Warning, /geant4/CHANGELOG/Patch4.9.1-2.txt is written in an unsupported language. File is not indexed.
0001
0002 Geant4 9.1 - patch-02 Release Notes
0003 -----------------------------------
0004
0005 9 May 2008
0006
0007 List of fixes included in this public patch since the public release 9.1.p01:
0008
0009 o Configuration:
0010 -------------
0011 + Fixed in liblist.c use of gets() by adopting safer fgets().
0012 + Linux-icc.gmk: updated setup to make use of icc-10.1 and ifort-10.1;
0013 changed compilation option "-mp" to more performant "-fp_model precise".
0014 + Linux-g++.gmk: Replaced g77 with gfortran compiler.
0015
0016 o Environments:
0017 ------------
0018 + Fixed compilation problems.
0019 + Make direct usage of tags in G4Version header.
0020
0021 o Geometry:
0022 --------
0023 + Cleared compilation warnings detected on gcc-4.3.0.
0024 + magneticfield:
0025 o Fixed 'unused' non-initialized values of dydx in classes
0026 G4EqMagElectricField and G4EqEMFieldWithSpin.
0027 + management:
0028 o Added missing protected copy-contructor and assignment operator
0029 for G4VSolid.
0030 o Added Shrink() and Reserve() methods to G4SmartVoxelNode to be used
0031 for debugging studies.
0032 + navigation:
0033 o G4PathFinder: commented out warning and exception in ComputeStep()
0034 for steps smaller than minimum threshold.
0035 + solids/CSG:
0036 o G4Cons: added protection against NaN in GetPointOnSurface() in the
0037 case the check is run for Rmax1==Rmax2.
0038 o G4Trap: corrections for arithmetic operations in 8-point constructor.
0039 Addressing problem report #934.
0040 + solids/specific:
0041 o G4ExtrudedSolid: bug fix in MakeFacets() to fill 'fTriangles' also
0042 in case of polygon with 3 or 4 vertices, as they are needed for
0043 Inside(p). Fix in Inside(p) for correctly handling the case of
0044 points located on the edges of triangles which are not the outline
0045 edges of the polygon.
0046 o G4TessellatedSolid: fixed bug in copy-constructor and internal method
0047 CopyObjects() for cloning of facets. Fix in DistanceToOut(p,v,..) to
0048 compute normal only if specified by flag in input.
0049 o G4Polyhedra: fixed problem of out-of-bound access to vectors in
0050 method GetPointOnSurface().
0051 o G4IntersectingVone: added protection against NaN in LineHitsCone1()
0052 in case of configurations where a conic section becomes a tube.
0053 Added kCarTolerance in definition of "low" and "high" cone limits
0054 in R and in Z in order to protect against numerical errors; this
0055 fixes cases of DistanceToIn(p,v) returning 'kInfinity' when p+t*v
0056 cross intersection between two cones.
0057 o G4EllipticalCone: removed some debug printouts.
0058 o G4Hype: refined polyhedron representation for correct visualization
0059 of the shape.
0060
0061 o Global:
0062 ------
0063 + Introduced inclusion of <cstring> and <climits> in G4String.hh and
0064 templates.hh respectively for porting on gcc-4.3.0 compiler.
0065 + Cleared compilation warnings detected on gcc-4.3.0
0066 + Removed useless specification of units in implementation of
0067 G4GeometryTolerance::SetSurfaceTolerance().
0068 + Changed date for patch-02 to release 9.1.
0069
0070 o Graphics Representations:
0071 ------------------------
0072 + Added G4Hype representation to HepPolyhedron and G4Polyhedron.
0073 + Added inclusion of <cstdlib> in HepPolyhedron.cc for porting on
0074 gcc-4.3.0 compiler.
0075
0076 o Materials:
0077 ---------
0078 + G4NistMaterialBuilder: using correct dimension for the density for
0079 ConstructNewMaterial() methods.
0080 + G4Material: improved printout in error conditions by specifying the
0081 name of the material affected.
0082
0083 o Particles:
0084 ---------
0085 + G4ParticleTable: fixed implementation of methods Remove() and
0086 RemoveAllParticles() so that particle definition objects are properly
0087 deleted when invoking these methods between runs. It's user's
0088 responsibility to delete dynamically instantiated particle objects.
0089 Added method DeleteAllParticles() to remove and delete all particles
0090 in the list. Addressing problem report #1007.
0091 + Fixed bugs in G4Iontable::Remove() and G4ShortLivedTable::Remove().
0092
0093 o Physics Lists:
0094 -------------
0095 + Fixed deletion using uninitialized pointers in FTF-BIC, for classes
0096 HadronPhysicsFTF_BIC, G4FTFBinaryNeutronBuilder and G4LEPNeutronBuilder.
0097
0098 o Electromagnetic Processes:
0099 -------------------------
0100 + Standard:
0101 o G4UrbanMscModel90, G4UrbanMscModel: fixed asymmetry of lateral
0102 displacement.
0103 o G4ionIonisation: adopt 'linLossLimit' 0.15 as it was originally in 9.1.
0104 o G4IonFluctuations: fixed discontinuity of width correction factor
0105 for small velocities and added a protection to its value
0106 (factor>1), which fixes very long alpha events.
0107 + Utils:
0108 o G4EmCorrections: set finite size correction to zero (provides flat
0109 distribution of vertex detector response for high energy hadrons).
0110 o G4ionEffectiveCharge: minor change of numerical constants to provide
0111 continuous change of charge versus energy.
0112
0113 o Hadronic Processes:
0114 ------------------
0115 + Cleared most compilation warnings detected on gcc-4.3.0.
0116 + cross_sections:
0117 o G4NeutronInelasticCrossSection: fixed bug which used to return zero
0118 cross section for 3He.
0119 o G4TripathiLightCrossSection: fixed bug in method IsApplicable() to
0120 use GetN() instead of GetA() for target.
0121 o Added capability to use element-based cross section data
0122 sets if no isotope-based ones are available. This is required to
0123 get the HP neutron code to work within the cross section framework
0124 introduced in version 9.0. Added internal method to class
0125 G4CrossSectionDataStore.
0126 + models/cascade:
0127 o Added Coulomb barrier in G4EquilibriumEvaporator.
0128 o Changed pp, nn total cross section parameterization in class
0129 G4CascadSpecialFunctions to more closely reflect the PDG values;
0130 fixed cross section bug responsible for returning low or zero total
0131 cross sections for pi-p, pi+n, np, and pi0 interactions.
0132 + models/chiral_inv_phase_space:
0133 o Modified G4QElasticCrossSection to fix problem with quasi-elastic.
0134 When run together with elastic, the parameters to sample T were not
0135 updated, this fix is to always update these.
0136 o Fixed case of exceeded boundary bounds in G4QCollision.
0137 o Check on tolerance is replaced by check on exact equality for fixing
0138 event reproducibility.
0139 + models/coherent_elastic:
0140 o G4HadronElastic: added extra protection against precision lost
0141 when projectile almost lost kinetic energy; set 'lowestEnergyLimit'
0142 to be 1.e-6*eV (was 0.0).
0143 Added extra protection: if sampled t out of the physics range then
0144 no scattering.
0145 o G4UHadronElasticProcess: call to G4QElasticCrossSection with option
0146 "false", so parameters for sampling of scattering are precalculated.
0147 + models/neutron_hp:
0148 o Fix against cases of division by zero.
0149 o Fixed erroneous condition in G4NeutronHPContAngularPar.
0150 o To increase performance, checking point of "neglecting doppler
0151 broadening" flag is changed from GetCrossSection() method to
0152 BuildPhysicsTable().
0153 + models/theo_high_energy:
0154 o Added missing Fermi momentum to quasi elastic scattering.
0155 + stopping:
0156 o G4MuMinusCaptureCascade: fixed case of NaN value in bound muon decay.
0157
0158 o Generic Processes:
0159 -----------------
0160 + Decay:
0161 o Added protection to properly handle shortlived particles with finite
0162 pre-assigned proper time.
0163 + Management:
0164 o Fixed bug related to invoking ResetNumberOfInteractionLengthLeft()
0165 at the zero length step
0166 o Allow zero mean free path for PostStepDoIt().
0167 + Parameterisation:
0168 o Cleared compilation warnings detected on gcc-4.3.0.
0169
0170 o Run:
0171 ---
0172 + Added G4Transportation to shortlived particles in G4VUserPhysicsList
0173 in order to fix problem of non-assigned process for shortlived particle
0174 decay detected in CMS.
0175
0176 o Examples:
0177 --------
0178 + advanced/underground_physics:
0179 o Fixed mis-sized arrays 'ring_phys_gas' and 'ring_phys_liq';
0180 both were less than the utilized size.
0181 + extended/field/field02:
0182 o Fixed setup to correctly handle electric field.
0183 + extended/field/field04:
0184 o Fix to avoid error for invalid rotation specification
0185 in F04DetectorConstruction::stringToRotationMatrix().
0186
0187
0188 ----------------------------------------------------------------------------
0189
0190 Technical Notes
0191 ---------------
0192
0193 o This patch should be applied on top of release 9.1
0194 o Technical notes distributed for release 9.1 are also applicable and
0195 valid for this patch.
0196
0197 The code and rebuilt binary libraries for release 9.1 are available
0198 through our "Download" Web page:
0199 http://cern.ch/geant4/support/download.shtml
0200
0201 Please refer to the Geant4 User Documentation:
0202 http://cern.ch/geant4/support/userdocuments.shtml
0203 for further information about using Geant4.
