Martin, Thanks for this patch. It will go into 21.1.12. - vin Martin> No changelog Martin> Index: PROBLEMS Martin> =================================================================== Martin> RCS file: /usr/CVSroot/XEmacs/xemacs/PROBLEMS,v Martin> retrieving revision 1.26.2.16 Martin> diff -u -w -r1.26.2.16 PROBLEMS Martin> --- PROBLEMS 2000/06/30 08:49:54 1.26.2.16 Martin> +++ PROBLEMS 2000/08/02 08:14:41 Martin> ＠＠ -523,7 +523,7 ＠＠ Martin> This is usually because xmkmf is not in your path or because you are Martin> using the default cygwin shell. The default cygwin shell (/bin/sh.exe) Martin> -is ash which appears to work in most circumstances but has some wierd Martin> +is ash which appears to work in most circumstances but has some weird Martin> failure modes. I recommend replacing sh.exe with bash.exe, this will Martin> mean configure is slower but more reliable. Martin> Index: ./src/event-Xt.c Martin> =================================================================== Martin> RCS file: /usr/CVSroot/XEmacs/xemacs/src/event-Xt.c,v Martin> retrieving revision 1.41.2.28 Martin> diff -u -w -r1.41.2.28 event-Xt.c Martin> --- event-Xt.c 2000/07/21 10:16:00 1.41.2.28 Martin> +++ event-Xt.c 2000/08/02 08:14:47 Martin> ＠＠ -3055,9 +3055,9 ＠＠ Martin> if ((d = get_device_from_display_1(dpy))) { Martin> visual = DEVICE_X_VISUAL(d); Martin> if (colormap != DEVICE_X_COLORMAP(d)) { Martin> - XtAppWarningMsg(the_app_con, "wierdColormap", "cvtStringToPixel", Martin> + XtAppWarningMsg(the_app_con, "weirdColormap", "cvtStringToPixel", Martin> "XtToolkitWarning", Martin> - "The colormap passed to cvtStringToPixel doesn't match the one registerd to the device.\n", Martin> + "The colormap passed to cvtStringToPixel doesn't match the one registered to the device.\n", Martin> NULL, 0); Martin> status = XAllocNamedColor(dpy, colormap, (char*)str, &screenColor, &exactColor); Martin> } else { Martin> Index: ./src/glyphs.c Martin> =================================================================== Martin> RCS file: /usr/CVSroot/XEmacs/xemacs/src/glyphs.c,v Martin> retrieving revision 1.23.2.76 Martin> diff -u -w -r1.23.2.76 glyphs.c Martin> --- glyphs.c 2000/07/22 19:33:24 1.23.2.76 Martin> +++ glyphs.c 2000/08/02 08:14:50 Martin> ＠＠ -887,7 +887,7 ＠＠ Martin> break; Martin> } Martin> - /* The image may have been previously finalized (yes that's wierd, Martin> + /* The image may have been previously finalized (yes that's weird, Martin> see Fdelete_frame() and mark_window_as_deleted()), in which case Martin> the domain will be nil, so cope with this. */ Martin> if (!NILP (IMAGE_INSTANCE_DEVICE (i))) Martin> Index: ./src/window.c Martin> =================================================================== Martin> RCS file: /usr/CVSroot/XEmacs/xemacs/src/window.c,v Martin> retrieving revision 1.41.2.54 Martin> diff -u -w -r1.41.2.54 window.c Martin> --- window.c 2000/07/22 19:33:30 1.41.2.54 Martin> +++ window.c 2000/08/02 08:14:55 Martin> ＠＠ -5228,7 +5228,7 ＠＠ Martin> font size change happened between saving and restoring, the Martin> frame "sizes" maybe equal but the windows still should be Martin> resized. This is tickled alot by the new "character size Martin> - stays constant" policy in 21.0. It leads to very wierd Martin> + stays constant" policy in 21.0. It leads to very weird Martin> glitches (and possibly crashes when asserts are tickled). Martin> Just changing the units doens't help because changing the