"Andreas Rosenberg" <[EMAIL PROTECTED]> wrote: > >From time to time I'm testing our application (IQ-FMEA) with Wine. > > Our GUI framework is checking the return codes of most API calls, > so several incompatibilities between Windows and Wine are detected > while many other programs simply ignore them. > > So I found several things that differ between Wine & Windows.
Many thanks for reporting this bugs! Please continue reporting anything you find that might prevent your apps to work correctly under Wine. > ############################# > 1.) > > If found out that InvalidateRect with a "zero extent rectangle" > returns FALSE with Wine, while it returns TRUE in Windows. > Wine builds from last year did not show this behavior. > > > Windows: > InvalidateRect(hwnd, RECT(10,10,10,15), TRUE); returns TRUE; > > Wine > InvalidateRect(hwnd, RECT(10,10,10,15), TRUE); returns FALSE; A fix for this bug has been committed to CVS. > #################### > 2.) > > Then I discovered an incompatibility with LB_SELITEMRANGE: > > > SendMessage(hWndControl,LB_SELITEMRANGE,TRUE, > MAKELPARAM(wFirst, wLast)); > > Windows accepts values for wLast that are greater than the > list size and performs the operation with the (internally corrected) > value - returning TRUE. > > Wine does not accept a value for wLast if it is greater than > the list size and returns FALSE (not performing the select operation, > which caused subsequent failures in our application). I just sent a patch to wine-patches which should fix this bug. > #################### > 3.) > There is also a different behavior of SetCapture(hWnd) in Wine. > If an overlapped window has captured the mouse input and a new > child window is being openend, Windows internally releases the > capture. Wine maintains the capture - dialogs in our application > may not be operated with the mouse, because the parent still gets > all the mouse messages. Could you please send a small test app demonstrating this behaviour? -- Dmitry.
