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
|
