On 26.02.2012 22:18, Ian MacArthur wrote: > > On 26 Feb 2012, at 16:30, Georg Potthast wrote: > >> I tried the example on Windows, Linux and DOS now - it did work. > > I'm sure it does - I'd hazard that it probably appears to work fine on > 80%-90% of current systems, then will occasionally fail oddly on most of the > others, and outright now work at all on the remainder. > In particular it can be very sensitive to how the graphics driver updates the > display, so we often see different behaviour on different graphics cards, on > otherwise similar host machines. > And I'm speaking from hard-earned experience here... > > What might be interesting to try is to run your example, minimise the window > then re-show it. My experience is that this will generally (but not always) > draw an empty window without any of the previously drawn features retained.
An even better test would probably be to make the window resizable and test resizing. Some window systems might keep the window buffer somewhere when the window gets minimized and redisplay the buffer when it is shown again. Resizing would really need updates calling draw(), and then the failure should be visible. Albrecht _______________________________________________ fltk mailing list [email protected] http://lists.easysw.com/mailman/listinfo/fltk
