Re: PHPNuke license
On Mon, Mar 03, 2003 at 03:53:22PM -0500, David Turner wrote:
> Actually, I think "Copyright 2003, FSF and others (see file /foo/bar for
> details) [no warranty]" would be an appropriate copyright notice. So,
> there's a minor problem, but not an unbounded problem.
> I'm just not sure I see an actual case when this would happen. One
> issue would be that any shell scripts relying on the output of the
> programs would have to be changed. This would be enough work to disuade
> any but the most determined from making this change. In cases other
> than coreutils, where far fewer shell scripts rely on output, the
> problem for everyone else is much smaller too.
Generally not, if the annoying text was output to stderr (or when on
a tty--which it will almost always be when interactive--even more safely
but even more annoyingly, to /dev/tty).
> Even if there is a problem, it's not even on the order of the BSD
> advertising clause -- it doesn't make the software non-free. And if
> there is a problem, it's a genuine problem which ought to be fixed.
A string of piped commands might output five such notices; a foreach
loop might output hundreds.  I agree that it's not on the order of the
BSD clause; at least we can disable it, though we might have to research
how it's done differently in dozens of packages. I believe it's
potentially extremely annoying, but (unlike the BSD clause) it generally
hasn't been, in my experience. ("Hasn't been" isn't "will never be",
Forcing me to mention the copyrights of underlying tools on my webpage
(or the existance of underlying tools at all, for that matter; users
don't care and can ask if they do, so don't bloat my pages) is orders of
magnitude more annoying, though. It's extremely non-free, in the view of,
well, my own subjective opinion.
 Continuing the idea that using these programs in a complicated
but user-typed shell string is still "interactive"; it's probably exactly
this problem that the "interactive" qualification was intended to prevent.
However, recalling the context, Branden's argument was, I believe, that if a
web session is interactive with respect to the tools generating them, then
manual shell scripting is, too.