summarylogtreecommitdiffstats
path: root/arprec.changelog
blob: 541c07b15098d8b1b494cf2cbccfd3318931cf29 (plain)
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
Changes for 2.2.19
  - Updated with a fix to write.cpp.

Changes for 2.2.18
  - Updated with changes for Intel compiler v15-16.

Changes for 2.2.17
  - Added an input parameter mpstrlen to mp_inp.f.

Changes for 2.2.15
  - Fixed intermittent memory problem in dotd.cpp.

Changes for 2.2.6
  - Fixed F-format bug in Fortran file mp_mod.f.
  - Fixed mpread bug in Fortran file mp_modx.f

Changes for 2.2.2
  - Clean up mp_mod.f, added mpeq_xz (assignment to complex)
    and cmplx (creation of mp_complex).
  - Fix dble to be more accurate, and avoid spurious underflow.
  - Make C++ I/O of mp_real closer to standard C++ I/O.
  - Added to_string function.

Changes for 2.2.1
  - Some bug fixes for Cygwin environment.

Changes for 2.2.0
This is a major update to ARPREC, there are several API incompatibilities.
  - Moved C++ main entry in libqdmod.a to libarprec_f_main.a.
    This allows to link Fortran code using ARPREC with custom C++
    main function.  Pure Fortran code will need to be linked with
    arprec_f_main library in addition to arprecmod and arprec library.
  - Constructor taking integers now just initializes the data.
    For example "x = mp_real(35)" now initializes x to 35, instead
    of creating an mp_real with 35 words.  To get the old behaviour,
    do something like mp_real(0.0, 35).
  - String output routines write and to_string routines now takes a single
    fmtflags.
  - Members are now all public, making it a plain-old-data type.
  - Removed mpr_location, set/getData accessors.
  - mp_real(double *) constructors are now explicit.
  - Renamed several static variables in class mp to more descriptive names.
  - Various constants are now in mp_real class (mp_real::_pi, _log2,
    _log10, and _eps).

Changes for 2.1.104
  - Fixed bug in division and increments for mp_int.
  - Output correct number of digits in c_mpwrite / c_mpout.
  - Fixed compilation errors and warnings.

Changes for 2.1.103
  - Minor changes to accomodate Windows build.

Changes for 2.1.102
  - Ignore CC compiler on Apple systems.
  - Fix C++ pslq demos to print out integers when appropriate.
  - Fix dependency of Fortran modules to allow parallel make.
  - Fix bug in mp_real::mpnorm that was accessing past end of an array.

Changes for 2.1.101
  - Added third display option in math toolkit.
  - Bug fix in math toolkit.

Changes for 2.1.100
  - Added fortran/Makefile.sample, a sample Makefile for using ARPREC
    library with Fortran programs.
  - Fixed bug in arprec-config where --fcflags had an extra @.

Changes for 2.1.99
  - Added read_binary and write_binary for binary I/O.
  - Fixed bug where decimal point and minus sign was not printed correctly 
    in certain cases.
  - General improvements in I/O routines, including adjustable exponent 
    width and column width.
  - Added mp_real constructor accepting std::string.
  - Perform some I/O tests during "make check".

Changes for 2.2.3
  - Removed dynamic precision from quadgs routine of Fortran and C++
    versions of the Gaussian quadrature demonstration program.
  - Fixed some bugs in the erf and erfc routines.