From me in 2001 posted to gnu.misc.discuss: https://groups.google.com/d/ms...
I definitely do not want to see a future world of only proprietary
intellectual property where basically everything I want to do requires
agreeing to endless licenses and royalty payments, such as described in
"right-to-read". My wife and I released a six person-year effort under
the GPL (a garden simulator application) around 1997 ...
so I am obviously sympathetic to encouraging free sharing of some
information and allowing derived works of some things.
However, on a practical basis, living in our society as it is right now,
any software developer is going to handle lots of packets of information
from emails to applications to program modules under a variety of
explicit or implied licenses. If a developer is going to do this in a
way that makes his or her work most useful to the community (under the
terms he or she so chooses), proper attention must be given to the
licensing status of all works received and distributed, especially those
that form the basis for new derived works to be distributed. Note that
even in the case of purely GPL'd works, one still needs to know that a
user contributing an extension to a GPL'd work was the original author
and/or he or she has permission to distribute the patch (if say an
employer owns all the contributor's work).
My question is: should software tools, protocols, and standards play a
role in easing this required "due diligence" ...
license management work (at least as far as copyright alone is
concerned)? ... Usually license management tools (e.g. for music or DVDs) are thought of
as keeping the end user from doing something they might wish to with
content they have paid for. Does it make sense as well to look at
license management tools from the perspective of allowing
(non-technical, non-lawyer) casual users to do things they otherwise
might not be legally sure they can do? Similarly, would such tools help
someone filter out proprietary content with licenses he or she does not
approve of (and would this provide incentives for artists to release
free versions if they want to reach people through those filters)? And
most of all, would such tools allow creative people to be more certain
that they could legally use certain freely licensed materials found on
the internet in making derived works? Would this provide a legitimate
defense of due diligence to minimize copyright infringement suit costs
(or reduce related liability insurance costs)?
For example, when you get an email it could come with a machine-readable
license (e.g. "redistribution OK in entirety", "for your eyes only",
"open content", "GPL"). Likewise, what if every file or zip archive came
with a specific machine-readable license? In effect, this would make the
license a fundamental part of the work.
In part, you may think, perhaps correctly, this it the "right-to-read"
nightmare. Such information could be used to prevent you from making
copies of things you might want to copy (legally or not) under some
notion of "fair use" ...
if the system enforced the license by preventing say you forwarding or
quoting an email that comes in with a license of "for your eyes only" or
with no explicit license at all. Perhaps the feeling that copy
protection systems will prevent fair use underlies much of the
resistance to such automation. It is not my point in this note to
advocate either for or against the enforcement of licenses by the end
user's system. Obviously though, enforcement would certainly be made
easier by machine-readable licenses, and this is a problematical issue
as far as "fair use" is concerned.
On the other hand, license management tools might force everyone to be
explicit about licenses for things they redistribute. Some authors would
explicitly choose free or open licenses. That might mean that when you
get free software (or open source software or anything else) you would
know what you at a minimum can and can't do with it. That clarity and
sense of peace of mind might help promote use and more derived works.
For example, even if MIT puts its course material on-line, that does not
necessarily mean you can make derived works from them or even share them
with a friend (other than by telling them to look at the MIT site). Yet,
without a good free license management system, that fact might not be
obvious to users and a truly free course library may never arise. (Note:
I don't know whether the MIT courses will permit derived works, so MIT
may surprise me.)
A LICENSE REJECTION PROTOCOL
Being explicit about licensing (especially in a machine-readable way)
may have great benefits. For one thing, you might decide to set your
email receiver to reject email from most people unless it came with an
acceptable (to you) license. There might be a "license negotiation"
protocol at the start of all transmissions of all works.
Sender: PERMISSION TO SEND "Windows NT Source" BY "misguided kiddy";
Receiver: WHAT LICENSE?;
Sender: LICENSE: NO-REDISTRIBUTE-39;
or perhaps instead:
Sender: PERMISSION TO SEND "GNU/Linux kernel mods" BY "Linus Torvalds";
Receiver: WHAT LICENSE?;
Sender: LICENSE: GPL-2;
If you ran a peer-to-peer file server, such a protocol might help ensure
only legally redistributeable works were redistributed on it (making it
legally safer to run one). Obviously, people could lie about the license
status of works when they inject them into the system -- but the point
is, it forces such people to explicitly lie, as opposed to just being
careless or neglectful. (Obviously, carelessness and neglect could
affect the system as well if the person injecting the information is
just confused, hopefully other factors like community awareness could
minimize this.) Nonetheless, it might gives users a legal defense from
extreme copyright infringement awards if they screen incoming data. This
in turn might make insurance for such situations affordable. Defenders
of such a file sharing system (in court) could then admit to there being
a few "bad apples" and take efforts to route out such illegally
contributed material in the same way people now use virus scanners or
other filters. This might make it more likely such systems would
prosper, with other attendant benefits for democracy or an open society.
To be clear: I personally am not for supporting sharing of material that
for legal or copyright reasons can't be shared (it's the law; change the
law peacefully if so desired). I instead want to make sure that it is
easy to share material that it is legal to share, and likewise I want to
ensure it easy to make derived works with clear legal titles from
material it is legal to make derived works from.
In the case of software, with such a system, when you build free
software packages (or "open source" ones), you could ensure that all
contributions were under an acceptable license, because that licensing
information would be already there in a machine-readable form (perhaps
including information pointing to works and their licenses from which
you made derived works). Presumably, if someone emailed you a
contribution using such a system, you could see at a glance from the
email record what license it (or the code part) was under. In addition,
information could also come along that was the equivalent of a statement
of either originality for the work or a statement the author had
permission to incorporate other works they used into the new work under
the license chosen. Such information might include an audit trail of all
works and licenses used by various authors in making the final product." ...