Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.1 6/24/83; site sdcsvax.UUCP
Path: utzoo!linus!decvax!ittvax!dcdwest!sdcsvax!laman
From: laman@sdcsvax.UUCP
Newsgroups: net.bugs.4bsd,net.bugs
Subject: YACB (Yet Another (4.2 BSD) Curses Bug)
Message-ID: <967@sdcsvax.UUCP>
Date: Mon, 9-Jul-84 13:30:56 EDT
Article-I.D.: sdcsvax.967
Posted: Mon Jul  9 13:30:56 1984
Date-Received: Tue, 10-Jul-84 07:46:03 EDT
Organization: EECS Dept. U.C. San Diego
Lines: 47

<Curses, foiled again>

The bug that Mike Lilley found in overlay() is where the last line of the window
being overlayed is not overlayed.  The following program will tell you
if you have the bug.  The fix follows the program.  It seems that the
extraction of this library from "vi" has bitten us again.

/* Beginning of program.  Cut here------- */
#include <curses.h>

main() {
	register WINDOW *win;

	initscr();
	noecho();
	win = newwin(0, 0, 0, 0);
	mvwaddstr(win, LINES/2, 12, "---Hit return to exit---");
	mvwaddstr(win, LINES-2, 0, "The middle of the next line should say \"Hi\".  If not you have the overlay bug.");
	mvwaddstr(win, LINES-1, COLS / 2 - 10, "Hi.  You don't have the bug");
	overlay(win, stdscr);
	refresh();
	getchar();
	move(LINES - 1, 0);
	refresh();
	echo();
	endwin();
}
/* End of program. Cut here--------- */

If you have the bug, the following fixes takes care of the problem:

	On the first for loop in overlay() is:

		for (y = starty; y < endy; y++) {

	It should be:

		for (y = starty; y <= endy; y++) {

	The problem is that the computation of "endy" takes into account
    the fact that "_maxy" is "1 relative".  The subtraction of 1 now makes
    it "0 relative", but the for-loop is treating it as "1 relative".
    Another solution is to remove the "- 1" from the assignment to "endy".
    Which ever you prefer.

		Mike Laman, NCR Torrey Pines
		UUCP: {ucbvax,philabs,sdccsu3,sdcsla}!sdcsvax!laman