Steven Brenner seb1005 at
Wed Feb 3 08:03:15 EST 1993

In article <BASHFORD.93Feb1234155 at>, bashford at (Don &) writes:
|> Going farther, I would argue the same for all types of naturally
|> occurring nucleic acid and protein sequences and their
|> three-dimensional structures.
|> Going even farther, I would argue the same for the source code of
|> computer programs whose output is considered significant enough to be
|> the substance of a serious scientific publication.  This is on the
|> grounds that the source code constitutes a description of one's
|> methods and should therefore be open to inspection by anyone.
|> Donald Bashford
|> Assistant Member
|> Department of Molecular Biology

Actually, I would argue that the very REASON that you don't see source 
code being distributed widely enough is that there it CAN'T be patented.

When you patent something, you make a "deal" with the govenment: they
give you exclusive rights to make a profit out of your work for a certain
number of years.  In return, you show everybody and anybody precisely what
it is that you've done--and give them the opportunity to do you one better
(or, in the worst case, do as good as you in 18 years without having to
spend any effort on research).

With software, however, patents are de facto no applicable.  Consequently,
there is really no way to protect the ideas in your software other than to
keep it as a "trade secret."  (You can protect the expression of your
ideas via copyright, but the ideas themselves are--unless patented--free
as air.)

Actually, there is a third way of protecting your ideas: publishing them
where a lot of people will see them.  You don't get any financial 
protection for having done so, but at least your colleagues give you credit
for coming up with the idea.  (And you also prevent anyone else from ever
patenting the idea, but they can still make money off of it.)

However, if I write a piece of software to do some sort of biological analysis
and then publish a paper on it, I am extraordinarily unlikely to include
in the paper *every* idea which is in the program, in part because many of
those ideas--while necessary to the program's operation--are of neglibible
interest to biologists.  Therefore, distributing this program means that
I essentially loose all rights to some of the ideas embodied therein, but
receive nothing in return.

I would therefore argue:

	1)	The amount of scientific software which is distributed
		freely is surprisingly large!  Those scientists who have
		distributed their software really should receive an 
		extra level of appreciation from all of us for having
		essentially given us their ideas.  Kudos to them!

	2)	Via law or regulation, better methods should be found
		for protecting intellectual property as embodied in


Note: these are just off-the-cuff ideas.  For the record, I do hope to
distribute most of the software that I write.  Please, no flames, but I'd
like to see dissenting views on this point

Steven E. Brenner               |  Internet    seb1005 at
Department of Biochemistry      |  JANET       seb1005 at
University of Cambridge         |  Laboratory  +44 223 333671
Tennis Court Road               |  Home        +44 223 314964
Cambridge CB2 1QW, UK           |  Lab Fax     +44 223 333345

More information about the Bioforum mailing list