[Patents] Publishing protects you from being sued?

[epa98@doc.ic.ac.uk (Edward Avis)]

On Wed, 9 May 2001 patents-request@liberte.aful.org wrote:

>From: "Jozef Halbersztadt" <halber@timsi.com.pl>
>Subject: Odp: [Patents] Limited Software Patentability

>To achieve one of your aims - not allow the others to patent your
>innovation - you don't have to patent yourself. It is enough to publish in
>good place.

If only that were true.  Many software patents are granted on old ideas,
ones which have been around for decades (the now-expired XOR patent is a
good example).  Patent examiners don't take into account most of the
body of published software.  I don't blame them for this, considering
that a million lines of code is written every day - even if only a tiny
proportion of that is published, it's still beyond the means of any
patent office to search through it all.  (In fact, they cannot even
search previously issued patents - eg LZW was patented twice.)

It is true that publishing will help you to get a patent declared
invalid once it gets to court.  But that is not much protection if you
cannot afford the legal bills in the first place.

So publish, it may help.  But it does not guarantee you legal security
to develop your application without being sued.  For that, the only
remedy is to decide that computer programs cannot infringe on any
patent.

-- 
Ed Avis
epa98@doc.ic.ac.uk