Databases Reference
In-Depth Information
The patentability of software has been evolving over the last 40 years. In 1972,
the Supreme Court ruled that algorithms could not be patented, because they were
a form of mathematics and as such were a completely abstract idea (Gottschalk v.
Benson, 409 U.S. 63 (1972)). The thin edge of the wedge was inserted in 1981 by
the case Diamond v. Diehr (450 U.S. 175 (1981)), when the Supreme Court allowed
a patent for a process to heat and cure rubber that included a computer program.
Another important case was State Street Bank & Trust v. Signature Financial Group,
Inc. (149 F.3d 1368 (Fed. Cir. 1998)), which permitted business processes and the
software that implemented them to be patented. Since then, software patents have
generally been allowed if the computer program were part of a business process or
a machine.
Despite their legal status, software patents remain controversial [
1
,
5
,
7
]. Many
software developers are adamantly against patents and believe they should all be
invalidated. It does not seem sensible for relational databases to be patented (U.S.
Patent 4,918,593), when there are undergraduate courses on the topic. Fortunately,
the patent on relational databases was granted on April 17, 1990, which means the
patent expired in 2010. However, improvements to the original patent may still be
in force, so we might not be entirely out of the woods yet.
Another example is U.S. Patent No. 8,406,721, “Unlocking a Device by Per-
forming Gestures on an Unlock Image” was granted to Apple in October 2011 [
4
].
A consequence of this patent is that no one else can use that slide gesture on a touch
screen to unlock a device. While this mechanism may have been novel in 2009, when
the patent was originally filed, but slide-to-unlock has since become commonplace
on mobile phones. There are many cases worldwide on this and other patents for
mobile and internet technologies [
15
].
At time of writing, the courts have not yet decided whether Samsung has in-
fringed on Apple's patent on slide-to-unlock. Therefore, we have no decided case
law to base our analysis. Nevertheless, it is sufficient to say that Lucy in our sce-
nario is treading on dangerous ground and that her invention is sufficiently similar
to warrant litigation by a patent holder, which is bad news for her. If Apple's patent
on slide-to-unlock is upheld, Lucy may have to change her design or seek a license
from Apple, which could prove expensive.
We have little helpful advice for Lucy and other software developers like her,
who must continue to create new software and earn a livelihood doing so. But this
is becoming increasingly difficult with the growing number of software patents. In
2010, the U.S. Patent Office granted 35,710 software patents [
16
]. With concerted
lobbying efforts by commercial industry groups, patents that were once relatively
difficult to obtain for software-based inventions are becoming more mundane.
Software changes quickly. Many inventions that were novel at the time quickly
become standard practice, such as scroll bars on a window, b-trees for file-based
data structures, or even a binary search algorithm. But when these inventions are
patented, there is no allowance for them to become scènes à faires. Where do we
draw the line between a reasonable patent and a soon-to-be best practice? Twenty
Search WWH ::
Custom Search