Re: Question about g_spawn_sync/child_setup

[Tristan Van Berkom <vantr touchtunes com>]

mike flyn org wrote:
> Hello everyone.
>
> The g_spawn_sync provides a handy option of providing a setup function
> (child_setup) that is executed by the child process before executing the
> child's argument vector.  However, child_setup returns void.  Because of
> this, there is now way to ensure the success of child_setup before executing
> the argument vector.
>
> I am interested in using child_setup to do some UID manipulation.  However,
> it is important to ensure that this manipulation does not fail before
> executing anything.  How may this be done?  Why does child_setup not return a
> gboolean that could then be checked by the g_spawn_sync function before
> executing anything?  I wish to avoid calling fork myself.

   Regardless of whether this is a design flaw or not,
cant you simply `exit()' ?

Cheers,
                      -Tristan