Re: lynx-dev peculiar behavior - ^Z in editor

[Klaus Weide]

> > The easiest way to prevent Bad Things from happening here is to 
> > disable the SIGTSTP handler in this situation (use SIG_DFL). 
> >  
> > Should it be (n)curses' job to do this (probably always while in  
> > 'shell' mode)? 
> 
On Sat, 10 Jul 1999, T.E.Dickey wrote:

> If Lynx establishes its own SIGTSTP handler before initscr, then
> ncurses won't (isn't supposed to).  It's simpler to just have one....
> 
> > Should lynx do this? 
> > (It could be put into LYSystem().  Or into stop_curses/start_curses. 

I didn't mean that lynx should implement its own handler (as done with
slang), and it shouldn't change anything permanently.

More like:

  if (some conditions) {
     ...  /* save current SIGTSTP state, and set to SIG_DFL */
  }
  code = system(...);
  if (same conditions) {
     ...  /* restore SIGTSTP state to what it was */
  }

I think this should go into LYSystem().  Conditions should include
(!dump_output_immediately && !LYCursesON), to make the change only
for "interactive" child processes for which lynx has already done a
stop_curses().

We should use sigaction() instead of signal() here when available.

I hope this would remove the problem.  The "fix" might belong in ncurses,
but fixing that would not help users of older ncurses and other curses
libs.

Any objections?

   Klaus