Matthias Hopf writes:
> 
> I have some problems using mutt (0.95.4) on IRIX 6.5.4m via rlogin or
> ssh.  Aparently the escape sequences produced by the cursor keys are
> not always mapped correctly to <up>, <down> and so on, but interpreted
> verbatim.  I always get a 'Key is not bound' in these cases.
> 
> As mutt behaves well when invoked locally, it most certainly has to do
> with vt100's brain-dead Esc schemes. I took a (quick) look at the
> sources but didn't find any possibilities to specify timeouts or
> anything else.

I recall this happening before to me on some machines when I was over a
particularly bad net connection.  The problem is that the arrow keys
actually generate a pair of keystrokes, and if there is too much time
between them, they are interpreted separately.

vi was especially vulnerable to this, and Bad Things would happen when
editing!

The following stuck in my .tcshrc file cured things.  I forget which
program (shell? curses?) made use of this variable, and what the units
were, but here it is:

# This next rather strange thing insures that DEC arrow keys are interpreted
# correctly in curses based applications(SC for example) otherwise the ESC
# is sometimes interpreted as a seperate character.
setenv ESCDELAY 5000

good luck!

     Alec

-- 
       Alec Habig, Boston University Particle Astrophysics Group
                           [EMAIL PROTECTED]
                       http://hep.bu.edu/~habig/

Reply via email to