Benjamin Shults wrote:
> Larry Troxler wrote:
> > investigated this, I remember finding some sort of re-centering function
> > that I was able to bind to a key, to at least manually shift the prompt
> > down to the last line of the buffer, after the output was finished. Now
> > I can't even figure out that much (the Xemacs docs are pretty poor).
> 
> M- - C-l
> 
> I think the docs are great.

Thanks! Somehow I missed in the docs how the recenter function behaves
with negative arguments. Actually, what the function that I was thinking
of was "comint-show-maximum-output", but now I remember that I was using
that in shell-mode with CLISP. I still don't understand why the Xemacs
help only helps you with functions that apply to the current mode. Oh
well, I guess the docs aren't too bad, in the sense that all the
information is there. Maybe it's just the help interface that isn't so
intuitive.

HOWEVER, I'd like to go the next step now and figure out how to have
this done automatically everytime eli detects that the Lisp prompt has
been received. But, nothing in the eli HTML doc mentions any kind of
hook for this, and a few minutes hunting the eli sources didn't turn up
any answers. Furthermore, I remember that someone on this list or
somewhere else, saying that in fact, this is the reason that this
behavior can't be easily implemented. Is my initial impression correct,
that the elisp interface does not detect the end of ACL output (the
prompt) as it occurrs?  Sorry for being a bit naive about how the emacs
socket-interface works. 

Doing the automatic recentering isn't the only situation where this
would be helpfull of course. For example, I started messing around with
elisp programming, and found that the following function doesn't work as
expected:

(defun cm-list-test ()
  (goto-char (point-max))
  (insert "list")
  (fi:inferior-lisp-newline)
  (recenter -1))

"list" in this case is a command to a custom top-level loop (Taube's
Common Music).

And what happens of course is that the "list" command is properly sent
to ACL, which properly prints the verbose many-line output. But the
"recenter" command doesn't work as expected, and I think this is because
at the time it is executed, the Lisp output from ACL has not yet been
received.

Any ideas on the right way to accomplish what I am seeking?

Larry
 


--  Larry Troxler --  <westnet.com at lt>  --  Patterson, NY USA  --