6 * Additional float constants: LEAST-NEGATIVE-NORMALIZED-FLONUM,
7 largest_negative_float, least_negative_float,
8 least_positive_normalized_float, and least_negative_normalized_float.
9 * Support for pre/post sub/superscripts (get_index_properties,
10 declare_index_properties and remove_index_properties).
11 (2) Make declare_index_properties
12 * Mark non-trivial results from gcfactor with a GCFACTORED flag
16 * pytranslate, which converts maxima to python
18 * descriptive: find_runs and find_runs_inverse
22 * Code clean ups and removal of functions that only had a historical
24 * In TEST-BATCH, ensure that list of unexpectedly passed tests is
25 nonempty before reporting it.
26 * :lisp-quiet can now be told not to use subsequent :lisp commands
27 * defmfun now adds an ARG-LIST property that allows frontends to
29 * An improved version of float_approx_equal
30 * Better handling of non-numerical input to find_root
31 * Better handling of defint errors triggered by laplace()
32 * Avoid variable capture in the WITH-NEW-CONTEXT macro
33 * Evaluate the body of a WITH-NEW-CONTEXT form as an implicit PROGN
34 * Numbers are no loger compared using EQ instead of EQL
35 * Characters are no loger compared using EQ instead of EQL
36 * Fixed bogus translations of apply2 and applyb2
37 * Fixed bogus translations of atan and atan2 with float arguments
38 * Fixed the bogus translation of atan2 with one float and one rational
40 * Fixed bogus and inconsistent translations of max and min when the
41 arguments were a mix of float and other modes
42 * Fixed an internal error in the translator when attempting to convert
43 a variable of mode rational to a float
44 * Fixed the inconsistent translation of signum with a float argument
45 * Fixed the translations of log and sqrt so they honor the flag
46 tr_float_can_branch_complex
47 * Updated gentran to v.5
48 * Improved the pattern matcher for "+" and "*" expressions, see
49 maxima-discuss, circa 2019-12-27: "Function to recognise series"
50 * allow zn_log to work on subgroups
51 * gamma_incomplete_lower is now returned in the noun form in many places
52 * operatorp(expr,op) was inconsistent with operatorp(expr,[op])
53 * sqrtdenest is now built-in, not a share function.
54 * load(sqdnst) now is a no-op.
58 * Interpol used funmake instead of apply
59 * An updated version of COMA
60 * A testbench for draw()
61 * plot2d with a single function now uses xlabel
62 * Many minor corrections
63 * In package amatrix, protect against empty matrices when converting
64 amatrix to matrix and vice versa
65 * Fixed contragrad in ctensor, added tests for cograd/contragrad
69 * #3412: Corrected the translation of conditionals
70 * #3587, +3489: Better handling of limits of logarithms
71 * #3521: In simplify_sum, avoid double-counting -min_ni term in
73 * #3265: gcfactor(0) -> division by zero
74 * #2839: gcfactor(9) => 9
75 * #3287: Cross product of scalar zero versus vector yields zero result
76 * #3583: Stack overflow for equality testing with assumptions
77 * #3608: logic_simplify handles inequalities incorrectly
78 * #3607: printprops displays matchdeclare properties incorrectly when value is a lambda
79 * #2174: Bogus translation of declare
80 * #3513: Emacs: wxdraw3d not working
84 * plot now works correctly with function names containing _, ^,
86 * Maxima now no longer becoms unusable with errset=t
87 * FREE-LISP-VARS property for LET was broken
88 * translator: improved error message when a throw is not in a catch
89 * PATCH-UP-MEVAL-IN-FSET had a bogus BARFO call
90 * *standard-input* was broking during gentranin runs
91 * Problem loading abs_integrate in imaxima mode (emacs)
92 * compgrind:true caused compfile to not write to the output file
93 * compgrind:true didn't ensure pretty-printing of compfile output
94 * Taylor expansion for gamma_incomplete_lower(1/2,z) now works
95 * compgrind:true didn't ensure pretty-printing of compfile output
96 * gamma_incomplete_lower now respects gamma_expand
98 Additional improvements:
99 ------------------------
100 * The documentation now is build use a stock makeinfo
101 * Many other improvements to the way documentation is generated
102 * The RPM files now end up in the build directory
103 * A way to run many draw() and plot() examples for manual tests
104 * More out-of-source-build improvements
105 * "make dvi" was subject to bit-rot and should now work again
106 * Resolved a few cases in which common variable names in
107 arguments caused unexpected behavior
108 * Better ABCL integration
109 * Added a help message (Option: /?) to the windows installer
110 * Fix errors when installing on MacOS via Homebrew
111 * run_testsuite(share_tests=only) now doesn't fail due to
112 unicode-sniffer.lisp not having been loaded
113 * Draw now translates the coordinates of labels, too
114 * Use DLMF 8.8.9 to expand gamma_incomplete(a+n,z)
115 * Expand gamma_incomplete for rational order
117 * Many improvements to the documentation
118 * Windows Installer: Updated wxMaxima to the new release.