Re: delwin() and deleting windows with subwindows

[Bill Gray]

On 8/13/22 19:20, Thomas Dickey wrote:
> On Thu, Aug 11, 2022 at 10:31:43PM -0400, Bill Gray wrote:
> >     I was tempted to keep recursive deletion (the code was all written and
> > tested),  but I've chickened out.
>
> good idea :-)

   (Hangs head) Yeah,  I got a little carried away when I pushed the 
recursive-deletion version.

> Today I made a modified test-program which made it simple(r) to fix that.

   The PDCurses makefiles can build a subset of ncurses test programs; 
 PDCursesMod has a larger subset,  to which I've added test_delwin.

> fwiw, NetBSD curses hangs (infinite loop) on the first delwin in your
> example, while Solaris dumps core on the last delwin.

   Well,  that's interesting.  It sounds as if trying to delete parent 
windows should be strongly discouraged.  I can't think of a case where 
you would legitimately try to delete a parent window anyway.  If you 
try,  you probably have a bug in your code.  (Similarly if you try to 
delwin() a window twice or delwin() a NULL pointer.)

   I've modified PDCursesMod so that in debug builds,  deleting a 
parent window triggers an assert().  Better that,  I think,  than 
letting things proceed and probably crash later on.

-- Bill