Re: nap and ncurses advice sought

[Thomas E. Dickey]

On Thu, 13 Dec 2001, Mike Castle wrote:

>
> I'm hoping the following will be of the "Oh, that's simple, do X" variety.
>
> I'm noticing an issue with the napster like client ``nap'' and ncurses (I'm
> seeing this on both my stable Debian system at work and home grown system
> at home).  It happens on xterm, but not the Linux console.

that's because the terminfo for Linux console doesn't set application mode
(probably because more people in that environment insist on hardcoding the
sequences into bash's .inputrc).

> What happens is, after a control-Z/fg, the sequence sent by the up arrow
> changes.  It changes from ^[OA to ^[[A.

(from memory - I'm not on Linux) it sounds as if you're saying that
suspending the process makes ncurses send the control sequence to turn
off keypad application mode (rmkx), but doesn't send smkx when returning.
That should be simple enough to check...

(with xterm you can recover by clicking the application mode checkmark in
the control/middle/mouse button - I have to do that occasionally with
non-ncurses stuff on IRIX).

> Oddly enough, the same thing happens with lynx, but it continues to
> correctly process the up arrow.

lynx has some fallback logic to handle vtxxx sequences that aren't in
termcap (it's probably using that)

> So, any pointers to the missing magic?  Or do I have a mis-configuration
> somewhere?
>
> Thanks,
> mrc
>

-- 
T.E.Dickey <address@hidden>
http://invisible-island.net
ftp://invisible-island.net