tim@cs.columbia.edu (Timothy Jones) (08/07/90)
We're experiencing some weird problems with gnu emacs 18.55 and X on our HP 9000 series 300 machines that hopefully someone can help me solve. Whenever emacs is started with X switches, like -font or -geometry, a new buffer is created under the name of the argument to that switch. For example, if I run it with "emacs -font 9x15", emacs pops up in the correct font, but now there's a buffer named "9x15"! Also, emacs doesn't seem to respond to mouse button clicks for setting the cursor position anymore. These problems reared their ugly heads a couple of weeks ago (everything had been working fine before then). The strange part is, we haven't made any changes anything that I can think of which might cause this. No new emacs, no new X stuff, no new anything. It's quite perplexing. If I run emacs without X switches, everything seems to work fine. For example, "emacs -u <user>" pops up a new emacs running <user>'s .emacs, and there's no buffer named <user>. If I combine this switch with another, X related, one, the problem comes back again, which leads me to think that the X command-line parser isn't removing things from argv correctly. These problems occur both on machines running HP-UX 6.5 and 7.0, and using both the X11R3 server supplied with HP-UX and the new R4 server recently released by HP. Has anyone experienced similar behavior, or can you think of something I'm not? Please email responses. Thanks in advance, Tim -- Timothy Jones tim@cs.columbia.edu Department of Computer Science ...!rutgers!cs.columbia.edu!tim Columbia University tim%cs.columbia.edu@cuvmb.bitnet
bd@hpopd.HP.COM (Bob Desinger) (08/15/90)
> Whenever emacs is started with X switches, like -font or -geometry, a > new buffer is created under the name of the argument to that switch. > For example, if I run it with "emacs -font 9x15", emacs pops up in the > correct font, but now there's a buffer named "9x15"! There may be something wrong inside x_term_init() from the 18.55 distribution's "src/x11term.c", where the X arguments are handled. (Maybe the routine is modifying local copies of the args? I haven't tracked it down.) But my copy of 18.55 does the right thing with both -fn and -font. > Has anyone experienced similar behavior, or can you think of something > I'm not? I load my X settings from my .emacs file, which eliminates them from the command line: (load "term/bobcat") ; DWIM with C-h and DEL ;; Set defaults for X11 or the terminal (cond ((eq window-system 'x) ; if running under X ;; Simulate the shell command-line arguments: ;; emacs -bg \#592020 -fg Wheat -ms Gold -cr Red -fn user8x16 & (x-set-background-color "#592020") ; dark brown (x-set-foreground-color "Wheat") (x-set-mouse-color "Gold") (x-set-cursor-color "Red") (x-set-font "user8x16") ; not present on all systems ;; (server-start) (message "On X11.") (put 'eval-expression 'disabled nil) ; live dangerously ;; Because I have so many screens open all the time, a mistaken ;; C-x C-c can really hurt. So make sure I don't exit without ;; confirmation by rebinding C-xC-c appropriately. (defun my-exit-from-emacs () (interactive) (if (yes-or-no-p "Exit? ") (save-buffers-kill-emacs))) (global-set-key "\C-x\C-c" 'my-exit-from-emacs) (x-set-bell (quote (4))) ; editors should be seen, not heard ) (t (setq visible-bell t) ; editors should be seen, not heard (message "On a terminal") ) ) > Please email responses. Okay; I'll mail a copy of this. Please summarize any interesting responses to the net! -- bd