DCC is patented

Vernon Schryver vjs@calcite.rhyolite.com
Wed Oct 27 19:33:32 UTC 2004

> From: "John R Levine" 

> I recently learned about patent 6,330,590.  It was filed in Jan 1999,
> which means the work it covers was done in 1998, and fairly broadly covers
> spam filtering by counting message signatures.
> A PDF copy of the patent is http://www.taugh.com/6330590.pdf
> If you haven't read patents before, the first part describes the invention
> but isn't legally very important.  The crucial bits are the claims which
> start on page 6.  Claims 1 and 2 and claim 30 look a lot like DCC.

I disagree.

I've been involved slightly in some patent disputes unrelated to spam.
What I was taught by expensive lawyers is that what matters are the
claims, and that reading patent claims is not as easy as seems it
should be to computer programmers or mathematicians (my background).
A set of claims is like a first order predicate expression of primitives
connected with ANDs and ORs, but with both the primitives and connectives
obscured as much as the Patent Office will allow.

My reading of that set claims is that it does not cover the DCC.
There are often important issues in patent disputes in addition to
prior art, dates, and what the claims really cover.  I suspect at
least one of those other issues would be relevant should this patent
become an issue for the DCC.

I won't discuss my reading of that patent's claims or any other issues
in public for two reasons.  It is quite likely that I'm wrong, because
I'm not a lawyer.  The bigger reason is I have also been told by
seasoned experts that, as in most legal circuses, it is vital to keep
quiet.  If you know about prior art or problems in patent claims, you
never tell the opposition until absolutely necessary.  You must listen
only to your own lawyers about what to say and when.

As with all legal issues, the best and only generally valid advice is
consult your own lawyers.  If you are selling spam filter software or
have other reasons to be concerned about spam patents, you should have
already done patent searches and already know what your lawyers think
about this as well as other anti-spam patents.

> Anti-spam vendor Commtouch recently bought the patent:
> 	http://www.commtouch.com/news/english/2004/pr_040913.shtml
> 	http://www.internetnews.com/xSP/article.php/3407601
> They say they're going to start selling licenses.  Uh, oh. How long ago
> did you start work on DCC?  If it's older than Jan 1998, you're OK.  If
> not, this guy probably got there first.

This guy got to his patent first, but I'd wait for a court or at least
my own lawyers to say more than that.  (not that I am going to pay
from my own purse the substantial money to get a competent opinion.)

Before pre-emptively shutting your DCC client or server, it might be
good to look at things from the point of view of the patent holder.
The DCC works better as more systems particpate.  Anything that reduces
participation in the global DCC network might reduce the value of the
patent.  If I owned that patent and had a financial stake in spam
fighting, I might use the patent to encourage cross licensing.

] From: Ruben Safir Secretary NYLXS 

] Such a broad patent will never hold up in court, along with the other 80%
] which are found invalid.
] Screw the patent.

I'm even less convinced of the wisdom of that view than the view
suggested by John's words.  Never mind my contrary impression that
more than 80% of patents are upheld.  All litigation is always too
expensive, but only people or organizations without assets can afford
to blithely ignore summons.

} From: Ruben Safir Secretary NYLXS 
} > Actually, this patent looks pretty good to me.  I don't know of anyone
} > else who was doing signature counts in 1998.  Do you?  Broad patents hold
} > up fine if they're really novel.
} It looks even worse than the SCO/IBM suit among many others.
} It's just too broad.

I thought that breadth makes a patent potentially more valuable without
necessary affecting its validity  either way.  Besides, I think the
SCO suit is more about copyrights than patents, which is unsurprising
if you know a little of UNIX history and famous UNIX patents.

} > You don't have any idea how much it costs to defent a patent suit, even if
} > you win, do you?
} Yeah.  I've defended against them and won.  Do you think the cost of
} defending a patent is a reason to submit to an unreasonable assertion.
} Fuck That.  

I doubt there is anything one can say about the cost of patent
litigation besides "lots."  Both sides must be willing to keep
spending before it can cost millions.

| In fact, I'll personally donate a substanial amount to the defense
| and put you in touch with an outstanding IP Lawyer.

Unless you are as rich as Bill Gates, that offer, generous as it
is, will not sway everyone.  Some patent suits have cost many tens
of millions of dollars.  I would be surprised if passing the hat
could get enough money to buy a week's worth of a lawyer's time,
and never mind the years that could be required to win.

One of the reasons for that dour view is that I have no financial
interests in fighting spam.  If my personal experience is any guide,
there is less than no money to be made fighting spam.  Judging from
their actions, some people disagree.  Good luck to them.

Vernon Schryver    vjs@rhyolite.com

More information about the DCC mailing list

Contact vjs@rhyolite.com by mail or use the form.