1. 07 Oct, 2007 3 commits
  2. 06 Oct, 2007 2 commits
  3. 20 Sep, 2007 1 commit
  4. 09 Aug, 2007 1 commit
  5. 14 Jul, 2007 2 commits
  6. 11 May, 2007 1 commit
    • Cyrille Berger's avatar
      - remove lcmsism from KoColorSpace · 7b98e442
      Cyrille Berger authored
      - might fix a bug in YCbCr's convertion to and from rgb (at least fix one of the function which was obviously wrong, can't remember if the problem was visible or not)
      - disable convertion from read-only colorspaces in png/jpeg/tiff (for now)
      
      svn path=/trunk/koffice/; revision=663371
      7b98e442
  7. 28 Apr, 2007 1 commit
  8. 09 Apr, 2007 1 commit
    • Adrian Page's avatar
      q3-- · 9f593072
      Adrian Page authored
      
      svn path=/trunk/koffice/; revision=651849
      9f593072
  9. 27 Mar, 2007 2 commits
  10. 26 Mar, 2007 2 commits
  11. 24 Mar, 2007 1 commit
  12. 23 Mar, 2007 1 commit
  13. 02 Feb, 2007 1 commit
  14. 18 Jan, 2007 2 commits
  15. 09 Jan, 2007 1 commit
  16. 06 Jan, 2007 1 commit
    • Cyrille Berger's avatar
      remove useless Q_ASSERT(false) · acd9e34a
      Cyrille Berger authored
      fix toLab/fromLab toRgb/fromRgb in lcms
      implement convert toRgb/fromRgb in HDR colorspaces
      
      svn path=/trunk/koffice/; revision=620488
      acd9e34a
  17. 27 Dec, 2006 1 commit
  18. 22 Dec, 2006 1 commit
  19. 26 Nov, 2006 1 commit
  20. 29 Sep, 2006 1 commit
  21. 25 Sep, 2006 2 commits
  22. 13 Sep, 2006 1 commit
    • Boudewijn Rempt's avatar
      Refactored pigment to allow pluggable composite ops -- the bits of · 17cbbe32
      Boudewijn Rempt authored
      code that blend two pixels together, a la porter-duff. I have disabled
      all colorspaces except rgbu8, alpha and lab, and rgbu8 is enabled only
      by a hack pending the discussion I expect will happen shortly on how
      to discourage code duplication through templates. For now, everything
      seems to compile and Krita starts up.
      
      Basically, every composite op for every depth for every colorspace
      is now an object with a composite() method that does the work; whether we
      need to expose this object outside pigment, or provide merely a proxy KoID
      object is up for discussion.
      
      Also added some Krita documentation, updated the todo, removed the old
      startkrita for binary (autopackage-type) releases and removed some warnings.
      
      CCMAIL: kimageshop@kde.org
      
      svn path=/trunk/koffice/; revision=583883
      17cbbe32
  23. 23 Jul, 2006 2 commits
  24. 18 Jul, 2006 1 commit
  25. 11 Jun, 2006 1 commit
  26. 04 Jun, 2006 1 commit
  27. 03 Jun, 2006 1 commit
    • Casper Boemann's avatar
      Change inheritance of colorspaces. · 4a218a1e
      Casper Boemann authored
      KisAbstractColorSpace is now known as KisLcmsBaseColorSpace, and the 
      idea is that not every colorspace will use that
      KisColorSpace is now a virtual base of both lcmsbase and each of the U08 
      type bases
      
      So say rgb08 would inherit from both KisLcmsBaseColorSpace and 
      KisU08BaseColorSpace
      
      With diamond shaped inheritance final colorspaces (like rgb) has to call 
      the constructor of KisColorSpace.
      
      Also the conversion colorspace used as fallback is now Lab. meaning that 
      _every_ colorspace _must_ implement to and from Lab function.
      
      
      svn path=/trunk/koffice/; revision=547735
      4a218a1e
  28. 02 Nov, 2005 1 commit
    • Boudewijn Rempt's avatar
      * Fix after-deselect-selection-reappears-if-select-tool-is-used again bug · 7829c8f8
      Boudewijn Rempt authored
      * Fix apidox errors & add some more documentation (at least, _I_ don't
        get any errors anymore)
      * Update the TODO (it's smaller now!)
      * Tools now have an empty widget if they don't have a widget -- still need
        to fix the captioning for non-paint tools.
      * Add a primitive dcop interface to KisColorSpace. It should now be possible
        to write a dcop client that inverts a paint device but I don't know how to
        do that.
      
      svn path=/trunk/koffice/; revision=476894
      7829c8f8
  29. 15 Oct, 2005 2 commits
    • Boudewijn Rempt's avatar
      A little better: now I still need to get rid of the color_strategy subdirectory. · e85c8f71
      Boudewijn Rempt authored
      svn path=/trunk/koffice/; revision=470943
      e85c8f71
    • Boudewijn Rempt's avatar
      * Move the colorspaces to their own shared library. · b6553212
      Boudewijn Rempt authored
        There are still a few problems, because I made a few mistakes with
        file placing that I can only fix by committing, fixing and commiting.
        Also, the library isn't completely independent of the rest of Krita yet,
        see the TODO.
      * Add an uml diagram of the histogram design.
      * Add a meta registry -- that needs to be expanded so it becomes a central
        database for all the stuff we have in the myriad of singletons.
      * Fix the HACKING file about getters and setters to conform to Qt standards.
      
      svn path=/trunk/koffice/; revision=470933
      b6553212
  30. 12 Sep, 2005 1 commit
    • Boudewijn Rempt's avatar
      * Fix memory leak in palette manager · 226a91f2
      Boudewijn Rempt authored
      * Fix focus bug with dockers
      * Removed superfluous headers here and there
      * Made sure we code against the interface, not the implementation of KisColorSpace.
      * More API separation from implementation
      * Replaced all QUANTUM with Q_UINT8 -- we will never do the ImageMagick trick of
        redefining QUANTUM to Q_UINT16 to have instant 16-bit support. We can do better 
        than that.
      * Removed obsolete and empty toolbar 
      * Added a little documentation about the sdk effort.
      
      svn path=/trunk/koffice/; revision=460097
      226a91f2