Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!tut.cis.ohio-state.edu!APPLE.COM!mrspoc!guinan!kayvan
From: mrspoc!guinan!kayvan@APPLE.COM
Newsgroups: gnu.emacs
Subject: Multiple shells
Message-ID: <8903161814.AA20416@apple.com>
Date: 16 Mar 89 18:14:40 GMT
References: <8903160713.AA03138@la.tis.com>
Sender: daemon@tut.cis.ohio-state.edu
Reply-To: mrspoc!kayvan@apple.com
Distribution: gnu
Organization: GNUs Not Usenet
Lines: 70

> Date: Wed, 15 Mar 89 23:13:40 PST
> From: apple!la.tis.com!ho (Hilarie K. Orman)
> 
> It is relatively easy to hack shell.el to allow multiple shell
> processes, each in its own buffer.

I got tired of people hacking various shell.el's to put in multiple
shell buffers. So I wrote my own function (which I bind to C-M-r)
to automatically rename the current buffer. It is appended to the
end of this message.

This function creates names like "1:*shell*" "2:*shell*", etc., making
it easy to switch-buffer with completion.

> This largely eliminates the need
> for job control, and it is an excellent way of organizing one's work.
> I normally work with 2-5 shells at a time.  Does anyone else work this
> way?

Yes. I love it. Once inside emacs I never have to leave. If I need to do
a screen-oriented thing, I just drop to a subshell (C-z on SysV) and
do it and return back to my emacs.

> The only major problem I have had with shell windows is that
> passwords from telnet and rlogins and su's are visible, and I finally
> had to resort to implementing a command to grab and stuff without
> echoing. 

Try using the new cmushell package that came out recently.
It does filename completions, among other nifty things as well.
My cmushell-load-hook for it is set in my .emacs as follows:

(setq cmushell-load-hook
      '((lambda () (define-key cmushell-mode-map "\C-c\C-i" 'send-invisible))))

This sets the key-esquence C-c C-i (i for invisible) to the send-invisible
function which grabs text invisibly. This means that after typing su in
my "1:*cmushell*" buffer, I type C-c C-i before typing in the password.

Anyways, here is the rename function.

			---Kayvan

---------------------------- cut here -----------------------------------
;
; rename.el - Functions to rename buffers.
;
; This is especially useful for shell process buffers that need to be
; renamed for multiple buffers to exist.
;
; Author: Kayvan Sylvan
;

(defvar rename-format "%d:%s"
  "*Format string that rename-to-new-name uses to generate new name.
The number comes first so that the least substring matches buffer names.")

(defun rename-to-new-name ()
  "Rename current buffer to new name not already taken. The new buffer names
are in the format specified by rename-format (using old name and a number).
This function is useful for creating multiple shell process buffers"
  (interactive)
  (let (new-name (old-name (buffer-name (current-buffer))) (i 1))
    (while (get-buffer (setq new-name (format rename-format i old-name)))
      (setq i (1+ i)))
    (rename-buffer new-name)
    (set-buffer-modified-p (buffer-modified-p)))) ; force mode line update

Kayvan Sylvan @ Transact Software, Inc. -*-  Mountain View, CA (415) 961-6112
Internet: mrspoc!kayvan@apple.com UUCP: ...!{apple,pyramid,mips}!mrspoc!kayvan