Subject: Re: Patents and Amazon?
From: Mark Shewmaker <>
Date: Mon, 20 Mar 2000 22:25:41 -0500

On Sun, Mar 12, 2000 at 05:15:31PM -0800, Tim O'Reilly wrote:
> That being said, I think that the most fundamental tool we have at our
> disposal doesn't depend on changes in patent regulation or new laws, but
> one something that is very much under OUR control:  we have got to
> create some kind of prior art database that serves as an easily
> searchable record of our collective memory.  This might include
> technical papers from organizations like ACM, IEEE, and Usenix, as well
> as various conferences, but it should also include searchable
> documentation, usenet, and the web.  

The Software Patent Institute at has been doing a subset of
this for years.  Perhaps they could expand their activities.

(But I have to admit confusion at their apparently contradictory stance
of trying to keep up a public database of prior art while also keeping
that same information a trade secret via an "I'll not redistribute any
material" click-wrap agreement.  I would hope they would be willing to
drop some of the more obnoxious requirements were they a real impediment
to growth.)

> I also think we could do a lot with the "comment period" thing. 
> Apparently, this is already something that could be done, since patent
> pending applications are apparently supposed to be made public 18 months
> after the application (which is usually before the grant.)

I'm very scared of the new "comment period".

Forgetting the expense involved with the possibly-necessary court cases
for the moment, after a patent is granted prior art can still be used to
invalidate the patent.  However, if someone brings that same prior art
to the attention of the USPTO before the patent is granted, that simply
gives the so-called inventor an opportunity to change the patent
application so that it can still be valid.

The result is that the patent-holder gets a stronger patent than would
otherwise have been awarded, probably on what to us would be necessary
and obvious implementation improvements.  What's worse, we couldn't
again use easily use that prior art to invalidate the patent, as it
would be assumed that the USPTO properly considered it during the
"comment period".

>  And even
> after grant, a site that allowed community comment on patents
> (especially one that asked for supporting data rather than just
> assertions) would provide both potential litigants and defendants better
> means for assessing their chances of prevailing were there to be a
> patent lawsuit.

That is part of what Greg Aharonian plans to do at

(Parts of his plan are somewhat controversial, btw.  I believe it
involves the idea of shorting the stock of companies that seem to be
over-valued because of patents that he or a client thinks are invalid,
and then working to invalidate those patents.  I personally don't see
anything UN-ethical with the idea of making money by exposing bad
patents.  It would be important to make sure the plan doesn't run afoul
of SEC regulations, but assuming it can be done I think it's a great
idea.  (But I figure you'd want to be aware of this controversial side of
his plan before publicly saying anything that could be construed as a
full endorsement of the idea.))

> But of course the biggest thing is to create an atmosphere of
> uncertainty around the public relations cost of attempting to enforce
> bogus patents.

I think it can become obvious to companies (more below) that the
offensive use of patents can be costly on accounting sheets as well, and
not just because of bad public relations.  I also want to promote a way
of turning their bad public relations good.

> I had a really interesting interview with a reporter from one law
> journal (IP Law Weekly or some such), who told me all the reasons why
> none of my suggestions would work, but who started to warm up when I
> brought up allusions to the environmental movement; that just as they
> had made people realize that there were certain common rights that
> superseded individual property rights, so too there were rights
> associated with the "internet commons" that has been built up by the
> cooperative community over the last thirty years.

To go with the "internet commons" theme for a bit:

The Tragedy of the Commons deals with the notion that an unowned shared

   o  from which anyone can take freely, and
   o  for which no requirement to assist with the upkeep is imposed
      upon the taker, and
   o  for which the taking lessons the resource in some way,
      (ie, most physical resources),

will not fare as well as one which is owned and for which the owner
has an incentive and the possibility of maintaining the resource in
some way.

But for intellectual property, the commons are not lessoned by the
takings, and in fact the takers sometimes have internally-imposed
incentives to improve the commons as well.

It may be the case that by forcibly dividing up the commons so that
fewer people are responsible for the upkeep, that the new owners will
have more of an incentive to maintain and improve the resource, or it
may be that that's not the case.

In the case of software patents especially, I personally think that
dividing up these intellectual commons will slow upkeep and improvements
in the field of software.  I think software patents are an example of a
Reverse Tragedy of the Commons.

(Quick rant:  I also think this sort of thing can easily happen when
governments create property rights out of thin air, where those property
rights don't really have a properly grounded basis economically or
morally, but that's a different issue.)

So if the splintering of a particular type of shared resource harms
maintenance and slows development of that resource, then coalescing the
resource back together again should improve matters, although that might
be difficult to prove, even after the fact.  (ie, if we could change the
laws to make software patents invalid, could we prove to the
satisfaction of those that currently strongly dislike software patents
that this was a good thing?)  But if the newly re-shared resource were
only compatibly shared between like-minded parties, thus forming
separately identifiable sharing and non-sharing groups, then I would
think that in any comparison between the sharing and non-sharing groups
that those who exercised their option of sharing would be seen to
benefit more than those who didn't, and that more and more parties would
decide it would be in their own interests to switch over to the sharing

That's the sort of thing that I want the Open Patent License under
development at to do.  If those who hold patents
defensively let each other use their defensive patents in products that
contain no non-defensive patents, they should have an advantage over
those who are at greater risk of patent infringement lawsuits.

(Also, the license allows the use of the submitted patents in Open
Source products, given some conditions.  Having the patents available
for use in either Open Source products, or proprietary but Open Patent
products mean that another participant could still either use, or if
necessary, reimplement a competitor's product.  Open Patent products
shouldn't give one participant a monopoly right over another.  Instead,
companies should benefit more as they continue to abandon their patent
and patent-like IP monopoly powers over other participants.  That
basically means that a one-size-fits-all license won't do.  Under this
license, there are multiple Options under which to submit IP, and
multiple Pools from which to use it.  That makes the license longer,
but I think it makes it a bit easier to understand and hopefully easier
to debug, and I think it will make it easier for companies to slowly
adopt the license once it reaches version 1.0.0.)

If the license can show by implication that software patents are harmful
overall by the eventual willingness of participants to "opt-out" of the
patent system when safe to do so, I would think that would help efforts
to get rid of software patents altogether.

 -Mark Shewmaker