Re: RE-PROPOSED: The Dictator Test
On Mon, Jul 12, 2004 at 08:23:22AM -0400, Michael Poole wrote:
> Branden Robinson writes:
> > If an innocent bystander is harmed through the operation of defective Free
> > Software, how can he or she be held to the warranty disclaimer, given that
> > he or she never received the corresponding copyright license?
>
> Can you elaborate on the situation you have in mind? I would think
> that in general, the liable person would be either the one who put
> Free Software in the medical device (construction machinery, etc) or
> the one who operated the device (the immediate actor might be an agent
> for a company that would be liable). Software by itself is pretty
> limited in how it interacts with the world, so it needs some enabler
> before its bugs could harm anyone.
"Shotgun suit" comes to mind immediately. Throw a summons at anyone vaguely
related to the problem at hand, and hope some of them stick. If you thought
your warranty disclaimer would protect you from any and all lawsuits, you're
in for a nasty shock.
More reasonably, in the absence of a binding disclaimer of warranty, the
usual rules of due care and whatnot come into play. If the injured party
can show that the software was dodgy, they *can* come after you directly, as
you were a contributing party to their injury. No cites off the top of my
head, but it's common to see several parties to a suit -- the operator
personally, the company the operator was working for (if any), and the
manufacturers of the equipment involved (again, if any) if it was
possibly a partial cause of the injury.
The injured party is in no way bound to your warranty clause, so you still
have regulation due care to ensure that your software isn't full of
nasssssty bugs. You *might* be able to palm it off onto the person who
deployed your software, saying "they shouldn't have used it because it's
unwarranted", but (i) that's a bad argument for free software (and software
in general); (ii) your warranty clause could be struck as being contrary to
the public interest, and (iii) you'd likely have to prove that the
deployer's QA procedures were woefully inadequate.
IANAL, NDIPOOTV.
- Matt
Reply to: