- update to v4.9.0:
  * Support for specifying a fixed denominator for display of
    fractions (e.g. "0.3 ft ➞ 1/8 ≈ (3 + 5/8) in")
  * Return gcd of numerators divided by lcm of denominators in
    gcd() with non-integer rational numbers, and vice versa for
    lcm()
  * Add units for mean Gregorian and tropical years
  * Ignore underscore in number
  * Replace defunct exchange rates source and fix bitcoin
    exchange rate
  * Fix asin(x)=a, acos(x)=a, and atan(x)=a, when a contains an
    angle unit and default angle unit is set
  * Fix output of value converted to unit expression with
    numerical multiplier in denominator, e.g. "➞ L/(100 km)"
  * Fix segfault when trying to solve
    "(xsqrt(x)-ysqrt(y))/(sqrt(x)-sqrt(y))=x+sqrt(x*y)+y"
  * Fix parsing of case insensitive object name ending with
    Unicode character when followed by another Unicode character
    in expression, e.g. "микрометр"
  * Add history command, listing expression history
  * Display all exponents 0-9 using Unicode superscript
    characters if these are the only exponents in the expression

• Files Changed
• Browse Source

- version update to 4.8.1
  *  Fix besselj() and bessely()
  *  Fix display of matrices and vectors with complex number form other than
     rectangular (default)
  *  Fix conflict between decibel units (e.g. "0dB") and duodecimal 0d… syntax
     (now requires at least two digits)
  *  Fixes for title options in plot() function
  *  Add additional buttons and button menus to general keypad (Qt)
  *  Add option to show all functions, units, and variables in tool button
     menus (Qt)
  *  Add recently used objects to functions, units, and variables menus (Qt)
  *  Add option to bypass dialog for functions in menu (Qt)
  *  Add option, in context menu, to display text for tool buttons (Qt)
  *  Open functions, units, and variables tool button menus using long-press
     or right-click (in addition to arrow button), and add keyboard shortcuts (Qt)
  *  Use keypad tool button menu as context menu and show keypad type in
     title bar of keypad (Qt)
  *  Auto-update number bases above programming keypad when entering simple
     integers, even if calculate-as-you-type is deactivated (GTK)

• Files Changed
• Browse Source

Line breaks in .changes.

• Files Changed
• Browse Source

- version update to 4.8.0
  * Concise (e.g. "1.2345(67)") and relative (e.g. "5+/-2%") notation for output
    and input (disabled by default for concise notation) of interval
  * Optimize rest of unit expression when converting to partial (insufficient)
    expression (e.g. 18MA×100ns to J = 1.8 J/V)
  * Round high precision intervals enclosing integers for integer arguments
  * Add support for specifying position of sign bit in floating point functions
  * New derived units: Debye, C·m, Daraf, F^-1, V/C, N/C, kcal, kcal/g, kJ
  * Use international foot as base unit for rod
  * Fix fibonacci() for non-integers and recursive use
  * Fix conversion options presented for speed units (and possibly other)
  * Update dialogs for editing of variables, functions, units, etc. to match
    the corresponding dialogs in the Qt UI (GTK)
  * Show user-defined objects in a separate category in menus and dialogs (GTK)
  * Add option to disable tooltips in keypad or everywhere (GTK, Qt)
  * Add option to disable units and to interpret unrecognized symbols as
    variables (unknowns) to preferences (Qt)
  * Make closing of application with Escape key optional and disabled by default (GTK, Qt)
  * Hide name column in conversion list when applied to current result (Qt)
  * Add temporary option to unknown variable edit dialog (GTK, Qt)
  * Add Microship (PIC16/17) 24 and 32 bit floating point formats (GTK, Qt)
  * Fix initial height of list and description in units window (GTK)
  * Minor bug fixes and feature enhancements

• Files Changed
• Browse Source

- version update to 4.7.0
  * Support for custom default angle unit, e.g. turn, arcsec, arcmin
  * Append default angle unit (instead of always radians) when converting
    value without unit to angle unit
  * More consistent addition and removal of angle unit from function arguments
  * Always interpret ./, .*, and .^ as entrywise operators if user intention is unclear
  * Change order of operations to place entrywise and ordinary operators on
    the same precedence level
  * Add function, kron(), for Kronecker product, and constants for Pauli matrices
  * Add radius to planets dataset and update other properties
  * Support replacement of unknown variables within variable values
  * Fix besselj(0, 0)
  * Fix incomplete calculation in tan() with try exact approximation
  * Fix 0/0=0 equality (do not return true) and output of 2/0 (and similar)
  * Fixes and improvements for newtonsolve() and secantsolve()
  * Fix segfault when MathStructure is deleted after Calculator, and in destructor
    of calculated DynamicVariable (called from Calculator destructor)
  * Do not save mode on exit if "-defaults" command line switch where used (CLI)
  * Allow multiple actions for keyboard shortcuts (GTK, Qt)
  * Add toggle precision, and min, max, or min and max decimals to available
    shortcut and button actions (GTK, Qt)
  * Add option to exclude units for unformatted ASCII copy (GTK, Qt)
  * Add optional value to copy result action, allowing expression copy and
    formatting selection (GTK, Qt)
  * Fix copy unformatted ASCII when local digit group separator is same as selected
    decimal separator (GTK, Qt)
  * Add option to automatically copy result (Qt)
  * Always set (primary) selection clipboard contents when whole expression is
    selected or selection is cleared, e.g. after calculation (Qt)
  * Improve support dark mode and high contrast modes, and change default style

• Files Changed
• Browse Source

- Update to 4.6.1:
  * Do not automatically by default set prefix higher than kilo for meter,
    gram, higher than 1 (no prefix) for second and litre, or lower than 1
    for byte and bit
  * Do not use recently introduced SI prefixes (R, Q, r, q)
    automatically by default
  * Do not set automatic prefix if resulting multiplier is
    less than 1 or greater than 1000, with some exceptions (e.g. km)
  * Default prefix unit property (use for gram, and few other units,
    instead of hard coded value)
  * Option to change symbols used for output of digits 10 and 11 in
    duodecimal numbers (use A and B as default digits)
  * Allow the same variable left and right of the equals sign (e.g. var=var+2),
    in variable assignment without colon, if the variable exists and the
    previous value is known
  * Use Unicode symbol for minus in scientific e notation
  * Do not show multiplier if exactly 1, in HTML output of scientific notation
  * Prefer Nm (torque) over J when the result is a vector with three components
  * Add "amp" abbreviation to ampere unit, and "thou" name for 1/1000 in (mil)
  * Fix unit order for Nm and Ws
  * Fix conversion from ounce (interpreted as fluid ounce) to litre with prefix
  * Fix automatic reactivation of global object after deletion of conflicting object
  * Fix parsing of scientific e notation when the number is extremely large and
    exponentiation fails because of floating point overflow
  * Fixes for output of scientific notation using number bases other than decimal
  * Fix prefix selection in denominator when multiplier is higher than the
    value of the largest prefix
  * Fix segfault in multisolve()
  * Don't show anything on empty input
  * Support "help [OPTION]" and "help set [OPTION]" command to display

• Files Changed
• Browse Source

- update to 4.5.1:
  * Fix retrieval of exchange rates from mycurrency.net and add alternative
    secondary exchange rates sources
  * Add m/s³, m/s⁴, m/s⁵, and m/s⁶ as derived units (improves automatic unit
    conversion)
  * Check that the output of command, in command() function, is a reasonably
    sane expression
  * EUR has replaced HRK as currency of Croatia
  * Fix MathStructure::countTotalChildren() with count_function_as_one set
    to false
  * Fix unit tests

• Files Changed
• Browse Source

- Update to 4.5.0:
  * Conversion to any type of expression
    (e.g. "1000 BTC to million USD")
  * Use current status of limit implicit multiplication setting
    during parsing of conversion expression
  * New SI prefixes (R, r, Q, q)
  * New units: ounce force (ozf), rack unit (U)
  * Use standard form (sorting) for polynomials
  * Improve consistency of behaviour when global units are replaced
  * More consistent addition of units to expression without units
    for unit conversion
  * Improvements and fixes for arg() and atan2() functions
  * Fix interpretation of Myr (megayear) and Gyr (gigayear)
  * Fix milli- and centimeter of water units
  * Fix genvector() with for negative step size or expression
    as step size/number of steps
  * Fix randuniform() function
  * Fix automatic/dual fraction format after conversion
    (e.g. "8 to score")
  * Fix inconsistently displayed parsing of "7d 5h" and similar
  * Minor bug fixes and feature enhancements

• Files Changed
• Browse Source

- Update to 4.4.0:
  * New prime number functions: primePi(), primes(), nthprime(),
    nextprime(), prevprime(), and isprime()
  * New functions for intervals and uncertainties: upperEndpoint(),
    lowerEndpoint(), midpoint(), and errorPart()
  * Add meter of water and related pressure units, and more
    derived flow rate units 
  * Support for vectors and matrices in uncertainty propagation
    calculation
  * Proper support for uncertainty propagation in total(), sum(),
    product(), and for() functions
  * Simplify display of parsed expression (primarily for quantities
    with units)
  * Avoid parsing of implicit multiplication as nested
    multiplication
  * Include +/- statement in function argument without parentheses
  * Return values for approximate variables in exact mode in
    dimension() and elements()
  * Improve parsing of quotation marks as feet/inch or
    arcmin/arcsec (e.g. do not parse s" as sin)
  * Fix exact calculation of re((a*i)^(n/2))
    (wrong sign in some cases)
  * Fix erroneous calculation in some cases with physical constants
    both outside and inside unit relations when interval
    calculation is set to interval arithmetic
  * Fix segfaults in some corner cases
  * Fix use of random numbers in iterative functions 
  * Fix writing beyond end of array during exchange rates check
    in some cases (results in activation of old style matrices)
  * Fix automatic reactivation of deactivated global object

• Files Changed
• Browse Source

- Update to 4.3.0:
  * Fix handling of Unicode powers for units in denominator, with
    adaptive parsing enabled (e.g. parse 10m/s² the same as 10m/s^2)
  * Fix "+" ignored after "E" in number bases where "E" is a digit
  * Fix scientific E notation with sign in argument when function
    is used without parentheses
  * Fix lambertw() for values very close to zero
  * Fix a × b^x + cx = d when a and c have different signs and d
    is non-zero
  * Fix a^x × b^x = c when a and b is negative, and c is positive
  * Fix segfaults in some corner cases
  * Fix potential issues in handling of leap seconds (e.g. during
    subtraction of seconds from date)
  * var=a syntax for variable assignment with calculated expression
  * Replace ounce with fluid ounce during conversion to volume unit
  * Solve a^x + b^x + … = c in more cases
  * Improve remainder/modulus for numerators with large exponents
  * Truncate number in output of parsed expression and end with
    ellipsis if unable to display all decimals
  * Improved floating point calculation and output speed,
    particularly for simple expressions with very high precision
  * New functions: clip(), qFormat(), qError()
  * "clear history" command and option to clear (not save) history
    on exit (CLI)
  * Replace selection (instead of wrap in parentheses) on operator
    input if selection ends with operator (GTK, Qt)
  * Act as if two arguments are required when applying base-N
    logarithm to expression (GTK, Qt)
  * When applying function to expression, exclude to/where
    expression and place cursor before closing parenthesis if
    operator is last in selection (GTK)
  * Show padlock (or "[P]") after protected expression (Qt)
  * Fix name field not working in argument edit dialog (Qt)
  * Minor bug fixes and feature enhancements

• Files Changed
• Browse Source

- Update to 4.2.0:
  - Option to plot real and imaginary parts separately (enabled
    by default)
  - Variable spacing of x values in plot (unless step size is used)
  - Updated plot() function with arguments for specification of
    plot options (in any order after the third argument)
  - Polar plot style
  - Ask the user for the desired sinc() variant (unnormalized or
    normalized)
  - Support for binary-coded decimals (BCD)
  - Show names with underscore capitalized and with underscore
    removed (with some exceptions)
  - Additional information units
  - Units for volumetric flow rate
  - Tradtional numbers (e.g. dozen, score)
  - Show warning when log() is used with a single argument, and
    always show second argument of log()
  - Show single digit suffixes using Unicode subscript (CLI)
  - Make names with a single character followed by digits
    case-sensitive with suffix by default
  - Accept more than two arguments in gcd() and lcm()
  - Interpret a single vector argument as a list of arguments if
    function requires more than one argument
  - Fix limit of floor(), ceil(), and trunc()
  - Fix plotting of derivatives and integrals
  - Fix randpoisson()
  - Fix reactivation of object, after object that caused
    deactivation is removed
- Update to 4.1.1:
  - Restrict use of reciprocal in unit conversion, e.g. invert in
    "5m/s to s/m" (= 0.2 s/m) but not in "500/lb to g" (= 1.1/g)
  - New mHg unit (allows use with prefixes other than milli)
  - Always display abbreviated unit name (symbol) with
    abbreviated prefix (when unit only has short name)
  - Minor improvements to automatic unit conversion
  - Fix display of (1 + f(x))/n
  - Save definitions on exit (only) if definitions has changed
    regardless of interactive mode status (qalc)
- Update to 4.1:
  - Use non-breaking narrow space as digit separator in
    reasonably short numbers, and between quantity and unit
  - Update value of austronomical unit
  - Fix segfault (when compiled with -D_GLIBCXX_ASSERTIONS) with
    local digit grouping when separator is empty (e.g. with
    ignore locale)
  - Fix freezes and crashes with extrememly large vectors and
    matrices
  - Fix segfault (when compiled with -D_GLIBCXX_ASSERTIONS) in
    matrix multiplication
  - Minor bug fixes and feature improvements (see the ChangeLog
    for more details)
- Update to 4.0:
  - New matrix syntax (column separated by space or comma, rows
    separated by semicolon, as in matlab)
  - Do not make any difference between a vector and a matrix with
    a single row, or between a scalar and matrix with a single
    element
  - Support input of functions, units, and variables with
    underscores omitted in names (e.g. µ0 = µ_0, FermiCoupling
    = fermi_coupling), with some exceptions
  - Option to disable "simplified" percentage calculations (e.g.
    interpret 100 + 20% as 100 + 0.2 instead of 100 × 120%)
  - Option to truncate numbers (instead of round to nearest) in
    result
  - Convert K, Hz, and 1/m to J (and derived units) using
    conversion factor constants (e.g. 1/(550nm) to eV ≈ 2.25
    eV)
  - Allow use of M as 10^6 without unit
  - New ksi unit
  - New functions: divisors(), rate(), vertcat(), horzcat()
  - New plastic number (ρ) constant, and new "hbar" name for
    reduced Planck constant, and add second Unicode codepoint for
    µ-sign in constants
  - Allow zero (1st and/or 2nd argument) in comb() and perm(),
    and return zero if second argument (k) is greater than the
    first (n)
  - Calculate statistical functions (e.g. mode(), min(), max(),
    percentile()) separately for each column of matrix
  - Return the lowest value if there are multiple values with
    equal frequency in mode()
  - Parse text argument as expression if concatenate function
    name is found in text string, and fix concatenate returning
    date
  - New layout for output of result of matrices and equations
    (qalc)
  - Fix Unicode handling on platforms where char is not signed
  - Fix bitrot()
  - Fix segfault in multisolve() when initial order of equations
    is not optimal
  - Fix segfault with empty sqrt inside sqrt, e.g. sqrt(1+sqrt())
  - Fix incompletely calculated nounit(), e.g. in unit conversion
- Update to 3.22.0:
  - New functions: linearfit(), quadraticfit(), cubicfit(),
    ramlatency(), parallel()
  - Merged inv() and inverse() functions
  - Improved and extended parallel operator (|| is interpreted as
    parallel if units are used)
  - Allow nested subfunctions
  - Do not stop testing function argument values after first failure
  - Solve x^(x^(-a))=b
  - Improved simplification: Im(-x)=-Im(x), Re(-x)=-Re(x)
  - Handle units in plot data (remove and set as y-axis label)
  - Avoid precision loss on output of interval when saving
    approximate variable
  - Fix pearson() and spearman()
  - Fix display of incompletely solved equation with dual
    approximation in some cases
  - Fix genvector() when step size requires evaluation
  - Fix a%%-b (interpret %% as mod, not percent)
  - Fix asin(x)+bcos(x)=c
  - Minor bug fixes and feature improvements
  - [qacalc] Support "--" as end of options
  - [qacalc] New --defaults command line option
  - [qacalc] Configurable SIGINT (Ctrl+C) action (exit orderly by
    default)
  - [qacalc] Vertical space option (vspace)
  - [qacalc] Clear command with Ctrl+L shortcut
  - [qacalc] Apply same logic with separate to command, as with "to" at
    end of expression, when result contains no unit
  - [qacalc] Fix "to unicode"
- Update to 3.21.0:
  - Ask (once) for correct interpretation of ambiguous implicit
    multiplication
  - Fix crash when using the dock menu to show the RPN dock stack
    for the first time
  - Fix link in new version dialog
  - Fix parenthesis highlighting when cursor is at end of
    expression
  - Fix initial value of interval display in preferences dialog
  - Fix saving of approximation mode
  - Improved simplification of radical expressions
  - Improved Im(), Re(), and arg() functions
  - Improved output of complex forms (other than the default
    rectangular) in exact mode
  - Function for drill bit size conversion
  - Fixes for equations with complex inverse trigonometric and
    hyperbolic functions with no solution
  - Fixes for complex x^(ax)=b

• Files Changed
• Browse Source

- Update to 3.20.1:
  * Save answer to dot and temperature questions in non-interactive
    mode and do not show the questions in non-interactive terse mode
  * Fix saving of sub functions with default values
  * Add support for function creation in save() function and with
    associated operator (e.g. func1():=x*y^2)
  * Add operators (.', .*, ./, .^) and functions (replace multiply(),
    divide(), and pow()) for transpose and element-wise multiplication,
    division, and power
  * Fix norm() and magnitude(), and add second argument, p,
    to norm() function
  * New text operators: dot, cross, comb, and perm
  * Add support for ^^ as XOR operator
  * Fixes and improvements in multisolve() function
  * Always return exact result in base() function, in exact mode
  * Improve parsing of functions without parentheses
  * Do not try to load data objects from global definitions dir
    if data set is local
  * Interpret "±" at end of expression as "±0"
  * Convert Celsius to Fahrenheit if no calculation was performed
  * Add support for QALCULATE_USER_DIR environment variable and always
    check for "user" directory on Windows
  * Output both original and factorized integer, and output abs(x) as |x|
  * Improved HTML output of mathematical expressions in libqalculate
  * Minor bug fixes and feature improvements

• Files Changed
• Browse Source

Update to 3.19.0

• Files Changed
• Browse Source

------------------------------------------------------------------
- Update to 3.18.0:
  * Improve handling of expressions with log-based units
  * Improve conversion of expression with multiple units
    to a single unit (e.g. m/W to W)
  * Output (kilo)gram instead of tonne with small prefix
  * New functions for statistical distributions: probit(), betadist(),
    cauchydist(), chisqdist(), expinv(), fdist(), gammadist(),
    tdist(), wblinv(), weibulldist()
  * Improve functions for statistical distribution
  * Improve sexagesimal input and output, and add geodistance()
    for calculation of distance between two GPS coordinates
  * Add command() function, which runs an external command
    and returns the output
  * erfinv() function and support for solving equations
    with error functions
  * Fix conversion with relative temperature units activated
  * Fix gammainc() function
  * Fix keyboard focus when run hidden automatically at startup
  * Temperature calculation modes (absolute, relative, hybrid)
  * Allow prefixes, with full name, without units (e.g. kilo = 10^3)
  * Extended and improved simplified Chinese translation
  * Save handle vector function argument property
  * Fix never ending loop when calculating []*[]
  * Minor bug fixes and feature improvements

• Files Changed
• Browse Source

New version

• Files Changed
• Browse Source

- Update to 3.15.0:
  * Automatic and dual approximation and fractions mode in qalc
    and Gnome shell search provider
  * Do not ask before overwriting temporary variables in qalc
  * Add keyboard shortcuts for approximation (Ctrl+E), fraction
    mode (Ctrl+F), and save variable (Ctrl+R) in qalc
  * Add "precalculate expression" argument (default false) to
    save() function (this also changes the behaviour of ":=")
  * Support multiple equivalent custom digits, separated by
    semicolon (e.g. "0;aA1;bB2;cC3"), and phoneword digit set
    in base() function
  * Always place unit with prefix first (fixes kWh)
  * Fix local time zone on Windows
  * Fix matrix/vector multiplication with approximate
    matrix/vector variable
  * Fix parsing of expression with multiple degree signs
  * Minor bug fixes and feature improvements

• Files Changed
• Browse Source

- Update to 3.14.0:
  * Support traditional calculator memory (MR/MC/MS/M+/M-)
  * Add Rutherford (Rd), millirem (mrem), mrem/s, and erg/g units
  * Add yr and annus names for year unit, röntgen name for
    roentgen (R) unit, and rem for roentgen equivalent man
  * Add matrix rank (rk) and reduced row echelon form (rref)
    functions
  * Support for different character encodings in code() function
  * Keep angle units in calculations involving both length and
    (plane) angle units (fixes lumen calculations)
  * Improve plot speed for functions defined using expression
  * Always perform conversion to optimal (SI) unit when the
    expression is a number multiplied by a unit and input equals
    output
  * Improvements and fixes for qalc on Windows
  * Compiled definitions without GIO and with support for
    out-of-tree builds
  * Fix availability of many currencies before manual exchange
    rates update (global exchange rates file was not installed)
  * Minor bug fixes and feature improvements

Corrected line lengths in this version.

• Files Changed
• Browse Source

- Update to 3.12.1:
  * Fix BTC, BYN, and BYR exchange rates
  * Fix exchange rates update time
  * Fix parsing of negative power in function arguments without
    parenthesis
  * Add arcsin, arccos, arsinh, arcosh, etc. as names for inverse
    trigonometric and hyperbolic functions
  * Interpret sin^-1, cos^-1, etc. as inverse functions
    (arcsin, arccos, etc.)
  * Add deftorad() and totient() functions

• Files Changed
• Browse Source

- Update to 3.12.0:
  * Improved exact simplification of roots (e.g. sqrt(2)-2/sqrt(2)=0)
  * Only apply negative exponents option to units, and always use
    negative exponents instead of division for unit expressions with
    only negative exponents
  * Support use of right arrow (including "->") as alternative to "to"
    for conversion (e.g. 5m->ft)
  * Use middle dot instead of multiplication x in unit expressions
  * Minor bug fixes and feature improvements

• Files Changed
• Browse Source

- Update to 3.11.0:
  * Allow multiple "to"-conversions in one expression (e.g. "5 m/s
    to hex to ft to h = 0xE6AF.1E*ft∕h)
  * Support specification of number of bits for hexadecimal "to"
    conversion (4 to hex16 = 0004)
  * Improvements and fixes for sexagesimal numbers and time format
  * Support for +, -, (, ) Unicode superscripts, and
    N-Ary Summation symbol
  * Syntax highlighting using colours in qalc
  * Enable Unicode by default in qalc
  * Improved display of unit expressions
  * Support for automatic updates of self-contained binaries
  * Fix use of equals sign in qalc command line arguments 
  * Fixes for the Gnome shell search provider
  * Minor bug fixes and feature improvements
  * Support for comments in expressions, using # (number sign)
  * Gnome shell search provider
  * Tau (2pi) constant
  * Allow angle as argument in nextlunarphase()
  * Fix automatic conversion to Newton
  * Fix x^2=25m^2 when x is assumed positive (was evaluated as
    false), and similar in exact mode
  * Fix display of repeating decimals
  * Fix segfault with empty sin() or sinh() in equation
  * Minor bug fixes and feature improvements
  * Fix segfault during calculation of atan(acos(x)),
    atan(asin(x)), and asin(atan(x))
    when angle unit is degrees or gradians
  * Fix order of prime factors
  * Fix immediate closing of plot window if qalc is started in
    non-interactive mode 
  * New icon (v3.9.0a)
  * Fix nextlunarphase(), for values > 0.8 and < 0.2 (v3.9.0a)
  * Fix segfault when running qalc in non-interactive mode
    without gnuplot (v3.9.0a)
  * Minor bug fixes and feature improvements

• Files Changed
• Browse Source