Re: getch always returns -1 after shell

[Tim Mooney]

In regard to: getch always returns -1 after shell, jon wackley said (at...:

>Hi,
>
>I have a problem with ncurses-5.3, I use the following pseudo-code;
>
>   def_prog_mode();
>   endwin();
>   /* spawn some shell processing */
>   refresh();
>
>Everything seems fine until the next execution of getch().  getch always
>returns -1, (ERR).  Is there something special I have to do to ensure
>the input state is consistent before and after spawning shell commands?
>Any suggestions where to look to track down this problem?  The term is a
>"sun-cmd".

Hmmm, I would think that you're on the right track.  *Years* ago (like
93-94) I hacked a curses-based menu program  to support shell escapes
via system, and the notes and comments I had say:

 * After carefully reading the *excellent* curses (sysv, curses.3v)
 * manpage on Solbourne OS/MP, here's the scoop:
 * 1. initscr should be called only *ONCE*.  It does global variable
 * setup.
 * 2. before each system() call or shell escape, you should call endwin.
 *    this will reset the terminal state, and get things ready for the
 *    the next program to execute.
 * 3. After finishing a shell escape or system call, either wrefresh or
 *    doupdate() should be called- and they will reset the proper terminal
 *    modes.

The menu made extensive use of getch, and I never had a problem with
errors.  My guess is you're on the right track.  Why it's not working,
I don't know.

Tim
-- 
Tim Mooney                              address@hidden
Information Technology Services         (701) 231-1076 (Voice)
Room 242-J6, IACC Building              (701) 231-8541 (Fax)
North Dakota State University, Fargo, ND 58105-5164