You learn something every day. Guess my hooking code was OTT. I notice that although some effort has obviously been made to implement CM_CANCELMODE, its effects look very similar to those of WM_CAPTURECHANGED (Rob, mightn't that work for you as well?). Can anyone enlighten me as to the difference?
Cheers, Carl > -----Original Message----- > From: [EMAIL PROTECTED] > [mailto:[EMAIL PROTECTED] Behalf Of Robert martin > Sent: Tuesday, September 05, 2006 7:42 PM > To: NZ Borland Developers Group - Delphi List > Subject: Re: [DUG] Component creation > > > Hi all > > I have this beaten now !!!! > > The required message is CM_CANCELMODE .... > > procedure CMCancelMode(var Message: TCMCancelMode); message > CM_CANCELMODE; > > > I found this using my expert skills (or looking at some third party > components we use and putting breakpoints in) ;-) > > I had tried this earlier but my component actually creates a child > component when expanded and I needed to better provide communication > between them. I am not an expert component creator (far from it) but > what I have is both components checking for this event and using the > ScreenToClient code provided by Todd to ensure the mouse isn't within > either component. If not it closes the 'component'. > > CM_CANCELMODE and its friend WM_CANCELMODE are really strangely > described my MS and others as 'providing special functionality in > certain circumstances' or some such. But essentially it seems to be > when a non focusable control or window is selected. > > Will test further but it seems this has hit the nail on the > head. A big > thanks to all who have contributed. > > > > Rob Martin > Software Engineer > > phone +64 03 377 0495 > fax +64 03 377 0496 > web www.chreos.com > > Wild Software Ltd _______________________________________________ Delphi mailing list [email protected] http://ns3.123.co.nz/mailman/listinfo/delphi
