Ben> This fixes a long-standing problem identified by Jeff Miller. Ben> Apparently back when I did the big Mule patch, I tried to fix Ben> a bug in set-window-point and make it more like it seemed Ben> like it was trying to do. This, however, introduced some Ben> problems. The FSF behavior is different from either our old Ben> or new behavior, and perhaps more logical, so I'm adopting Ben> it. Ben> Not committed yet because CVS commit is broken. Ben, CVS at least works for Ville and myself. Please try something like the following and send output to cvs-admin(a)xemacs.org. C:\>ssh -2 -l xemacs sunsite.dk Enter passphrase for key '/cygdrive/c/Users/AichnerAd/.ssh/id_dsa': version M Concurrent Versions System (CVS) 1.11.1p1 (client/server) ok E cvs [server aborted]: received interrupt signal Connection to sunsite.dk closed. C:\> Ben, you'll have to use -1 instead of -2 above until you send us your SSH2 key. Best regards, Adrian Ben> src/ChangeLog addition: Ben> 2002-12-15 Ben Wing <ben(a)xemacs.org&gt; Ben> * window.c (Fset_window_point): Correct to FSF behavior. This Ben> fixes some bugs in GNUS reported by Jeff Miller. Ben> latest source patch: Ben> Diff command: bash -ci "cvs-diff -no-changelog " Ben> Files affected: src/window.c Ben> Index: src/window.c Ben> =================================================================== Ben> RCS file: /pack/xemacscvs/XEmacs/xemacs/src/window.c,v Ben> retrieving revision 1.69 Ben> diff -u -r1.69 window.c Ben> --- src/window.c 2002/12/08 10:25:11 1.69 Ben> +++ src/window.c 2002/12/15 12:09:25 Ben> ＠＠ -1809,20 +1809,24 ＠＠ Ben> DEFUN ("set-window-point", Fset_window_point, 2, 2, 0, /* Ben> Make point value in WINDOW be at position POS in WINDOW's buffer. Ben> -If WINDOW is the selected window, this actually changes the buffer's point Ben> -instead of the window's point. (The equivalence of the selected window's Ben> -point with its buffer's point is maintained throughout XEmacs.) Ben> +If WINDOW is the selected window, and window's buffer is the current Ben> +buffer, this actually changes the buffer's point instead of the window's Ben> +point. (The equivalence of the selected window's point with its buffer's Ben> +point is maintained throughout XEmacs. However, enforcing the additional Ben> +restriction on the current buffer is "bug compatible" with FSF and is Ben> +perhaps more logical.) Ben> */ Ben> (window, pos)) Ben> { Ben> struct window *w = decode_window (window); Ben> CHECK_INT_COERCE_MARKER (pos); Ben> + Ben> /* Don't dereference selected-window because there may not Ben> be one -- e.g. at startup */ Ben> - if (EQ (wrap_window (w), Fselected_window (Qnil))) Ben> - /* Even though window selected, buffer may not be current */ Ben> - Fgoto_char (pos, w->buffer); Ben> + if (EQ (wrap_window (w), Fselected_window (Qnil)) Ben> + && EQ (w->buffer, Fcurrent_buffer ())) Ben> + Fgoto_char (pos, Qnil); Ben> else Ben> set_marker_restricted (w->pointm[CURRENT_DISP], pos, w->buffer); -- Adrian Aichner mailto:adrian＠xemacs.org http://www.xemacs.org/