PROSA II - PROtein Structure Analysis available

Lynn F. Ten Eyck teneyckl at sdsc.edu
Wed Feb 9 19:10:42 EST 1994


The issue of expensive software (or in fact anything but free software)
has generated more interest than anything else I have seen on this list
lately.  It is an issue near to my heart, so please excuse the length . .
.
or hit your kill key now.
	-- Lynn Ten Eyck

In article <1994Feb8.092347.11043 at abo.fi> Adrian Goldman,
GOLDMAN at ALA.BTK.UTU.FI writes:
> ..............................As far as
>I am aware most, if not all, of the programs being commercialised are
>developed using government funding (Austrian, US, etc., etc.) and with a
>committment to sharing "results".  If the results of some government
funding
>is a coordinate set, we are obliged to deposit it in the PDB for all to
use
>(sometimes without reference to the original work!).  If the result is a 
>piece of code, that person feels the right to copyright it and charge
>money for it -- which doesn't quite seem to me to be sharing.  
>
>The logic behind this escapes me.  It's not that I don't appreciate the
lure
>of money -- but might I make a Swiftian suggestion?  Maybe the coordinate
>generators should strike back by  creating a commercial repository of
>coordinates and charging a royalty for each use?  
>
>What do others think? - or am I being too old-fashioned?  When I started
>as a macromolecular crystallographer some 13 years ago, all code
basically
>came free....

Here is my $0.02 . . . Not in any particular order.

First, good software is not easy to write, takes considerable effort to
maintain, and often involves significant research to develop.  However,
it gets essentially *NO* career credit save for a very few individuals. 
My
reputation for software development rendered me unemployable as a
crystallographer.  George Sheldrick succeeds at this business because he
continually develops new things and new combinations of old things,
and while he is at it solves a ton of structures.

Second, despite this, there are people who spend a great deal of effort to
develop and distribute quality software.  They do it for personal
satisfaction,
because of strong opinions about how certain things should be done, and
out of social responsibility.  They also find these problems personally
interesting and satisfying to solve.

Third, it IS very difficult to obtain funds to pay for the development  of
software.  NSF will do it; Peter Arzberger in Biological Instrumentation
and Resources funds our work here at SDSC and Axel Brunger's work at
Yale, among others.  If you have a good idea for a program, try asking
parzberg at nsf.gov for funding -- but there is not a lot available.

Fourth, the government and the universities both strongly encourage
people to SELL software instead  of give it away.  The idea is called
technology transfer, and I think it just gets in the way.  Socially,
our society at large runs on money, but universities run on intellectual
credit.  Mixing the two gives rise to some pretty ludicrous situations,
like entrepreneurs confusing "intellectual property" and "trade
secrets" with easily observable phenomena just waiting for people
to trip over them.  Also look at the Apple Computer "Look and feel"
lawsuits.

Fifth, it is also difficult to get any money to buy software, even when
you need a commercial, industrial strength product.  This is one of the
first things (along with computer maintainence) to get deleted from
the budget.

My personal opinion (not the rules under which I sometimes have to
operate) is that if the development was paid for with government funds,
the results ought to be free to all.  This specifically includes somebody
who wants to improve it and sell it, provided they are willing to put
their improvements up against the competition of the original free
version.  However, you might consider the analogy of a person who
writes a book while on grant support . . .

Incidently, there is another issue with commercial packages -- you
can't check out the source code.  If you get a strange result, how can
you validate it?  I *REALLY WORRY* about this one.

There is also the issue of the two price system -- industrial users, some
of whom are very innovative software producers, are charged a fortune
for programs which the rest of us get very cheaply.  I don't think this is
fair unless they also get very good software support services, manuals,
and so on.

In an "ideal world" the Free Software Foundation would rule, source code
would always be available, and people would be able to get money to
support the development and distribution of software.

All of that said, I should also mention that the aforementioned Peter
Arzberger (and before him, John Wooley in the same job) do support us
for the development and distribution of software.  All things are
possible in time . . . If you want to see some of the results of that
support,
check out XtalView by sending the message "get xtalview" to 
ccms-request at sdsc.edu.

BTW, XtalView also deals very comprehensively with the human interface
issues mentioned in other articles in this thread.  It can be run either
from
windows or from the command line.  The windows interface does a lot more
than just feed the input files.

Lynn Ten Eyck
teneyck at sdsc.edu




More information about the Xtal-log mailing list