1. 04 Feb, 2008 2 commits
  2. 03 Feb, 2008 9 commits
  3. 02 Feb, 2008 1 commit
  4. 01 Feb, 2008 3 commits
  5. 31 Jan, 2008 5 commits
  6. 30 Jan, 2008 13 commits
  7. 29 Jan, 2008 7 commits
    • Ariya Hidayat's avatar
      prevent potential crash · 1a6e2319
      Ariya Hidayat authored
      svn path=/trunk/koffice/; revision=768399
      1a6e2319
    • Jan Hambrecht's avatar
      add configuration widget for snapping guides · b3b219f0
      Jan Hambrecht authored
      
      svn path=/trunk/koffice/; revision=768398
      b3b219f0
    • Cyrille Berger's avatar
      give back to caesar what's belong to caesar : add the copyright and licence of... · ab0eed66
      Cyrille Berger authored
      give back to caesar what's belong to caesar : add the copyright and licence of the hypenation dictionnaries
      
      
      svn path=/trunk/koffice/; revision=768390
      ab0eed66
    • Bart Coppens's avatar
      Whoops, seems like I forgot to commit a file. · 7d6d25d6
      Bart Coppens authored
      
      svn path=/trunk/koffice/; revision=768356
      7d6d25d6
    • Bart Coppens's avatar
      *Lightning Flash* *KABOOM!* · e52910b1
      Bart Coppens authored
      Bet you didn't see this one coming, eh? ;-)
      
      This commit adds an experimental new tile backend, which features actual shared tiles and so. It's also supposed to be more robust for threading. Emphasis on the 'supposed'.
      
      It's still very experimental, a lot of things probably are just plain wrong and bad code, and it might crash your Krita. It seems almost all regression tests pass (including some new ones written specifically for it), but sometimes there is a random SIGSEGV, I haven't figured out yet what's causing those (but you can be absolutely sure that the amount of crashes is significantly lower than while I was coding this all). It also doesn't do a lot of things yet, like clean up its tempfiles at exit and all that dull stuff about deleting unused data and so ;-)
      
      Cyrille will be (not very) pleased to find out it is also significantly slower, at least when just timing the make tests. I'm suspecting either bad coding (possibly memory leakage, or just badly written), or too much time being spent in locking.
      
      This codepath is disabled by default (Duh!). To try it out, edit the krita/image/CMakeLists.txt, and comment out the 'set(USE_OLD_TILESYSTEM 1)'.
      
      Have fun with it! (Or not... :P)
      
      CCMAIL:kimageshop@kde.org
      
      
      
      svn path=/trunk/koffice/; revision=768355
      e52910b1
    • Jan Hambrecht's avatar
      - some restructuring to support more than one snapping strategy at once · 6fcf3164
      Jan Hambrecht authored
      - added a second snap strategy which snaps to path nodes
      
      
      svn path=/trunk/koffice/; revision=768347
      6fcf3164
    • Thorsten Zachmann's avatar
      o use the KoVariableRegistry and Factory · bd6a7a24
      Thorsten Zachmann authored
      o do no longer compile the KoInlineFactory and KoInlineRegistry as the
        job is now done be the KoVariableFactory and KoVariableRegisty. With
        this the loading of the variables is done in the variable plugins
        where it belongs too.
      
      svn path=/trunk/koffice/; revision=767938
      bd6a7a24