September 24, 1991

Memorandum to:	Members of the X Consortium

Subj: U. S. Patent No. 4,555,775 (Pike), Issued to AT&T on November 26, 

In February 1991, AT&T sent letters to MIT and to members of the X
Consortium, notifying the community that the "backing store"
functionality available in the X Window System is an implementation of
a patented AT&T invention, and that consequently, companies or
institutions commercially marketing or internally developing products
based on an X Window System implementation will need a license from
AT&T.  Since MIT developed the X Window System and distributes it
widely, we were deeply concerned by AT&T's action, and we conducted a
thorough technical analysis of the patent and similar window system
technology.  MIT also retained counsel to provide us with an expert
legal opinion.  Both our technical analysis and the legal opinion
conclude that AT&T's claim with respect to the X Window System cannot
be sustained.

Patent #4,555,775 was applied for in 1982 by Robert Pike and issued in
1985 to his employer, Bell Laboratories.  It describes a variation of
graphics drawing algorithms that enables a single display screen to
support overlapping windows, in which several programs can be active
simultaneously.  It is our opinion that this same capability was
present in window systems that predated the Bell Laboratories' work.

The Bell Laboratories' work, the X Window System, and previous window
systems all use different variants of the same basic methodology
(storing obscured regions of windows as off-screen bit maps).  Both
our legal and technical analyses agree that each of the patent's
claims is either so broad as to have been anticipated by prior art, or
so narrow as to apply only to the `775 patent's specific algorithm and
not to the X Window System.

AT&T has requested the Patent Office to open the `775 patent for
reexamination, which will give AT&T an opportunity to clarify the
patent's claims.  We hope that AT&T will seek there to recognize the
contribution of the work on which it is based while not casting its
claims beyond the specific methodology involved.  MIT is very willing
to work with AT&T toward reaching this goal and is exploring ways in
which this can be done.

From a broader perspective, the confusion over the `775 patent
illustrates the difficulties raised by the increasing number of
patents issued on software technology.  We would welcome the
opportunity to join with AT&T, other companies, and the university
community to explore ways to alleviate these problems.  Software is an
area in which prior art is often difficult to identify and in which
obviousness differs significantly from one perspective to another.
The recent proliferation of software patents is creating pressure on
software developers to seek patent protection for straightforward
ideas, and increasing pressure to claim broad applicability of
software patents, well beyond the scope of any original invention.
These practices introduce a chilling effect on university research,
and they undermine our ability to pursue ideas in a free and open
climate.  In as much as these practices attack the university's
central goals and its obligation to disseminate information, they are
a matter of utmost seriousness to MIT.