kjones@talos.pm.com (Kyle Jones) (04/14/90)
pastor@PRC.UNISYS.COM writes: > There are many operations that I try to perform that cause GNU Emacs to > *FORCE* me to respond to one or more questions; for example, if I > 'touch' a c source file to force make to do a recompile (say, to > recompile with a debug switch after a program bombs), when I try to > edit the source file again GNU Emacs will not let me proceed until I > respond to its "Are you sure?" message. Furthermore, in this and other > situations, GNU Emacs forces me to respond "yes" or "no". This can be fixed by the putting the following in your .emacs file: (fset 'ask-user-about-supersession-threat '(lambda (&rest ignore) (clear-visited-file-modtime))) If you go behind Emacs' back and touch a file, Emacs will say nothing when you try to edit the old buffer or when you save it later. This code can be easily expanded to force Emacs to be quiet only if a buffer local variable is non-nil, or the function ask-user-about-supersession-threat can be made to diff the buffer and the file, flip a coin or anything else you want in order to decide whether editing an `obsolete' buffer should be allowed. > In an editor that is supposed to be user-reconfigurable, and is > supposed to be usable by both novice and expert users, I regard the > inability to disable what I've come to call "Nag mode" as a serious > bug. I want the ability to tell Emacs to warn me if I'm about to do > something "stupid", but not to impede me. But that's exactly what Emacs was doing. It didn't stop you frmo editing the file, it just demanded that you show that you understand the consequences of your actions. After all, someone else may have modified the file you're editing, and it would important indeed that you be told that you were about to edit a version of the file that's vastly different from what's on disk, or that you were about to overwrite someone else's work. > I want the ability to tell Emacs to permit me to type 'y' or 'n' in > response to *ANY* question (i.e., always to act as though 'y-or-n-p' is > being called, rather then 'yes-or-no-p'). I don't think this would be wise. Since Emacs allows typehead, you could answer one of the y-or-n-p questions without even knowing it, with disaterous results. yes-or-no-p slows you down enough to read the question and to answer it consciously. There's much less likelihood of your typeahead doing you in. > I don't want to be mothered by my editor -- if I say "do it" I want the > editor to do it *UNLESS* I have specifically requested it to look over > my shoulder. The problem is that you don't always know when you need help, because you can't keep track of everything. Handling tedium is what computers are good at; let them help you. It's better to err on the side of caution. > While I'm at it, nobody has been able to tell me how to cure another > behavior that I also regard as a bug if it is not used-controllable. > When I delete a window, I do not want the space thus liberated > distributed among all other windows; I have arranged my screen the way > I want it, and I don't want Emacs to mess with it. Ok, you've got me there. I don't see a way to do this in Lisp. Presumably those laboring on version 19 read this group (or the mailing list thereof) and will hear your cries. Whether they will hearken to them or not is anyone's guess. Graduate students are a peculiar breed. (-: kyle jones <kjones@talos.pm.com> ...!uunet!talos!kjones "I'm a grad student. Why would I want to paint my nails?" -- Louise Penberthy