Re: fndir introspection, site-packages documentation

[Bart Schaefer <schaefer@xxxxxxxxxxxxxxxx>]

On Mar 16,  9:08am, Roman Neuhauser wrote:
}
} but then i'm bound to miss replies.  i don't want to subscribe to
} zsh-workers (yet)...

We're pretty good about Cc'ing people when they ask to be.

Everybody, please Cc Roman.

} > Can anyone give me a situation in which an
} > oridinary user (i.e., not a software packager) would need this?
} 
} how about: a site admin wants to install a function for all users.

And the site admin can't determine where to do this by simply looking
at the value of $fpath?  If he's automating it for a bunch of dissimilar
platforms he's back to being a software packager.

} the behavior with *default* --prefix is, well, completely contrary
} to the intent described: "no matter how the shell was configured" is
} blatantly false thanks to this special case.

I'm leaving this to PWS as he made the changes to configure.ac.  I will
however note that I think --disable-site-fndir ought to do exactly that,
and the documentation blurb is what needs fixing.  If the shell is being
configured for some kind of secure/embedded environment, removing all
access to locally added functions is perfectly reasonable.

} > Maybe the right thing is to throw all of this into a module (and doc
} > for that module) that can be loaded by package maintainers
} 
} zmodload or pkg-config?

I meant a zmodload module.