Skip to content
Hugo edited this page Sep 16, 2019 · 35 revisions

News

New in version 0.4.1

  • enhancement: add HSLuv and HPLuv color spaces (contributed by masonium)

New in version 0.4.0

  • incompatible change: gen-cat-function and define-cat-function can now take only XYZ space as target
  • incompatible change: every slot of structures rgbspace, illuminant, cat, and observer are read-only
  • rename: optimal-spectrum1, optimal-spectrum2 to optimal-spectrum-peak, optimal-spectrum-trough
  • enhancement: users can inline all of the converters by explicitly declaring inline
  • enhancement: make-cat can take various types of objects as matrix.
  • optimization: faster CIEDE2000 and other miscellaneous optimizations

New in version 0.3.1

  • enhancement: better extrapolation of the Munsell Renotation Data

New in version 0.3.0

  • major incompatible change: The system names and package names now use / instead of - as separator: e.g. dufy-extra-data -> dufy/extra-data. However, the old package names are left as nicknames.
  • minor incompatible change: The white points of (parts of) built-in standard illuminants are calculated with the same precision as CIE 15.3:2004. Specifically, the white points of illuminants A, B, C, and D are slightly changed.
  • bug fix: hsl-to-rgb previously returned a wrong result when it received a hue outside the interval [0, 360].

New in version 0.2.9

  • Now dufy works on ABCL.

New in version 0.2.7

  • add an example of processing spectral data

New in version 0.2.6

  • rename: packed RGB is now called rgbpack (formerly int)
  • enhancement: add RGBA space: rgba (gamma-corrected RGBA), qrgba (quantized RGBA) , rgbapack (packed RGBA). Byte-order can be ARGB or RGBA.
  • enhancement: add several predefined illuminants to dufy-extra-data package: FL 3.x series, illuminant B, several gas discharge lamps.
  • enhancement: add CMC l:c delta-E
  • optimization: faster CIEDE2000 (-deltae00 functions)
  • some minor bugfixes

New in version 0.2.3

  • major incompatible change: Now most converters don’t use &optional arguments but &key arguments.
  • incompatible change: make-illuminant doesn’t receive xy values but XYZ values.
  • rename: def-cat-function -> define-cat-function
  • rename: deltae -> deltaeab
  • rename: deltaeXX -> lab-deltaeXX
  • enhancement: Now hsv and hsl converters can handle a non-normal RGB space.
  • enhancement: quantize and dequantize functions are introduced, which (de)quantizes a single value.
  • optimization: faster quantization and dequantization of RGB values

New in version 0.2.2

  • modularization: dufy is splitted into several independent modules: dufy-core, dufy-munsell, dufy-extra-data, etc. You can use the main package dufy in the same way as before.
  • bug fix: Wrong illuminant in wide-gamut RGB space #9
  • enhancement: add illuminant series F (by wasserwerk)
  • add an example module dufy-examples

New in version 0.2.0

  • major incompatible change: The term hex is changed to int: e.g. All hex-to- / -to-hex converters are now int-to- / -to-int instead.
  • incompatible change: gen-cat-function and def-cat-function don’t take &optional arguments but &key arguments now.
  • incompatible change: Now the default value of :clamp argument of all the -to-qrgb functions is T, which was nil before.
  • incompatible change: change the function name max-chroma to max-chroma-in-mrd.
  • incompatible change: add the suffix -illum-c to all the Munsell converters whose illuminant is C: mhvc-to-lchab-illum-c, munsell-to-lchab-illum-c, lchab-to-mhvc-illum-c, lchab-to-munsell-illum-c
  • enhancement: gen-cat-function, def-cat-function can now take not only XYZ but also xyY, L*a*b*, L*u*v*, LCHab and LCHuv as target.
  • improvement: The LCHab-to-Munsell converter works better, if an achromatic color is given. Besides, the action can now be specified with :if-reach-max argument if inversion fails.
  • optimization: faster conversion between Munsell and LCHab than before

New in version 0.1.14

  • enhancement: add approximate-spectrum which returns a piecewise linearization of a given spectrum.
  • enhancement: spectrum-to-xyz can now take the range and the sample interval of wavelength as optional arguments.
  • enhancement: gen-spectrum can now take not only (simple-array double-float (*)) but also a sequence of double-floats as spectral data.
  • optimization: faster matrix operations.

New in version 0.1.13

  • enhancement: add gen-rgbspace-changer which is for changing RGB working spaces.
  • bug fix: Build fails in SBCL 32 bit. #6