1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250
|
MPQC REVISION HISTORY
Following is a brief summary of changes made in each release of MPQC.
* 2006-03-22: MPQC-2.3.1
o MCSearch added for cubic interpolation during quasi-Newton line
searches.
o Added KMLYP method.
o Updated libtool to version 1.5.22.
o SumDenFunctional returns correct HF exchange coefficient when using
nested ACM functionals. This could change the results from certain,
uncommon inputs.
o Other minor bug fixes, enhancements, and documentation improvements.
* 2005-10-16: MPQC-2.3.0
o Common Component Architecture interfaces for energies and
integrals are now available.
o Updated to a more recent version of the EMSL basis sets. This
will change results in many cases.
o Polarization consistent basis sets added.
o Test suite inputs and results have been moved from a separate
file into the main release.
o "make" targets are now available to automatically check results.
o "mpqcrun" can be used to simply starting MPQC runs, particulary
parallel runs.
o Complementary Auxiliary Basis Set (CABS) version of the MP2-R12
method [see Chem. Phys. Lett. 395, 190 (2004)].
o MP2-R12/A' method with split virtual space (virtual orbitals
constructed from a separate Gaussian basis).
o Novel Brillouin condition-free versions of the MP2-R12/A' method.
o A new exception infrastructure has been added.
o Numerous numerical improvements, compiler bug work-arounds,
and bug fixes.
o Checkpoint files made with earlier versions of MPQC utilizing
more than one processor will not work 2.3.0 or later versions.
* 2004-12-18: MPQC-2.2.3
o Documention errors and omissions corrected.
o Command man pages generated automatically.
o bison and flex are no longer required.
o Will now compile with MPICH2.
o Maintain symmetry in optimizations of large molecules.
o Improved ease of builds within the source directory.
o Fixed density value computations.
o Port to Cray X1 (correctness of results not checked).
o Port to G5/OS X (correctness of results not checked).
* 2004-04-28: MPQC-2.2.2
o Can now compile with GCC 3.4.0.
o Switched from ieee_{set,get}_fp_control to fenv functions.
o More architecture types recognized, including pentium4.
o The getnwbas.pl and parsenwbas.pl work with recent versions
of the EMSL Basis Set Database.
o Libtool flags are now accessible with sc-config.
o Psi 3 compatibility classes have been updated.
* 2003-12-26: MPQC-2.2.1
o Fixed a problem where FORTRAN files caused builds of shared
libraries to fail.
o Duplicate identifiers for AMD64 shared library builds have been fixed.
o The SONAME of this release is 5:1:0. GCC 3.2 is the official C++ ABI.
* 2003-12-04: MPQC-2.2.0
o (ABS) MP2-R12 method added (Edward Valeev).
o ARMCIMemoryGrp added.
o Better threading parallelism.
o The SONAME of this release is 5:0:0. GCC 3.2 is the official C++ ABI.
* 2003-09-26: MPQC-2.1.5
o Fixes for shared library support (Michael Banck)
o Minor documentation updates.
* 2003-07-22: MPQC-2.1.4
o Fixes a problem that occurs when optimizing with GCC 3.3.
o Now works with more recent versions of bison, flex, and autoconf.
o Upgraded libtool to 1.5 (this requires recent versions of
automake/autoconf).
o Bug in the charge computation for Extended Huckel is fixed.
o sc-config --cppflags should always give the correct include path now.
o The SONAME of this release is 4:1:0. GCC 3.2 is the official C++ ABI.
* 2003-01-10: MPQC-2.1.3
o Fixes a problem that occurred with recent versions of bison. Note
however that a new problem has emerged in bison 1.875 which is not
fixed in this release.
o Applied changes to make RPM's easier to build.
+ The installroot=path assignment can be given on the make command
line to specify a temporary install root.
+ --with-sc-includedir=dir can be used to specify an installation
directory for the include files.
* 2002-10-14: MPQC-2.1.2
o Bumped SONAME to 4:0:0. GCC 3.2 is the official C++ ABI for
4:0:0. See the --with-build-id configuration documentation for more
info.
o Remove use of MPI-2 1-sided as this has proved inadequate.
o Minor documentation updates.
* 2002-04-21: MPQC-2.1.1
o Support recent versions of bison and flex.
* 2002-04-18: MPQC-2.1.0
o Everything is now in the "sc" namespace.
o Fixed a problem causing redundant output lines with gcc3.
o Added extended Huckel theory guesses.
o Added configure options:
+ --enable-always-use-mpi
+ --with-default-parallel={mpi2,mtmpi}
o Begin conversion to use exceptions instead of abort.
o Improved MP2 gradient code performance.
o Greater than 2GB support for memory keywords.
o Symbolic notation for memory keyword supported. Examples: 2KB, 1MB,
1.3GB, 8KIB, 1MIB, 6GIB.
o Thread safety issues addressed in MessageGrp classes.
o Obsolete code removed.
o Fixed accuracy problem affecting mainly single point DFT gradients.
* 2002-03-01: MPQC-2.0.4
o Fixed a checkpoint/restart bug introduced in 2.0.2.
* 2002-02-17: MPQC-2.0.3
o Problems using newer versions of bison were fixed. Versions up to
1.33 should work.
o Avoid duplicate MemoryGrp constructor calls.
* 2002-01-24: MPQC-2.0.2
o Several configure problems were fixed.
o The keep_guess_wavefunction option was added. This is used to allow
the guess wavefunction to be reused for lower symmetry displacements
during a frequency calculation.
o Unrestricted wavefunctions will now correctly recompute occupations
when displaced into lower symmetry.
o A stack overflow for very large calculations was fixed.
* 2002-01-08: MPQC-2.0.1
o Incorporate some of Michael Banck's Debian porting work.
o Improve usability on parallel machines without pthreads.
o Minor portability enhancements.
* 2001-10-04: MPQC-2.0.0
o Add support for KAI KCC.
o Put template instantiations in libraries for Compaq C++.
* 2001-08-31: MPQC-2.0.beta.6 html build 2
o Inheritance and collaboration diagrams added to documentation.
* 2001-08-29: MPQC-2.0.beta.6
o Make sure MPI_Finalize gets called.
o Fixed scpr.
o Fixed parallel DFT printing.
o Added --enable-production configure option.
o Improved MP2 performance.
* 2001-07-10: MPQC-2.0.beta.5
o Added Gram-Schmidt orthogonalization of basis functions.
o Changed the following basis sets to use 5D instead of 6D:
6-311++G(2d,2p), 6-311++G(3df,3pd), 6-311++G**.
o Fixed bugs in MTMPIMemoryGrp exposed by MPI/Pro 1.6.3.
o Improved multi-threaded MP2 code.
* 2001-04-06: MPQC-2.0.beta.4
o Major API changes
+ Use RTTI for dynamic casts
+ Uses smart pointer templates instead of macros
+ Requires ISO standard C++ compiler
+ May require patched flex
o Clean up geometries when symmetry is specified
* 2001-03-08: MPQC-2.0.beta.3
o Object directory can be a subdirectory of the source directory.
o Documentation updates.
o --enable-stl fixed for standard C++ compilers.
* 2001-03-03: MPQC-2.0.beta.2
o New, simplified input format is supported in addition to the
objected-oriented input.
o Many improvements to the Density Functional Theory code. It is now
ready for production use.
o Documentation converted to doxygen.
o Supports ISO 14882 standard C++. MPQC now might not work with some
older compilers.
o Nonstandard symmetry frames now work.
o All HF and DFT objects can use multi-threading.
* 2001-02-25: MPQC-1.2.5
o "make install_devel" fixed for unified source and object dirs.
o Symmetry detection in presence of ghost atoms fixed.
o GCC 2.95.2 now can be used to compile MPQC. GCC 3.0 and above will
not work (this will be fixed in next major release).
o Nonstandard symmetry frames do not work. They are detected and will
cause an abort (this will be fixed in next major release).
o "make -j2" problems fixed.
* 2000-09-07: MPQC-1.2.4
o MTMPIMessageGrp will be linked into MPQC if threads and MPI are
available.
* 2000-01-31: MPQC-1.2.3
o install_devel will now install scconfig.h
* 1999-11-04: MPQC-1.2.2
o Turned off core dumps. These could cause problems on the IBM SP.
* 1999-10-24: MPQC-1.2.1
o Fix minor problem preventing a smooth compile on AIX.
* 1999-10-20: MPQC-1.2
o The 1.1 input files will not work with 1.2
o Ported to IRIX and IRIX64
o Documentation improvements
o "make interface" no longer needed
* 1999-08-19: MPQC-1.2alpha6
o Use DOC++ for all documentation
o Bug fixes and ports to glibc 2.1 and gcc 2.95
o Canonical orthogonalization option
* 1998-10-23: MPQC-1.2alpha5
o Multi-threaded CLHF and MP2 gradients
o Removed all use of libg++
* 1998-05-25: MPQC-1.2alpha4
o Save/restore and result accuracy bug fixes
o Reducing memory requirement for integrals
o configure improvements
* 1998-04-03: MPQC-1.2alpha3
o Faster integrals with arbitrary angular momentum
o Finite displacement Hessian can be used as guess Hessian
o Checkpoint override capability added
* 1997-10-30: MPQC-1.2alpha2
o Fix a bug affecting optimizations
* 1997-10-29: MPQC-1.2alpha1
o Checkpoint files have directory information that can be used to
selectively restore objects
o Class hierarchy changed under SCF
o The molecule class was cleaned up
o Many other bug fixes and enhancements
* 1997-6-11: MPQC-1.1.2
o Fixed SYSV IPC configuration problem on sgi-irix platforms.
o Fixed sample inputs in HTML user manual.
* 1997-5-23: MPQC-1.1.1
o GNU-CC specific construct removed.
* 1997-5-22: MPQC-1.1.0
|