> 	if there's dlopen() call, we cannot predict what kind of code will be
> 	loaded into, hence then cannot test enough.
> 
> 	well, the decision is up to people using the system, not us.

That doesn't make sense for me.
User can predict the locale module by setting LC_* environment
variables.

If a user cannot control their environment variables,
then something is seriously broken in the environment.

If that made sense, we could not predit what kind of code will be
called by rcmd(3) interface, because we could not assume the value of
RCMD_CMD environment variable. Thus, we had to provide libnorcmd.a.

But that's wrong.
We never provide libnorcmd.a.

We don't have to provide dlopen()-less library for dynamically linked
binaries for same reason.
--
soda