Subject: Re: Exploring the limits of free software: Cygnus, and GPL
From: craig@jcb-sc.com
Date: 24 May 1999 08:18:52 -0000

>Furthermore, the "Doomsday scenario" with truly unowned (public
>domain) text, that someone will take it, alter it slightly, copyright
>it, and then be in a position to enforce copyright against future
>users of the original text, is surely not possible precisely because
>BSD-licensed code is owned.  (I don't know that the Doomsday scenario
>is actually a legal possibility, it's just what I was told was the
>reason why BSD-like licenses are needed.  Seems more likely that the
>advantages are (a) they provide convenient documentation against the
>Doomsday scenario for the legitimate users (== everybody) and (b)
>require attribution to the owner.  Any lawyers out there know?)

That's not a genuine "doomsday scenario", in the sense that, as long
as people who use the original text can show that they got it from
a genuinely PD source, they're fine -- able to continue using the
work, and in fact able to continue distributing copies of it to
others.

I wouldn't call *this* a doomsday scenario per se, but it's pretty
bad: what really happens is that the people who still have the PD
source (this is pre-Internet, mind you) end up with much less incentive
to share it with others *as* PD, once a 3rd party has turned it (or a
portion thereof) into an expensive proprietary product.  After all,
making copies is expensive, and while it might previously have been
done on a "friendly" basis, seeing someone else make piles of money
off their *copyrighted* versions can make the possessors of the PD
version rather, well, downright unfriendly about continuing to make
those expensive copies.

The end result is that the provably-PD copies of the work are
like money under peoples' mattresses -- not in circulation, not
benefitting anybody during that time, and easily lost due to
accident, death, whatever.

The Internet makes this less of a problem for the kinds of things
we talk about (software -- the percentage of significant software
being published as PD per year that is *not* published via the
Internet surely has declined to near zero in recent years, I would
think).

Nevertheless, think about this: the same organizations (web sites,
for example) you might trust to preserve the historic integrity of
the PD release of a work over the decades have, in fact, no *legal*
requirement to do so.

The mere fact that they're PD means those organizations can, years
later, eliminate their archived postings saying "Hi, I'm Mike, here's
the source code for my web browser, it's public domain", and instead
offer the exact same source code under those organizations' copyrights,
or even offer only the binaries derived from that code, complete
with proprietary-license restrictions.  (I'd like to think this won't
ever happen, but it's good to remember it can.  E.g. MS could buy
Dejanews and eliminate every "hereby put in Public Domain" post,
legally "hijacking" the works for its own use under its own copyright.)

Your "Seems to me" is fairly on-target.  Having a legally recognized
instrument recording ownership helps.  Having it usable (in most or
all countries) as a means to provide incentive for distributing copies
of the work helps as well.  The BSD, AL, GPL, and other licenses all
do this to one extent or the other (though the incentives, and their
degrees of value, differ).

        tq vm, (burley)