software ethics

[Ed Gould]

I'm not a lawyer, either, but in the process of setting up a software
business I spent a lot of time talking to lawyers about just these questions.
Being a Unix business, too, we often discussed the particulars of the
AT&T license.

> It would help if people were a bit more careful about using phrases like
> "trade secret".  I'm sure a lawyer can do better, but let me try to define
> some relevant terms:
> 
> 	trade secret -- some knowledge that is truly kept secret by the
> 		owner, for commercial gain.  An example is the formula
> 		for Coca Cola(R).  It is *legal* to attempt to deduce
> 		a trade secret by reverse engineering, and to use this
> 		knowledge commercially.  Thus, if I buy a computer and
> 		find all the secret ROM entry points by disassembling
> 		a readout, I can write programs to use them.  I can also
> 		use the algorithms contained therein if they are not
> 		otherwise protected.

That's why ROMs are often copyrighted - see below.
 
> 	patent -- a limited-term right to use an original invention, in
> 		exchange for publication of the patent.  Because of the
> 		limited duration, and because essential information must
> 		be disclosed, many companies choose not to patent some
> 		items -- again, Coca Cola is a good example.  (Btw, the
> 		patent on Valium expired last week....)

[Agrees with my understanding.]

> 	copyright -- a very general sort of protection, applicable to a
> 		wide class of works, including computer programs and (now)
> 		ROM masks.  Copyright vests in the creator of a work by
> 		virtue of the act of creation; however, an appropriate
> 		notice must be affixed at time of publication or the work
> 		reverts to the public domain.  Copyrighted works need
> 		not be published to retain their protection; I've even seen
> 		some programs with explicit notices describing the file
> 		as an unpublished, copyrighted work.  If I write a program
> 		on my home computer, and you break in and steal it, I can
> 		sue you for copyright infringement.  Derivative works are
> 		also protected.  That is, if you write a book, no one can
> 		make a movie of that book withou your consent.  Copyrights
> 		expire, too, but after a much longer period of time -- I
> 		believe that current (U.S.) law specifies the life of the
> 		author plus 50 years.  Additionally, copyrights can be
> 		renewed under certain conditions.

Copyright protects "representations", not ideas.  Thus, algorythms may
not be copyrighted, although implementations may.

> 	contract protection -- in general, the owner of any object can convey,
> 		under any sorts of terms, more or less limited rights to use
> 		that object.  Most software falls in this category, combined
> 		with copyright protection to establish ownership.  That is,
> 		AT&T Technologies (formerly known as Wester Electric) *owns*
> 		the set of programs we know as UNIX.  In return for some
> 		consideration (i.e., large quantities of bucks), they'll
> 		sign a contract giving you certain specified rights to use
> 		their software.

Parties are free to enter into *any* contractual relationship that is
not prohibited by law.  That's how trade secrets are let out by their
owners.

> Now -- algorithms are in general not eligible for any of these forms of
> protection except trade secret.  (I assume that one can license a trade
> secret, but I'm not certain of that.)

One certainly can.  The contract by which it's licensed specifies how
the licensee is required to protect the secret.

>					 I just don't know what class a
> protocol falls in.  If protocols are in the same category as algorithms,
> then Lauren could *probably* look at some (copyrighted, licensed)

Copyrighted, yes, if the *implementation* were "substantially" different
from the original.  Licensed, probably not.

>								    UNIX code
> to understand it, then sit down and write his own version.  (Note:  before
> you do this, please remember again that I'm *not* a lawyer.)  On the other
> hand, he might have to prove to some big hairy lawyer that his code is not
> derivative from the protected code.  That can make for expensive lawsuits,
> and AT&T has lots of money and lots of lawyers...  And of course, whatever
> source license he's working under might contain some restrictive covenants
> that would bar such use -- remember that the UNIX source code is someone's
> property, and you can only use it with their permission.

Exactly.  The Unix license requires the licensee to protect AT&T's trade
secret.  They also specify, in a fairly reasonable way, what *is* secret
in the Unix code.  Not *all* of it is.  What is secret are the (I don't
remember the exact wording here) methods and concepts embodied in the
code.  Most, if not all, of the trade secret licenses I've ever seen
contain the release that anything the licensee learns from a third
party without restriction, or is found to be in the public domain,
is not part of the secret.

>							    I don't know what
> UNIX licenses say today; I do know that 5 years ago, an educational license
> prohibited using UNIX source code in the classroom.
> 
> You don't like this, you don't think it's right?  Well, what do you do about
> other property laws you don't think are right?  It's really the same thing;
> the only question at issue here is what is the definition of property.  If
> you believe in the concept of private property (and if you don't, please reply
> to net.politics....), I suggest that you lobby to have your own definitions
> accepted instead by Congress.  Personally, I do accept the concept of
> intangible property, including intellectual property; given that, I admire
> Lauren for declining to steal someone else's property.

I agree completely!  If, as our economic system and culture embrace, people
are to benefit from their labors (again, disagreements/discussions to
net.politics) then one must include programmers' labors, too.

> 		--Steve Bellovin
> 		ulysses!smb
> 
> P.S.  The opinions expressed herein are *mine* and mine alone, and do not
> necessarily represent the opinions of AT&T Bell Laboratories, its lawyers,
> etc.  And I decline to accept any responsibility for any actions anyone takes
> based on my understanding of the law.

-- 
Ed Gould		    mt Xinu, 739 Allston Way, Berkeley, CA  94710  USA
{ucbvax,decvax}!mtxinu!ed   +1 415 644 0146