🚀 go-pugleaf

I was looking through the sample program *MiniEdit* discussed in
Stephen Chernicoff's book *Macintosh Revealed Vol II* when I stumbled
across what I think maybe a bug, please correct me if I'm wrong.

It involves the interaction between the procedure *CloseAppWindow* and
the procedure *DoActivate*:

In *CloseAppWindow*, TEDispose is called to dispose of the program
window's edit record, which is part of a structure whose handle is the
window's reference constant.

Then HideWindow is called to force a deactivate event, upon which
Chernicoff's Activate/Deactivate event handler is invoked:

	TEDispose(editRec);
	...
	HideWindow(TheWindow);	/*	force deactivate event	*/
	if GetNextEvent(ActivateEvt, TheEvent) then
		DoActivate;					/*	handle deactivate event	*/
	if GetNextEvent(ActivateEvt, TheEvent) then
		DoActivate;					/*	handle activate event	*/

However, in *DoActivate*, if the event is a deactivate event, the
window's edit record is deactivated:

	if BitAnd(modifiers, ActiveFlag) <> 0 then
		/* handle activate event	*/
	else begin
		/*	handle deactivate event	*/
		...
		TEDeactivate(editRec);
		...
	end;

Will this not bomb, since the program is referencing (trying to
deactivate) a nonexistent (NULL?) edit record in TEDeactivate?

Also, aren't the parameters to GetNextEvent incorrect? Shouldn't the
first parameter be ActivMask instead of ActivateEvt?

thanks in advance for any help
-- 
Frank Ng			ACSnet:	frankng@basser.oz
Department of Computer Science	ARPA:	frankng%basser.oz@seismo.css.gov
University of Sydney NSW 2006	UUCP:	seismo!munnari!basser.oz!frankng
AUSTRALIA