Well actually it depends. when you move the cursor you just blit the buffer with no cursor to the topmost buffer, and repaint the cursor. Bliting the buffer should be quite fast, so if your image is like lots of shape objects you render every time this is many times faster.
If this is analogous to what you are already doing (ie your image renders fast), you could do this (as you were thinking): have a small buffer the size of the cursor contain what is under the outline, and just blit that back and make a new one in the right position before repainting the outline. Another method is to paint the outline using the 'XOR' method instead of 'REPLACE', which means you just have to paint the cursor again in the same place to erase it and return the picure to it's original state; and te paint it in the new position. just in case: XOR is exclusive OR; 0xor0 = 0, 0xor1/1xor0 = 1; 1xor1 = 0 so b xor c xor c = b (paint c twice on b yields b : cool ) HTH, Jonathan On 9/30/07, Keith Feesh <[EMAIL PROTECTED]> wrote: > > Hmm, so no matter what you do, you end up redrawing the whole image again. > I guess with some time you could make it only redraw the last place you were > at instead of the whole image, but if that's the only method then luckily > I've been doing it right. Cheers Jonathan. > > On 30/09/2007, Jonathan Winterflood <[EMAIL PROTECTED]> > wrote: > > > > There is indeed a method. > > > > One would probably call it double buffering (anthough when you're done > > with your program you might have moe like quad buffering... > > > > The idea is to render everything in layers: > > > > bottom layer = your image (data, etc) > > above: rulers, marks, selection box, etc (semi-static stuff) > > above: brush outline > > > > these are hopefully in the order of least to most often > > changing, and what is painted on top of what. > > > > the data is held in a pixbuf > > you paint that pixbuf and the selection etc into another pixbuf > > you paint that pixbuf and the outline into the topmost buffer > > > > then that pixbuf is painted onto the screen. > > > > HTH, > > Jonathan > > > > On 9/30/07, Keith Feesh < [EMAIL PROTECTED]> wrote: > > > > > Hello, I have been developing a drawing application for quite some > > > time now > > > and I have implemented an outline system > > > which shows where your next drawing operation will go (just load up > > > the > > > gimp, go on the paintbrush tool and select a size 11 > > > brush if you don't quite get what I mean). > > > > > > The way that I do things now includes redrawing the entire canvas > > > every time > > > the mouse is moved, in order to clear the old > > > outline's drawing, and then to redraw the outline once again. This > > > brings my > > > system to it's knees and I figure there must be > > > a better method, just I can't think of any method. Any hints would be > > > great. > > > _______________________________________________ > > > gtk-app-devel-list mailing list > > > gtk-app-devel-list@gnome.org > > > http://mail.gnome.org/mailman/listinfo/gtk-app-devel-list > > > > > > > > -- <Morpheus> linux, c'est une question de VI ou de MORE _______________________________________________ gtk-app-devel-list mailing list gtk-app-devel-list@gnome.org http://mail.gnome.org/mailman/listinfo/gtk-app-devel-list
