NetNews Usenet Archive 1992 #26/NN_1992_26.iso/spool/comp/org/eff/talk/6918

Internet Message Format | 1992-11-10 | 14.4 KB

Xref: sparky comp.org.eff.talk:6918 misc.int-property:1422 alt.suit.att-bsdi:609 comp.unix.bsd:8704
Newsgroups: comp.org.eff.talk,misc.int-property,alt.suit.att-bsdi,comp.unix.bsd
Path: sparky!uunet!think.com!spool.mu.edu!decwrl!csus.edu!netcom.com!mcgregor
From: [e-mail] (Scott Mcgregor)
Subject: Re: Software Patents: Promotion of science and the useful arts?
Message-ID: <[e-mail]>
Organization: Netcom - Online Communication Services (408 241-9760 guest)
References: <[e-mail]> <[e-mail]> <[e-mail]>
Distribution: usa
Date: Wed, 11 Nov 1992 00:56:34 GMT
Lines: 267

In article <[e-mail]> [e-mail] (Joe Kelsey) writes:

>Since people who use the Internet and Usenet actually *pioneered* the
>use of electronic mail and also created most of the standards used
>today, I think we can clearly state that if RSADSI and PKP had not
>actively prevented the implementation of public key cryptography, a
>standard PEM could have existed years ago.  We might even have
>privacy-enhanced netnews!

PUBLIC key cryptography is not necessary. Private key cryptography
could provide the same benefits. Even PKE systems are becoming hybrids
with private keys used for authentication to servers, etc.  And we
have a Private Key standard, DES. Why isn't it in every mail system?
Why isn't it in every mail system today?  Why isn't it in net news?
No I think Joe trivializes the other factors.  Encription is often
suited to its requirements. Netnews DOES have encryption, rot-13 is
sufficient for the purposes it is used for in rec.humor and various
movie reviews.

Electric cars are possible. Some aspects are patented. But that is not
the primary reason that they aren't in more common use, even though
they cause less vehicle air polution.

>  The very existence of the medium we use to
>discuss this issue shows how standards arise from the very real need
>to interoperate with a wide variety of computers in a network.  Your
>argument about too many different implementations preventing
>standardization has no basis in the history of the very networks you
>use now!

As someone who was active in working out some of the standards for
interconnecting UUCP (! format) mail networks with ARPA Internet
descendent networks (@ format) and proprietary systesm (IBM Profs, HP
DeskManager, MCI ), as well as someone who worked with interconnections
between net news, net notes, Confer, and Gnotes, I can assure you that
there is tons of evidence in the appropriate discussion groups in the
archives that this has been and continues to be a problem. It was a
major problem at HP when I was at their corporate offices trying to
help build an HP internet, and to connect that internet with the rest
of the world.  Tens of thousands of people have regular difficulties
posting, or replying or exchanging mail with other seemingly
interconnected people.  The knowledge to make it work all the time is
beyond most system administrators. This is impeding many people from
taking advantage of this interconnectivity.

>However, the history of PKP shows an agressive posture over tha past
>ten years or more.  Only recently has PKP softened its tone and now
>suddenly allows the so-called personal use of the RSA protocol.  Why
>did they suddenly change their tune?

Perhaps because business conditions changed? Maybe because more people
have systems capable of the performance necessary for such a system
today than 10 years ago, and more of those PCs are now networked
together? And because more people are now using computers to
communicate private information across insecure media?

>  I think that they finally
>realized that their policies have actually prevented any products
>based on their precious patents from gaining widespread acceptance and
>now realize how much they have hurt themselves and everyone else in the
>business!

Possibly. Or maybe they go the timing right and widespread acceptance
wasn't likely ten years ago because of the scarcity of powerful
networked PCs then. But is the point that because some companies make
business decisions we think are boneheaded that they should not be
granted a patent and thus saved from their stupidity?

>The policies of RSADSI and PKP show how a patent holder can prevent
>innovation in the software industry.

What innovation was prevented?  I can see how copying RSA's Public Key
ideas was prevented. But that's copying an already known idea, not
innovation. If someone else came up with a different, and better, one
way trap door function to use in a PK system it could have been
patented too (TRW?). Then they could have cross licensed with RSA.

If RSA is so good, why haven't people licensed it at the prices that
RSA was offering them licenses?  Why wasn't it a bargain?

What WAS prevented was a freeware version (PGP from Zimmerman).  Even
that might have been okay if some philanthropist had been willing to
shell out the money for the royalty.  If this was so important to
society why didn't foundations (like the MacArthur fund) fight to
contribute in this way?

Certainly royalties are contributing to innovation in that RSA Data
Security has gone on to do additional (patentable) work in
authentication, digital signatures and public key servers.  Is this
not evidence that patents can be an encouragement to ongoing research
by small firms (and I assure you RSA Data Security is small!)

And innovations in private key encryption are not covered by the patent.

PKP now must realize how much
>they have lost through their policies, since, in spite of thier much
>vaunted list of licensees, we still see absolutely no widespread use
>of licensed software based on these protocols, and thus no license
>fees accruing to the patent holders!

We don't see widespread use of DES in mail protocols either and it is
a standard and not burdened by RSA's patents. I think there might be
other issues here.

Also, don't confuse licenses with royalties. Sometimes people pay for
licenses in advance. Sometimes even for a fixed feee that may be
indepedent of sales.  I don't believe that PKP's nor RSA DS's
agreements are all public, so I don't think we can know for sure what
the terms were.  But I do know that RSA Data Security is a growing
company. It might be reasonable to speculate that they have some
source of income.

>So now we argue about tradition?  Clearly, the policies of RSADSI and
>PKP have hurt the general software economy by preventing people from
>producing products based on their IDEAS, so in this case patents do
>not help the economy work!

At the same time RSA DS has been a growing company increasing its
employment of software engineers and other professionals in Silicon
Valley. So it is contributing to society too. Their patents have
generated revenue that paid for these people to develop additional
innovations, products and services, some of them patentable as well in
authentication, digital signatures and public key servers. This work
might have been delayed without such financial support. Joe's right
that some people have been kept from competing. That cost has to be
weighed against some of these other benefits.  How many jobs would PGP
have created? We don't know.

>I think we need to change the argument from using the word benefits to
>using the Constitutional terms: promotion of science and the useful
>arts.  Then the question becomes, how has the RSA patent promoted
>science and the useful arts?  The answer: the policies of PKP have
>STIFLED progress on implementation of public key cryptography in the
>USA.

This is not at all obvious.  RSA has implemented public key crypto
systems and brought them to market. Why aren't they adopted? Other
companies and individuals could have developed other innovations in
Private Key cryptography, or enhanced PK cryptography using
improvement patents.  Where are all those other people. Maybe there is
another innovation as startlingly unexpected  as PUBLIC key
cryptography was in the 70's. Maybe there could be KEYLESS
Cryptography. Where is this innovation?

>  So this patent does not fall under the purpose stated in the
>Constitution.

If it contributes to "progress" (not promotion) in the sciences and
useful arts it is.  Now you can buy a PKE system, before RSA you could
not. Now there have been additional innovations in digital signatures,
authentication and public key servers--financed by the proceeds of
RSA's patent.  You'll be able to get these built into your PC or Mac
soon without paying extra.  You didn't have that before.  You don't
have that now with unpatented private key systems. I think that there
is enough here for a policy maker in congress to take a moments pause
before concluding as Joe does that there is no contribution to
progress from this patent.  And certainly I think they would pause
about concluding that there is not progress resulting from the sum of
all software patents (including Dolby's et. al)

>I believe that patents have both good and bad sides.  Clearly,
>traditional patents provide incentive to invent and market products
>for personal benefit as well as making the ideas available for
>everyone to (eventually) use.  I, however, can make a clear
>distinction between the traditional patents and the mental process
>patents that have served to slowly stifle growth in software over the
>past few years.

This distinction seems illusionary to me. I can see a distinction
between processes and devices.  I can see a distinction between
processes that don't change physical states in the world (non-useful
or mental, take your pick) and those that do. But the notion of a
process patent involving a computer that accomplishes a  real world
result (LZW concerving disk storage) being different from a computer
program accomplishing some other real world result (cured rubber)
using unpatentable off the shelf devices looks difficult and
unpredictable to me when judges and juries try to decide what is
appropriate. Also, because what can be done in software can be done in
hardwired circuitry, relays or mechnanical means, (and more
importantly vice versa) I think a distiction between software and
hardware that performs a process causes havoc with the doctrine of
equivalents. Dolby S NR software would not be patentable, even though
the equivalent Dolby NR analog hardware that uses precisely the same
transformation would be patentable.  But a patent holder of the latter
has no means to defend against an infringer using the former method.
And that would cause a lot of confusion

>So Scott, please tell me exactly how software patents serve to promote
>science and the useful arts.  Clearly, software engineering qualifies
>as a useful art, but your promotion of RSA shows how this patent has
>stifled progress.

It has been proposed by some here that competition benefits society.
But if competition is too intense no one can make a profit. Is the
airline industry healthier when most  of the major players lose money?
Is the US auto industry healthier when all the players lost money? I
think not.  An industry is healthy (and beneficial to society) when
the companies in it are profitable and growing and generating new
jobs, and spending more money on new innovations and on increased
distribution. We've seen in the microcosm how patents have helped
companies like RSA Data Security, and Dolby labs to be more
profitable, add jobs to economy and developing and distributing more
products.  And there are thousands of other small companies in the
same postion. Sum these up and you can get a measure of the economic
benefits. Now that's not the end of the story. You need to sum up the
affects on the competitors that were kept out of the market too and
all the other drawbacks that have been discussed here. And you have to
look at opportunity costs and benefits (what would have happened and
when if there had been no patents).  It is tough to get a good
magnitude for any of these, and tougher to compare and weigh. So
that's why I think there's lots of room for disgareement on which side
of dead even things go. But this is precisely the same set of benefits
and costs that apply to non-software patents, and there is little
evidence that I can see that the balance (whatever it may be) would be
different for software than  in other cases.  And the difficulty in
defining a software patent in such a way that it won't have
unintentional affects on other process patents worries me if they were
different balances.

>  You also continue to claim that Dolby holds some
>sort of mythical software patent.

Please don't take my word for it. I'm not a patent depository library.
I'm not the LPF on-line archive. Write directly to the source: Dolby
Laboratories, 100 Potrero Avenue, San Francisco, California
94103-4813. Ask for their company brochure which cites: "Continuing
research at Dolby Laboratories has resulted in over 650 patents and
patent applications worldwide. Note the reference "to the Dolby AC-2
digital audio coding system for such purposes as satellite audio and
telconfencing links, and Dolby S-type noise reduction for
higher-performance cassette recording". Note also the reference to
Dolby S NR, Dolby HX Pro and Dolby Surround in the licensed SOFTWARE
list. Ask for the list of patents and patent applications.

>the Dolby patents mostly promote consumer electronics
>and we haven't (yet) seen Dolby actively suppressing software
>developers.

This is precisely why I am worried about the fuzziness of the
"software" patent categorization. MOST software does perform some
application specific value. Consumer electronics is just one of them.
Most users of Stacker or AutoDoubler (compression algorithms for
disks similar, if not identical to what is claimed in the LZW patents)
are not software developers but "consumers" who use PCs and macs and
don't even have a compiler. I'm worried when someone hints that
somehow a lay jury or judge is going to look at such a situation and
conclude Dolby S NR patent okay, and Stacker Disk compression not
okay.  The only thing that I get out of this is that as a programmer
who does not intend to avail himself of the patent system Kelsey is
personally concerned by the limitations that such patents might lead
to, but concerned about people in other fields.  Similarly, a company
like RSA DS or Dolby might be  pro patent for similar selfish
interests. I don't think it is right to damn only the latter. Both
have chosen their strengths and act in self interest alike.



--

Scott L. McGregor        [e-mail]
President            tel: [phone]
Prescient Software, Inc.    fax: [phone]
3494 Yuba Avenue
San Jose, CA 95117-2967