Re: delete-process bug

[David Kastrup]

address@hidden (Kim F. Storm) writes:

> And to Eli:
>
> I also saw how MS-Windows handles process termination without SIGCHLD.
> Actually, it would do harm if SIGCHLD was ever defined on MS-Windows,
> so perhaps we should explicitly add something like this after
> all includes:
>
>
> /* MS-Windows has its own way of detecting process termination.
>    It does not normally define SIGCHLD, but just in case...  */
> #ifdef WINDOWSNT
> #undef SIGCHLD
> #endif

I seem to remember that Cygwin emulates an awful lot of Unix/Posix
functionality in its DLL.  Others will be more qualified to comment on
this, but the mere use of WINDOWSNT might not necessarily mean that
SIGCHLD can't work: it might be provided by the emulation layer.

Could somebody who actually knows Cygwin by more than heresay fill in
the details and tell whether I am completely wrong here?

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum