The following improvements have been added between Version 4.0.3 and 4.0.4:

issue000122 Improvements to Monte-Carlo capabilities, as required in Physor-2010 paper
issue000123 Correct delayed neutron data processing in dragr
issue000124 Implement the possibility of many fissionable isotopes in Trivac
issue000125 Correct badly behaved correlated weight matrices in USS:
issue000126 Use the lumped depletion chain in EDIMIC instead of the complete one
issue000127 Bring NJOY to 99.304 level
issue000128 Correct WR approximation in USS: with SUBG probability tables
issue000129 Bring Version4 to 3.06G level
issue000130 Correct a bug in the Saphyb generation system
issue000131 Correct the identification of fissionable isotopes in LIBMIX.f and recover H-FACTOR in COMPO: with MACR option
issue000132 Set NCOR=1 in EXCELT hexagonal trackings
issue000133 Correct a bug in SPN/Raviart-Thomas-Schneider method detected with Intel compiler
issue000134 Improve Donjon capabilities for the simulation of PWRs
issue000135 Correct the calculation of power factors (H-FACTOR)
issue000136 Set NOLJ as default in SHI: and correct module CRE:
issue000137 Correct Monte-Carlo capabilities in 2D and 3D
issue000138 Implement boundary condition ACYL in Trivac and Donjon
issue000139 Correct edition in MAC: when iprint>3
issue000140 Correct a regression in ISOTXS file generation
issue000141 Modifications related to micro-depletion in Donjon
issue000142 Introduce coherent elastic scattering in Draglibs made with PyNjoy
issue000143 Add two more non-regressions tests in hexagonal geometry
issue000144 Add the capability to force a specific isotope to be at saturation
issue000145 Correct a convergence problem with the subgroup projection method
issue000146 Include a non-regression test for TINST: module in Donjon
issue000147 Correct a segmentation fault in Cartesian SPN case
issue000148 Correct an instability with physical probability table generation
issue000149 Implement bundle refuelling for CANDU cases in Donjon4
issue000150 Extension of modules USPLIT:, RESINI: and FLPOW: to hexagonal prismatic geometries
issue000151 Modernization of space-time kinetics in Trivac
issue000152 Correct a bug related to depleting chains in dragr
issue000153 Correct the homogenization of Microlibs in module EDI:
issue000154 Bring Version4 to 3.06H level
issue000155 Correct a bug in situations where cross sections are coming from both Multicompo and Draglib
issue000156 Initial support of script instver4.bat for building Version4 on MS-DOS
issue000157 New modules DETINI: and DETECT: for extracting flux at detector locations
issue000158 Correct a regression with nup=0 in EXCELT:
issue000159 Correct free format reader to accomodate SIMEX-HQ
issue000160 Enable 64-bit build and MPI capabilities
issue000161 Bring NJOY to 99.336 and upnea070_up336 level
issue000162 Correct a bug in the isotope lumping algorithm for depletion
issue000163 Correct the identification of depleting mixtures in NCR:
issue000164 Correct a regression in issue000155
issue000165 Add a missing LCMSIX in hexagonal EXCELT: module
issue000166 Initial support of the Snow Leopard operating system
issue000167 Implement micro-depletion capabilities in DRAGON and DONJON
issue000168 Update PyNjoy datasets for recent evaluations
issue000169 Move and rename two LCM API calls
issue000170 Increase the maximum number of power iterations in EVO:
issue000171 Correct the CHI record of zero-burnup residual macroscopic isotope
issue000172 Use /usr/tmp/ as working directory on AIX
issue000173 Correct two errors in the documentation (including the readme file)
issue000174 A buckling correction is applied more than once
issue000175 Continue implementation of micro-depletion capabilities
issue000176 Add more history-based and MPI capabilities