🚀 go-pugleaf

In article <1992Apr23.001851.10636@algor2.algorists.com> jeffrey@algor2.algorists.com (Jeffrey Kegler) writes:
>   My understanding (I am not a lawyer) is that if you simply wish to
>   defend against patents, you clearly establish prior art, and this will
>   invalidate any attempts to patent your invention.  It's a lot cheaper,
>   too.

In article <JTW.92Apr22213301@mercury.lcs.mit.edu> jtw@lcs.mit.edu (John
Wroclawski) writes:
>It's important to understand that this is true in theory, but not
>necessarily in practice. The PTO's definition of prior art is strongly
>biased towards "previously issued patents". This is no surprise: the
>patent database is what they know. It's well-indexed, organized the
>way the PTO wants it to be, and the model works OK in fields where
>most important advances are patented.

John makes many other good points in his article.  Unfortunately,
small companies are generally telling the truth when they say they are
getting software patents for defensive purposes; at the rate that software
patents are increasing, you soon won't be able to play unless you have
something to deal with at the cross-licensing negotiation table.

Given this, I don't think that organizations like LPF should be hostile to
someone for the mere act of getting a software patent (even though I think
they should be abolished).  The heat shouldn't be turned up unless an
organization attempts to get other products withdrawn from the market by
suing for patent infringement.

In fact (though I expect not to be listened to on this one), I think that
FSF should get software patents and engage in cross-licensing agreements;
given their general approach, I suspect that organizations like IBM would
be willing to play.  Once Unisys found themselves stepping on an FSF
patent, the most prudent course for them might just be to cross-license
the LZW patent for whatever FSF patent they needed access to.  While this
is a worse world than a world without software patents, it might be the
best possible world if the GATT treaty (with its mandatory software
patents) goes through.

The proposed FSF policy would go like this:

1.  Anyone would be free to use FSF patents in any freely redistributable
code (freely redistributable meaning GPL or the more permissive agreements
like BSD's or X11's).

2.  Commercial organizations could use FSF patents in proprietary software
if and only if they sign a cross-licensing agreement permitting FSF to
distribute programs using any patent belonging to that organization, that
is, access to all FSF patents in exchange to access to all Company XYZ
software patents.
--
Joe Buck	jbuck@ohm.berkeley.edu