Version 0.3.2 o Bugfix in Taylor-Arithmetic o Fixed some large memory leaks o Added rand(n,m) for an n*m random matrix o Added gnuplot postscript output o Added 2d parametric plot pplot[...] o Added 3d surface plot splot[...] o Added matrix conjugation o Added abs() and relational operators to taylor-arithmetic o Changed readline syntax again for 4.2 (when are they done??) o Switched to C-XSC 2.0 beta o Preliminary parser fix thus "2* 3+5" results in 11 not 16 o n-ary predefined functions may now also accessed as f(,,,..) o Introduced xsc-toolbox modules rpeval, cpzero, linsys, linopt o Introduced polynomial arithmetics o added mid,inf,sup,re,im to matrix&poly, mid,inf,sup to complex o Made standalone version "valexample" work again Version 0.3.1 o Fixed "longinteger^(-longinteger)"-bug (thanks nam!) o Fixed "(-longinteger)^(longinteger)"-bug o Added tests for above cases o Use of "persist" option in gnuplot-module Version 0.3.0 o Added desktop menu entries o User defined functions now may have multiple variables o Experimental gradient support o Added matrix creation functions zeros(,) and ones(,) Version 0.2.9 o Initial release frontend "grascal" o Added colour support o Added "clear" command o Initial purenum support Version 0.2.8 o Added more documentation to online help o Removed (reappeared) matrix-complex type predence bug o Added generic type predence tests o Improved plotting support (no more fraction as input for gnuplot) o Vector valued functions are plotted as several lines Version 0.2.7 o Improved testing environment o Added constants "version" and "configuration" o Parsevalue statistics o Included online help environment o Added help topics for most modules o Added module list command 'modules' Version 0.2.6 o Long integer fraction now honor SIGINT signal o Removed bugs in power(taylor,int) and in sqr(taylor) functions o Derivatives of functions can now be evaluated, like f'(2) o Precision(*,-n) forces output to be most accurate up to n decimals Version 0.2.5 o Improved man(1) page o Improved support for stand alone c++-class value o Removed "complex multiplication results in string" in full-version bug o Removed staggered arithmetics power bug o Increased speed and decreased memory demands during compile o Compiler optimization now possible, but high memory demands o Added some more tests Version 0.2.4 o Added transpose operator on functions for derivation (try entering cos') o Added "rows" "cols" and "size" functions for matrices and vectors o Added "strlen" function for strings o Added some more tests o Parsevalue output now in "parsevalue.log" o Added man(1) page Version 0.2.3 o Corrected minor bugs in forwarded operators in generated value.cpp o Added experimental "for(...;...;...)" language feature o Rascal now honors SIGINT (ctrl-c) signal Version 0.2.2 o Added ShowPrecision query procedure o Added Precision function for light, medium, long versions o Exceptions are now being caught o Removed compiler option that was problematic with egcs-versions o Built-in functions do not appear twice in "functions" any more Version 0.2.1 o Added "variables" and "functions" procedures o Updated modules creation documentation o Floating point number string now converted directly to c-xsc datatypes o Added interval compares o Corrected some tests Version 0.2.0 o Added complex compares o Corrected result types of string-compares o Added abs(LReal) o Added old readline library compatibility o Added automated test environment o Added sqr(complex) o Command-line options o rc-files o Using %g printf output for generic double output o Corrected 1/0 bug in long integers Version 0.1.9 o New more accurate implicit cast determination method o Implicit cast from long integer and fraction now correctly defaults to double and not to string! Version 0.1.8 o Long integer module added o Long integer fraction module added o Double and integer sqrt and pow-functions now overridden by complex o Complex number output now uses units i or j if defined o Numbers like "2." are now also being accepted as floating point numbers o Parsing of numbers (integer and floating point) may now be overridden o Introduced functions det(matrix) and inv(matrix) o Additional spaces in function definition like "f(x) =" now ignored Version 0.1.7 o Added output format control to interval module o Overriding unary functions now possible Version 0.1.6 o All functions are now values that can be evaluated, enabling function as variable features o Experimental plotting support via gnuplot o Added complete pow-support to taylor arithmetic Version 0.1.5 o Added all standard functions (incl. hyperbolic) to taylor arithmetic o Corrected XSC-urls in documentation Version 0.1.4 o Alpha-support of C-XSC interval, l_real, l_interval datatypes o Experimental functional support Version 0.1.3 o Operator concept overhaul o Readline support o Procedures o Help o rpm Version 0.1.2 o Added hyperbolic functions and inverse for double o Added standard functions for complex Version 0.1.1 o Matrix inversion now prevents division by zero o Minor documentation improvements o Complex comparism operators Version 0.1 Initial Release