--- Tomas Frydrych <[EMAIL PROTECTED]> wrote: > > At the heart of the problem is the fact that our > colour classes cannot > represent transparent colour. My suggestion is that > we change the > m_red, m_grn, m_blu members of UT_RGBColor and > related from > unsigned char to signed short and use -1,-1,-1 as a > representation > of transparent color. This will do away with the > need to re-examine > the bgcolor property in > fp_Run::updateBackgroundColor(). > > Any objections? If not, I will do this.
Yes: 1) It's a hack 2) You will consume more memory than need (ok, not an issue here, but anyway...) 3) you gain a whole range of invalid colors, for instance (-2, -4, -3)? 4) It's a hack :) Until we gain alpha handling all over abiword, the easiest way to cope with this problem is just to store next to the color a boolean to say if it's transparent. That's also not ideal, but it's localized. Only the fp_Run will be affected, and not too much. Alternatively, you all can just wait until night falls, I reach my computer and make a diff against current CVS :) I guess that I didn't make anything too stupid, and that it should work ok. Cheers, ===== Joaquin Cuenca Abela [EMAIL PROTECTED] __________________________________________________ Do You Yahoo!? Yahoo! Health - Feel better, live better http://health.yahoo.com
