into event time tag

Version 2.4 - 1999-02-17 (UL)
-----------
+ fixed initialization problem in src/Ccf.cc

Version 2.3 - 1999-02-12 (UL)
-----------
+ minor change in documentation

Version 2.2 - 1999-02-12 (UL)
-----------
+ tests in cxxtests no longer need pgplot

Version 2.1 - 1999-02-12 (UL)
-----------
+ fixed problem with failing tests in f90test

Version 2.0 - 1999-02-10 (UL)
-----------
+ renamed environment variable CCF_DIR to SAS_CCFDIR; the old name is
no longer recognized which might break several tasks/test harnesses
+ fixed SPR 436: Task subroutines should not initialise the CAL
the usage of CAL_openCcf in task code is now deprecated; the global
calServer object will access the default CCF pointed at via the
environment variable SAS_CCFDIR upon first usage of CAL services;
alternatively, taskmain loads a particular CCF which is triggered by
naming a CCF directory via the generic task parameter "--ccfdir"

Version 1.21 - 1999-02-03 (UL)
------------
+ needed file src/Xmm.h was missing from 1.20 by mistake

Version 1.20 - 1999-02-03 (UL)
------------
+ code overhaul: there is now only a single global instance of a CalServer
object
+ CAL can now be built as shared library
+ Xmm.h: new global object XmmPayLoad
+ use of 'status' and 'error' state variables is now deprecated - usage
of these in setState/getState will produce a corresponding warning
message

Version 1.19 - 1999-01-29 (UL)
------------
+ make the Cal compile under egcs 1.1.1

Version 1.18 - 1999-01-26 (UL)
------------
+ make the Cal compile under gcc 2.8.1

Version 1.17 - 1998-11-13 (UL)
------------
+ new build procedure in-line with what is done in the Dal: separate
libaries are build in each src-subdir which get combined in a
new directory 'interface'; this also contains now all needed symbolic
links to *.h/*.f90/*.mod; this closes
o SPR 294: Can't the library be built without additional rules?
o SPR 297: Can't the library be built without additional rules?
(f90/Makefile)

Version 1.16 - 1998-11-12 (UL)
------------
+ in *.f90: made declaration of function return types explicit (fixes
compilation failure problem when 'f90 -u' is used

Version 1.15 - 1998-11-12 (UL)
------------
+ fixed SPR 273: Need symbolic link to cal.f90 and caloaldefs.f90

Version 1.14 - 1998-11-04 (UL)
------------
+ minor problem in ccftools/Makefile fixed

Version 1.13 - 1998-10-30 (UL)
------------
+ in documentation:
o fixed min/max pixel values for EMOS, i.e., 1/600 _not_ 0/599
- eduThreshold parameter in CAL_getEventPatterns now optional
+ fixed SPRs:
SPR 455 : caltest_emos fails (core dumps)
SPR 474 : Definition of optional arguments in CAL_getBadPixelList
SPR 476 : ccdNodeId for EPIC PN
+ fixed crash of calexample_emos due to arithmetic exception

Version 1.12 - 1998-10-15 (UL)
------------
+ in f90:
o fixed typo: rgs_CTIcorrrection -> rgs_CTIcorrection
o fixed version of rawxy2mm, gainCorrection, CTIcorrection
which work on Dal Columns (Dal now returns Handle)
o minor change in filterIdToString()
o new routine OAL_rgsOffset()
+ in f90test:
o added test for CTI and gain correction to caltest_rgs
o fix in caltest_emos.f90 (one CAL_releaseMemory too many -
caused crash under OSF)

Version 1.11 - 1998-10-06 (UL)
------------
+ fixed broken link in documentation

Version 1.10 - 1998-10-06 (UL)
------------
+ added #include in several places

Version 1.9 - 1998-10-01 (UL)
------------
+ three new routines added to API (CAL_getParticleBackground,
CAL_getDiffuseXBackground, CAL_rgsGetRefinedWaveScale);
calexample_emos.f90/caltest_emos.f90 extended accordingly;
documentation updated
+ CAL_mm2thetaPhi/CAL_thetaPhi2mm propagated to f90 layer
+ fixed broken link in doc to CCF ICD
+ fixed SPR 438: wrong format of atom in warning
+ fixed 442: Do not use STOP, call error instead
+ reading of EMOS pattern lib mode dependent
+ setStateFromScienceFile() implemented and corresponding tests added
to f90test
+ gainCorrection() now CCF-based (for PN uses ADUConv file provided by
hardware group)
+ added missing includes (iomanip.h) (MB)
+ fixed improper return (ALTRETURNV) (MB)
+ upgraded message calls in test harness (MB)
+ now creates links to perl files in $SAS_DIR/lib/perl (MB)

Version 1.8 - 1998-08-26 (UL)
-----------
+ changes to Xmm.h (new methods in class `EPNcam')
+ new output operator for AttitudeMatrix
+ closed tiny memory leak in CalOalUtility::stringColumnRow()
+ internal changed to Ccf.h

Version 1.7 - 1998-07-24 (UL)
-----------
+ CAL_getEventPatterns is now CCF-based
+ various changes following testccf upgrade to v1.7

Version 1.6 - 1998-07-20 (UL)
-----------
+ fixed SOC-SPR 360 about linking problem on Dec
+ changes to Makefiles following upgrade of package 'sas'
+ moved 'common.h' -> utils

Version 1.5 - 1998-07-09 (UL)
-----------
+ fixed SOC-SPRs:
o 242: CAL documentation
o 265: Package documentation has wrong name
o 295: Remove the spurious subdirectory ccftools
o 296: Rename calapi.tex to cal_description.tex and add cal.tex
o 363: doc makes use of raw HTML directives
o 375: Positions of detector centres.
o 360: calviewdummy fails to link on Digital Unix
o 380: calsample fails

Version 1.4 - 1998-07-08 (UL)
-----------
+ blocks in CCF constituents are now accessed by name - necessary
due to changes in Dal-1.0 -> Dal-1.1 - because of this, testccf
needs to be in version 1.4 or higher

Version 1.3 - 1998-07-07 (UL)
-----------
+ added test harness for StringPool class

Version 1.2 - 1998-05-06 (UL)
-----------
+ fixed SOC-SPRs:
o 95: Symbol bloat
o 174: Wrong CIF file
o 193: OAL/CAL should use error instead of exitval
o 196: getEventPatterns returns transposed patterns
o 201: Detector coordinates for EPIC MOS
o 205: MOS pixel size

Version 1.1 - 1998-05-14 (UL)
-----------
+ Initial version for SAS v1 integration