On Sat, Oct 14, 2017 at 07:06:06AM +0000, Jean-Marc Lasgouttes wrote: > Le 14 octobre 2017 04:39:49 GMT+02:00, Scott Kostyshak <skost...@lyx.org> a > écrit : > >Are you proposing this patch for 2.3.x? I remember that I could not > >reproduce the issue on 2.3.x, but I think that others could, right? > > > No, this is only related to the new painting scheme in the properpaint > branch. Basically what happens is that, when one has consecutively > processEvents(Force); > processEvents(FitCursor); > everything was Ok in the 2.3 world : the screen would be fully redrawn with > the first call, and then probably not touched with the second. > > In the new world, the first call will recompute metrics and ask for full > redraw and the second will ask for no redraw. So when the paint event occurs, > no painting is done. The patch makes sure that update flags are cumulative. > > But since this is a big semantic change, I expect collateral effects.
OK thanks. I'm taking this issue off of my 2.3.0 radar then. Scott
signature.asc
Description: PGP signature