Path: utzoo!utgpu!jarvis.csri.toronto.edu!clyde.concordia.ca!uunet!cs.utexas.edu!samsung!think!eplunix!das
From: das@eplunix.UUCP (David Steffens)
Newsgroups: comp.sys.mac.programmer
Subject: question about QuickDraw regions
Keywords: QuickDraw regions CloseRgn
Message-ID: <850@eplunix.UUCP>
Date: 28 Feb 90 22:17:40 GMT
Organization: Eaton-Peabody Lab, Boston, MA
Lines: 50

My question concerns proper use of the QuickDraw CloseRgn call.

According to my reading of _Inside Macintosh_ Volume I, pages 181-183,
the following Pascal program fragment should always work correctly:
	RH:RgnHandle;
	for i:=1 to 100 do begin
		RH:=NewRgn;
		OpenRgn;
		{ draw something }
		CloseRgn(RH);
		{ do something with the region }
		DisposeRgn(RH)
	end

The description of SetEmptyRgn on page I-183 seems to imply
that the same function can be performed with one less call in the loop:
	RH:RgnHandle;
	RH:=NewRgn;
	for i:=1 to 100 do begin
		SetEmptyRgn(RH);
		OpenRgn;
		{ draw something }
		CloseRgn(RH);
		{ do something with the region }
	end;
	DisposeRgn(RH)

Depending how the description of CloseRgn is interpreted, however,
it seems reasonable to conclude that old region data, if any,
is first thrown away before the new region data is installed.
If true, the call to SetEmptyRgn seems unnecessary, and the number of calls
in the loop can be minimized while still maintaining correct behavior:
	RH:RgnHandle;
	RH:=NewRgn;
	for i:=1 to 100 do begin
		OpenRgn;
		{ draw something }
		CloseRgn(RH);
		{ do something with the region }
	end;
	DisposeRgn(RH)

I'm not loooking for an answer like: "Yeah, I tried it and it works."
I already did, and it does.  The real question here is --
can I afford to bet the farm on it _always_ working?
Can anyone provide a definitive answer?
-- 
{harvard,mit-eddie,think}!eplunix!das	David Allan Steffens
243 Charles St., Boston, MA 02114	Eaton-Peabody Laboratory
(617) 573-3748 (1400-1900h EST)		Mass. Eye & Ear Infirmary