Hi, I just noticed that Compiz' window base size calculation in the function recalcNormalHints() is broken. If both PMinSize and PBaseSize are set, the base size is overwritten by the minimum size first and after that the minimum size is overwritten by the base size again. This leads to the base size information being completely lost. I have attached a patch which fixes this problem, however, I do not yet fully understand the idea behind the original code. Any objection against putting this patch in? Regards, Danny -------------- next part -------------- A non-text attachment was scrubbed... Name: fix-base-size-calculation.diff Type: text/x-patch Size: 1128 bytes Desc: not available Url : http://lists.freedesktop.org/archives/compiz/attachments/20070515/323f5b67/attachment.bin
On Tue, 2007-05-15 at 09:46 +0200, Danny Baumann wrote:> Hi, > > I just noticed that Compiz' window base size calculation in the function > recalcNormalHints() is broken. If both PMinSize and PBaseSize are set, > the base size is overwritten by the minimum size first and after that > the minimum size is overwritten by the base size again. This leads to > the base size information being completely lost. > I have attached a patch which fixes this problem, however, I do not yet > fully understand the idea behind the original code. Any objection > against putting this patch in?Yes, that code was not doing the right thing. I've pushed out changes that should make it work according to the ICCCM spec. [1] http://tronche.com/gui/x/icccm/sec-4.html#s-4.1.2.3 - David