Tech Insider					     Technology and Trends


			      USENET Archives

Path: gmdzi!unido!fauern!ira.uka.de!yale.edu!qt.cs.utexas.edu!
zaphod.mps.ohio-state.edu!samsung!munnari.oz.au!metro!cluster!benn...@math.ksu.edu
From: benn...@math.ksu.edu (Andy Bennett)
Newsgroups: comp.patents
Subject: Ideas in the air and AT&T
Message-ID: <3043@cluster.cs.su.oz.au>
Date: 3 Oct 91 04:18:49 GMT
Sender: n...@cluster.cs.su.oz.au
Organization: Kansas State University
Lines: 20
Approved: pate...@cs.su.oz.au

I am neither a lawyer nor an expert at bitmapping or much of anything else.
But I think people need to be more careful than they are being in talking
about the obviousness of AT&T's backing store patent. Just because an idea
is in the air doesn't mean it can't be patented by the person who gets to
the patent office first. Two people applied for the patent on the first 
telephone on the SAME DAY; Alexander Graham Bell and the second person whose
name I've forgotten (along with a great many other people). That was an idea
that was clearly in the air and even developed to a practical design by more
than one inventor, but Bell got the patent and it stood under challenge. And 
today everyone has heard of Bell and the other person is forgotten. 

Of course, AT&T is very familiar with this case.

Andy Bennett

--
Andrew G. Bennett         benn...@hilbert.math.ksu.edu     Happiness is being
Dept. of Mathematics      Voice: (913) 532-6750            at the top of the
Kansas State University   Fax:   (913) 532-7004            food chain.
Manhattan, KS 66502       STRICTLY MY OWN OPINIONS                     -DN

Path: gmdzi!unido!mcsun!uunet!zaphod.mps.ohio-state.edu!samsung!munnari.oz.au!
metro!cluster!olivea!bu.edu!nntp-read!jbw
From: jbw%bigbird.bu....@metro.ucc.su.OZ.AU (Joe Wells)
Newsgroups: comp.patents
Subject: Re: Ideas in the air and AT&T
Message-ID: <3137@cluster.cs.su.oz.au>
Date: 13 Oct 91 00:52:09 GMT
References: <3043@cluster.cs.su.oz.au>
Sender: n...@cluster.cs.su.oz.au
Organization: Boston University Computer Science Department
Lines: 15
Approved: pate...@cs.su.oz.au
To: comp-pate...@uunet.UU.NET
In-Reply-To: bennett@math.ksu.edu's message of 3 Oct 91 04:18:49 GMT


In article <3...@cluster.cs.su.oz.au> benn...@math.ksu.edu (Andy Bennett) writes:

   [stuff about Bell applying for telephone patent on same day as
   competitor Gray deleted.]

The point is not whether the law allows for giving patents for ideas that
are "in the air", but whether it is good for society.  Clearly such
patents have been given.  In most cases, the existence of the patent
causes more harm than good, because the specific idea would have been
rediscovered anyway.

-- 
Enjoy,

Joe Wells <j...@cs.bu.edu>

[pjt - patents are indeed purely a creation of law and should have a purpose
	to serve social need...
	Oh, another can of worms... :-) ]
====================================================================
Peter Treloar - comp.patents Moderator
pate...@cs.su.oz.au

Path: gmdzi!unido!mcsun!uunet!spool.mu.edu!munnari.oz.au!metro!cluster!mitel!
Software!...@uunet.uu.net
From: mitel!Software!...@uunet.uu.net (Kim Letkeman)
Newsgroups: comp.patents
Subject: Examples of software patent harm sought.
Message-ID: <3176@cluster.cs.su.oz.au>
Date: 18 Oct 91 08:00:26 GMT
References: <3043@cluster.cs.su.oz.au> <3137@cluster.cs.su.oz.au>
Sender: n...@cluster.cs.su.oz.au
Organization: MITEL Public Switching, Kanata, Ontario, Canada
Lines: 43
Approved: pate...@cs.su.oz.au
In-Reply-To: jbw%bigbird.bu.edu@metro.ucc.su.OZ.AU's message of 
13 Oct 91 00:52:09 GMT


In article <3...@cluster.cs.su.oz.au> jbw%bigbird.bu....@metro.ucc.su.OZ.AU 
(Joe Wells) writes:

| In article <3...@cluster.cs.su.oz.au> benn...@math.ksu.edu (Andy Bennett) writes:
| 
|    [stuff about Bell applying for telephone patent on same day as
|    competitor Gray deleted.]
| 
| The point is not whether the law allows for giving patents for ideas that
| are "in the air", but whether it is good for society.  Clearly such
| patents have been given.  In most cases, the existence of the patent
| causes more harm than good, because the specific idea would have been
| rediscovered anyway.

This is at least the third person that has made the sweeping "more
harm than good" statement in relation to software patents. Could
someone please enumerate all the bad things that have happened as a
result of software patents?

I'm looking for hard facts like:

        - the JPEG guys can't give away the "arithmetic coding"
	  algorithm, even though it would make them incompatible with
	  all the commercial implementations anyway; or

        - the JPEG guys are paralyzed with fear even though they
	  haven't actually approached IBM to see if a deal can be
	  worked out.

not:

	- it's a stupid race and I won't play; or

        - no one can get anything done because there are no areas that
	  aren't patented.

As far as I know, companies are not regularly put out of business as a
result of software patents. It does limit their flexibility and
freedom, but that's the whole point. If everybody could duplicate
every good idea they see and make money from it, the incentive to
disclose would be nil. At least with the patent system reasonable
companies can share the technology.
--
Kim Letkeman	k...@Software.Mitel.COM

====================================================================
Peter Treloar - comp.patents Moderator
pate...@cs.su.oz.au

Path: gmdzi!unido!mcsun!uunet!spool.mu.edu!munnari.oz.au!metro!cluster!
egg...@twinsun.com
From: egg...@twinsun.com (Paul Eggert)
Newsgroups: comp.patents
Subject: Examples of software patent benefit sought
Message-ID: <3207@cluster.cs.su.oz.au>
Date: 22 Oct 91 07:47:33 GMT
References: <3176@cluster.cs.su.oz.au>
Sender: n...@cluster.cs.su.oz.au
Organization: Twin Sun, Inc
Lines: 41
Approved: pate...@cs.su.oz.au

mitel!Software!...@uunet.uu.net (Kim Letkeman) writes:

>Could someone please enumerate all the bad things that have happened as a
>result of software patents?

You've got the question backwards.  The tradition in the US was that
you could not patent software.  Recent lower court decisions have cast
doubt on that tradition, although the issue is still unsettled.  But if
you want to change the tradition, you must argue why software patents
are a good thing.  Arguments that it can't hurt are not enough to
change the status quo.  Our most successful software companies grew
without benefit of the supposed protections of software patents, but
are are now being threatened by legions of patent lawyer leeches.
Do you really think the recent software patent craze will help the
industry keep growing?

Could someone please enumerate _any_ good thing that has happened as a
result of software patents?  Anything that has made the expense of
millions of dollars on software patents worthwhile?


>As far as I know, companies are not regularly put out of business as a
>result of software patents.

That's as silly as saying ``As far as I know, companies are not
regularly put out of business as a result of health insurance
premiums.'' When companies are put out of business, it is typically due
to some management dispute or cash flow emergency that makes the
newspapers, even though the underlying cause may be something mundane
like health insurance premiums or paying patent lawyers.


>At least with the patent system reasonable companies can share the technology.

Nonsense.  Companies share software technology all the time without
using the patent system.  If we were _forced_ to use patents to share
software technology, most progress in the field would stop.

====================================================================
Peter Treloar - comp.patents Moderator
pate...@cs.su.oz.au

Path: gmdzi!zeus.ieee.org!europa.asd.contel.com!uunet!spool.mu.edu!
munnari.oz.au!metro!cluster!mitel!Software!...@uunet.uu.net
From: mitel!Software!...@uunet.uu.net (Kim Letkeman)
Newsgroups: comp.patents
Subject: Re: Examples of software patent benefit sought
Message-ID: <3234@cluster.cs.su.oz.au>
Date: 25 Oct 91 03:23:13 GMT
References: <3176@cluster.cs.su.oz.au> <3207@cluster.cs.su.oz.au>
Sender: n...@cluster.cs.su.oz.au
Organization: MITEL Public Switching, Kanata, Ontario, Canada
Lines: 71
Approved: pate...@cs.su.oz.au
In-Reply-To: eggert@twinsun.com's message of 22 Oct 91 07:47:33 GMT


In article <3...@cluster.cs.su.oz.au> egg...@twinsun.com (Paul Eggert) writes:

| mitel!Software!...@uunet.uu.net (Kim Letkeman) writes:
| 
| >Could someone please enumerate all the bad things that have happened as a
| >result of software patents?
| 
| You've got the question backwards.  The tradition in the US was that
| you could not patent software.  Recent lower court decisions have cast
| doubt on that tradition, although the issue is still unsettled.  But if
| you want to change the tradition, you must argue why software patents
| are a good thing.  Arguments that it can't hurt are not enough to
| change the status quo.  Our most successful software companies grew
| without benefit of the supposed protections of software patents, but
| are are now being threatened by legions of patent lawyer leeches.
| Do you really think the recent software patent craze will help the
| industry keep growing?

I disagree with most of this. The point I was making was that there
are a lot of people willing to take very strong anti-patent stands
because of several questionable software patents (e.g. XOR, backing
store, etc.) I was simply asking to have something more than anecdotal
discussed because I really would like to know of specific cases of
harm caused by a software patent.

| Could someone please enumerate _any_ good thing that has happened as a
| result of software patents?  Anything that has made the expense of
| millions of dollars on software patents worthwhile?

That might be quite interesting as well. However, the lack of any good
things doesn't really go very far in documenting the bad things.

| >As far as I know, companies are not regularly put out of business as a
| >result of software patents.
| 
| That's as silly as saying ``As far as I know, companies are not
| regularly put out of business as a result of health insurance
| premiums.'' When companies are put out of business, it is typically due
| to some management dispute or cash flow emergency that makes the
| newspapers, even though the underlying cause may be something mundane
| like health insurance premiums or paying patent lawyers.

Again, I was looking for a specific case of a company forced out of
business because their technology (I'm sure that lots of small start
ups are based on a single, very new technology) was patented by
someone else. This would be a very direct attribution, and not subject
to something as vague as the price of patent lawyers being too high.

| >At least with the patent system reasonable companies can share the technology.
| 
| Nonsense.  Companies share software technology all the time without
| using the patent system.  If we were _forced_ to use patents to share
| software technology, most progress in the field would stop.

Again, I don't agree. I don't think most companies are out there
sharing technology. I think they keep their's secret as long as
possible (certainly until it hits the marketplace.)

The sharing comes from academia (where profit is essentially a
non-issue) and/or cross-fertilization from employee turnover. Also,
when a dazzling product hits the marketplace, lots of people spend
lots of money reengineering it.

Note that I am not defending the patenting of algorithms (which is
where this all started as I remember), but rather am asking for some
enumeration of actual negative incidents, rather than such vague
statements as "companies share technology all the time without using
the patent system."

Off course, opinion bashing is fun too.
--
Kim Letkeman	k...@Software.Mitel.COM

Path: gmdzi!zeus.ieee.org!europa.asd.contel.com!uunet!zephyr.ens.tek.com!
uw-beaver!cornell!batcomputer!munnari.oz.au!metro!cluster!ji...@microsoft.com
From: ji...@microsoft.com (Jim ADCOCK)
Newsgroups: comp.patents
Subject: Re: Examples of software patent benefit sought
Message-ID: <1991Oct24.185506.1864@microsoft.com>
Date: 29 Oct 91 05:35:21 GMT
References: <3176@cluster.cs.su.oz.au> <3207@cluster.cs.su.oz.au>
Sender: n...@cluster.cs.su.oz.au
Organization: Microsoft Corp.
Lines: 90
Approved: pate...@cs.su.oz.au

In article <3...@cluster.cs.su.oz.au> egg...@twinsun.com (Paul Eggert) writes:
|mitel!Software!...@uunet.uu.net (Kim Letkeman) writes:
|
|>Could someone please enumerate all the bad things that have happened as a
|>result of software patents?
|
|You've got the question backwards.  The tradition in the US was that
|you could not patent software.  Recent lower court decisions have cast
|doubt on that tradition, although the issue is still unsettled.  But if
|you want to change the tradition, you must argue why software patents
|are a good thing.  Arguments that it can't hurt are not enough to
|change the status quo.  

The status quo is where we are today.  The status quo yesterday is not
the status quo today is not the status quo tomorrow.

----

Let me relate two experiences I had regards software patents -- 
experiences that happened before I joined Microsoft.

1) As a high-school student I became interested in computers and music, 
and started developing a new idea combining the two.  In college I
seriously persued the idea using microprocessors.  This was in the
1975-1978 time frame -- ie at the very start of microcomputers.
I bought one of the first batch of 40 6502 microprocessors that
were sold commercially, for example.  I pursued the idea for many years,
spending tens of thousands of dollars.  And then gave up.  I could
not figure out any way of protecting my idea that would not be instantly
ripped-off by the Japanese, who are very active in this area.
-- that I could afford.  I could have afforded a software patent if
that were available to me.  I couldn't afford to spend the $100,000
it would have cost me to re-cast my software as a custom integrated
circuit -- and then get a hardware patent on it.  Why if I embodied that
invention in hardware should it be patentable, but not if I embodied it
in software in a microcontroller?  The result is the same by God --
a 40-pin plastic chip -- even the pin-outs are the same!  Yet some
people want to look at one IC and say: "Software -- you can't patent
it" and look at the other IC and say: "Hardware -- you can patent it."

I say this is baloney.  

My software invention can hardly be called "trivial" 
	-- its been 15 years and no one else has thought of it!


2) So, I gave up on the invention I could not profit from, and went
to work at a major company.  I did pretty boring work for a number of years, but
then along with a friend we started persuing a new idea for how to
make fundamental electronic measurements on a wide variety of electronic
devices.  We convinced management to allow us to persue this idea for 
several years, at a probably total cost of more than a million dollars.
Why did management pay for this research? -- because they thought they
could get their money back.  And this was because that company's lawyers
thought [correctly] that they could successfully argue a patent application
even if it contained an element of software programming.  And they
were right.  If we couldn't have patented it our Japanese would have
reverse-engineered it, just as they have ripped off everything else we came
up with.  Coming up with a fundamental new measurement technique is hard
and expensive.  Reverse engineering it is trivial.  Copyrights do nothing
to protect these kinds of inventions.

This was not a trivial piece of software.  The engineering reports on
its progress ran thousands of pages.  Fortunately, because we could
get a patent on it, we were able to describe how it works to the
industry.

So, in my personal experience, if I am to have the freedom to think up
new ideas in an industrial setting, management has to be able to believe
that they can protect those new ideas.  They can't afford to spend hundreds
or millions of dollars of their investor's money to sponsor research into
new techniques that their competitors are free to rip off instantly.
And, thus without protection, I am forced to do stupid grunt work
that isn't *worth* ripping off.  The few people who have the freedom
to work as professors at universities at the general publics expense
have much to be thankful for -- but they don't represent the norm.
The public should be thankful when such professors freely share
their work with society -- and many of them don't -- and such 
professors should be thankful that society supports their research.
But, we should not confuse the situation of the university
professor with the situation of engineers in industry.  What
is good for one is not necessarily good for the other.

The issue is not software patents verses no software patents.  The
issue is the mal-performance of the patent office in allowing trivial
ideas.  These evil actions by the patent office are the result of
"reforms" introduced under the Reagan administration.  Allowing trivial
patents has nothing uniquely to do with "software."  There is nothing
unique about software.

Fix the patent office administration.  Don't distort the patent system
by introducing special cases for software verses hardware embodiments.

Path: gmdzi!zeus.ieee.org!europa.asd.contel.com!darwin.sura.net!gatech!
bloom-beacon!micro-heart-of-gold.mit.edu!uw-beaver!cornell!batcomputer!
munnari.oz.au!metro!cluster!egg...@twinsun.com
From: egg...@twinsun.com (Paul Eggert)
Newsgroups: comp.patents
Subject: Re: Examples of software patent benefit sought
Message-ID: <3300@cluster.cs.su.oz.au>
Date: 31 Oct 91 07:45:50 GMT
References: <3176@cluster.cs.su.oz.au> <3207@cluster.cs.su.oz.au> 
<3234@cluster.cs.su.oz.au>
Sender: n...@cluster.cs.su.oz.au
Lines: 41
Approved: pate...@cs.su.oz.au

mitel!Software!...@uunet.uu.net (Kim Letkeman) writes:

>... there are a lot of people willing to take very strong anti-patent
>stands because of several questionable software patents (e.g. XOR,
>backing store, etc.) I was simply asking to have something more....
>I really would like to know of specific cases of harm caused by a
>software patent.

The very patents you mentioned both cause harm.  Cadtrak, which holds the XOR
cursor patent, regularly shakes down graphics vendors who use this obvious
technique.  AT&T has written letters to the MIT X Consortium members, saying
that X infringes on the backing store patent, and right now lawyers from both
sides are no doubt running up huge bills.  These activities harm progress in
software development, and benefit only patent lawyers and their ilk.

>I was looking for a specific case of a company forced out of business.

To paraphrase you: ``Software patents aren't driving companies bankrupt, so
they must not be so bad after all.''  You misunderstand how patents are used
in the real world.  Patent holders don't want to drive licensees out of
business, because bankrupt licensees are a poor source of royalties.  You'll
rarely find patents causing bankruptcies in any field, much less in software
where their legal status is still so unsettled.

Most of software patents' harm comes from taxing existing companies or from
preventing companies from forming in the first place.  For example, Public Key
Partners has aggressively prevented public key encryption technology from
being widely used in the US.  They have even sent threatening letters to a
college student who wanted to distribute such software freely.  Without these
patents, public keys would be much more widely used.  This is a specific
example where software patents have caused direct harm on our use of software
technology.  Even though nobody has gone bankrupt, we are all the poorer for
it.

>I don't think most companies are out there sharing technology.

Um, have you read the papers lately?  Apple and IBM are sharing Pink.  IBM and
Microsoft share (if that's the word :-) OS/2.  Apple and Microsoft share True
Type.  DEC, HP, IBM, etc. share Motif.  A cast of thousands shares X.  Even
Mitel, your company, shares software technology by using the GNU C compiler.
We don't need patents to share software technology profitably.

Path: gmdzi!zeus.ieee.org!dorm.rutgers.edu!rutgers!ub!zaphod.mps.ohio-state.edu!
sdd.hp.com!spool.mu.edu!munnari.oz.au!metro!cluster!brns...@KRAMDEN.ACF.NYU.EDU
From: brns...@KRAMDEN.ACF.NYU.EDU (Dan Bernstein)
Newsgroups: comp.patents
Subject: Re: Examples of software patent benefit sought
Message-ID: <3307@cluster.cs.su.oz.au>
Date: 1 Nov 91 01:27:12 GMT
References: <3176@cluster.cs.su.oz.au> <3207@cluster.cs.su.oz.au> 
<3234@cluster.cs.su.oz.au>
Sender: t...@cluster.cs.su.oz.au
Organization: IR
Lines: 18
Approved: pate...@cs.su.oz.au

Kim Letkeman writes:
> I was simply asking to have something more than anecdotal
> discussed because I really would like to know of specific cases of
> harm caused by a software patent.

If it were not for software patents, we'd all be using public-key
encryption, and mail and news forgery wouldn't exist.

If it were not for software patents, people would be able to use any
compressor available, without fear of a lawsuit.

In plainer words, patents on encryption and compression have screwed the
communications infrastructure of the world, and will continue to do so
until the next millenium. That's ``harm'' in my book.

Where's the benefit to justify this cost?

---Dan

Path: gmdzi!zeus.ieee.org!europa.asd.contel.com!darwin.sura.net!uvaarpa!
caen!spool.mu.edu!munnari.oz.au!metro!cluster!patents
From: pate...@cs.su.oz (patents newsgroup moderator)
Newsgroups: comp.patents
Subject: Re: Examples of software patent benefit sought
Message-ID: <3326@cluster.cs.su.oz.au>
Date: 4 Nov 91 06:03:24 GMT
References: <3176@cluster.cs.su.oz.au> <3207@cluster.cs.su.oz.au>
Sender: n...@cluster.cs.su.oz.au
Organization: Hacker's haven
Lines: 79
Approved: pate...@cs.su.oz.au

In article <3...@cluster.cs.su.oz.au> egg...@twinsun.com (Paul Eggert) writes:
>mitel!Software!...@uunet.uu.net (Kim Letkeman) writes:
>>Could someone please enumerate all the bad things that have happened as a
>>result of software patents?
>You've got the question backwards.  The tradition in the US was that
>you could not patent software.  Recent lower court decisions have cast
>doubt on that tradition, although the issue is still unsettled.  

First error of fact. The U.S. Supreme Court ruled that software patents
are valid in Dyer, 1981. That essentially settles the issue for the U.S.
at least.

>Our most successful software companies grew
>without benefit of the supposed protections of software patents, but
>are are now being threatened by legions of patent lawyer leeches.

I don'y know just who you mean by "most successful", but the Japanese
computer and software companies have been very active with patents. To
name a few U.S. companies, IBM, DEC and AT&T also seem to create quite a
lot of software. Motorola has the disctinction of writing the most
executed piece of code in existance, the idle loop for the Ford
electronic ignition!

>Do you really think the recent software patent craze will help the
>industry keep growing?

Yes I do, in fact I see software patents as ESSENTIAL to move software
out of the "attic hacker" phase and into something where you can get
investors to put up $500M or more.

Patents are essentially a form of private property. It is convenient to
trespass on and use someone else's beach, but if you can't do that you
must go find one of your own.

>Could someone please enumerate _any_ good thing that has happened as a
>result of software patents?  Anything that has made the expense of
>millions of dollars on software patents worthwhile?

Yes indeed. The investment in quality software and the interst in making
workable software has increased enourmously. All of a sudden you find
people willing to part with source code - once it is protected by a
patent so you can't steal it!

>>As far as I know, companies are not regularly put out of business as a
>>result of software patents.
>That's as silly as saying ``As far as I know, companies are not
>regularly put out of business as a result of health insurance
>premiums.'' When companies are put out of business, it is typically due
>to some management dispute or cash flow emergency that makes the
>newspapers, even though the underlying cause may be something mundane
>like health insurance premiums or paying patent lawyers.

It may be silly, but its true. Of course you could argue "overpaid
software writers" as much as patent lawyers.

>>At least with the patent system reasonable companies can share the technology.
>Nonsense.  Companies share software technology all the time without
>using the patent system.  If we were _forced_ to use patents to share
>software technology, most progress in the field would stop.

Until now no company has been willing to release source code unless it
was part of (a) a joint development or (b) a product purchase. Simply
because the source code WAS the product.

Circuit design is very similar to software in that it is mostly
intellectual, no "touch and feel" product, much of the field is
expressed or expressible using mathematics of some sort, and the results
are incorporated into products.

Somehow I don't see all these terrible things happening in circuit
design. Why will software be different? As one who has worked for many
years in both areas, I don't see that there is a difference.

Sure it hurts to have to pay to use property that you used for nothing.
Unless of course it is YOUR front lawn they are using!

           *Mike Waters    AA4MW/7  wat...@nddsun1.sps.mot.com *
Q: Why did the tachyon cross the road?
A: Because it was on the other side.

Path: gmdzi!zeus.ieee.org!europa.asd.contel.com!darwin.sura.net!uvaarpa!
caen!spool.mu.edu!munnari.oz.au!metro!cluster!mitel!Software!...@uunet.uu.net
From: mitel!Software!...@uunet.uu.net (Kim Letkeman)
Newsgroups: comp.patents
Subject: Re: Examples of software patent benefit sought
Message-ID: <3341@cluster.cs.su.oz.au>
Date: 4 Nov 91 09:52:52 GMT
References: <3176@cluster.cs.su.oz.au> <3207@cluster.cs.su.oz.au>
	<3234@cluster.cs.su.oz.au> <3300@cluster.cs.su.oz.au>
Sender: n...@cluster.cs.su.oz.au
Organization: MITEL Public Switching, Kanata, Ontario, Canada
Lines: 60
Approved: pate...@cs.su.oz.au
In-Reply-To: eggert@twinsun.com's message of 31 Oct 91 07:45:50 GMT


In article <3...@cluster.cs.su.oz.au> egg...@twinsun.com (Paul Eggert) writes:

| To paraphrase you: ``Software patents aren't driving companies
| bankrupt, so they must not be so bad after all.''  You misunderstand
| how patents are used in the real world.  Patent holders don't want
| to drive licensees out of business, because bankrupt licensees are a
| poor source of royalties.  You'll rarely find patents causing
| bankruptcies in any field, much less in software where their legal
| status is still so unsettled.

Actually, I do understand how they are used in the real world. They
are often used as currency in transactions between major corporations.
They are used for cross-licensing purposes.  And, as you say, they are
used to shake down those who have not licensed a particular technology
>from the inventor.

Unfortunately, the negative impressions most people get are still
caused by a few very obvious "bad" patents. At least, that's how the
majority of anti-patent postings on this group read.

| Most of software patents' harm comes from taxing existing companies
| or from preventing companies from forming in the first place.  For
| example, Public Key Partners has aggressively prevented public key
| encryption technology from being widely used in the US.  They have
| even sent threatening letters to a college student who wanted to
| distribute such software freely.  Without these patents, public keys
| would be much more widely used.  This is a specific example where
| software patents have caused direct harm on our use of software
| technology.  Even though nobody has gone bankrupt, we are all the
| poorer for it.

Granted. But there must be equivalent examples in non-software areas,
because this does not seem to be an issue of software only. My point
in saying this is that the above paragraph is a consequence of the
patent system in general, not of software patents in particular.

| >I don't think most companies are out there sharing technology.
| 
| Um, have you read the papers lately?  Apple and IBM are sharing
| Pink.  IBM and Microsoft share (if that's the word :-) OS/2.  Apple
| and Microsoft share True Type.  DEC, HP, IBM, etc. share Motif.  A
| cast of thousands shares X.  Even Mitel, your company, shares
| software technology by using the GNU C compiler.  We don't need
| patents to share software technology profitably.

I only wish Mitel were my company :). My employer does, in fact, make
use of FSF software. FSF is good enough that, were it necessary, we
would probably pay money for it. However, we are not sharing
technology with the FSF in any sense.

Your other examples are either: companies collaborating for potential
profit; or academic institutions giving away their technology in hopes
of establishing a standard (and succeeding of course.)

In neither case is it obvious to me that there are companies (i.e.
institutions that exist to make money) sharing technology without
entering into agreements intended to take a slice of the profits.
"Sharing", as used by the first poster to use that term in this
thread, did not mean (IMO) collaberating.
--
Kim Letkeman	k...@Software.Mitel.COM

Path: gmdzi!unido!mcsun!uunet!munnari.oz.au!yoyo.aarnet.edu.au!aegir!
dmsadel.adl.dms.CSIRO.AU!dmssyd.syd.dms.CSIRO.AU!metro!cluster!egg...@twinsun.com
From: egg...@twinsun.com (Paul Eggert)
Newsgroups: comp.patents
Subject: Re: Examples of software patent benefit sought
Message-ID: <3381@cluster.cs.su.oz.au>
Date: 6 Nov 91 07:10:38 GMT
References: <3176@cluster.cs.su.oz.au> <3207@cluster.cs.su.oz.au> 
<3326@cluster.cs.su.oz.au>
Sender: n...@cluster.cs.su.oz.au
Lines: 41
Approved: pate...@cs.su.oz.au

I wrote: ``The tradition in the US was that you could not patent software.
Recent lower court decisions have cast doubt on that tradition,
although the issue is still unsettled.''  Mike Waters responded:

>First error of fact. The U.S. Supreme Court ruled that software patents
>are valid in Dyer, 1981. That essentially settles the issue for the U.S.
>at least.

Waters is greatly mistaken.  In Gottschalk v Benson (1972), the US Supreme
Court ruled that computer program algorithms and mental processes are not
patentable.  Waters's reference to ``Dyer, 1981'' is no doubt a misspelling for
Diamond v Diehr (1981), where the Court ruled that a patent claim should not be
rejected merely because it includes a computer program as an element; the only
requirement is that the basic process being claimed must be patentable.  (In
Diehr the basic process was curing rubber.)  However, the basic principle laid
down by the Court is that a program isn't patentable in and of itself.

Naturally, the Patent Office and patent lawyers dislike this principle and are
trying their best to reverse it, and in lower courts they have won some
decisions that many people find questionable.  Because the Supreme Court
_hates_ patent cases, it may be some time before it addresses these issues.
In the mean time, it's fair to say that the issue is still unsettled.

For more on this subject please see Pamela Samuelson, ``Should program
algorithms be patented?'', Communications of the ACM 33, 8 (August 1990),
23-27.


>Patents are essentially a form of private property. It is convenient to
>trespass on and use someone else's beach, but if you can't do that you
>must go find one of your own.

Nonsense.  When you use my beach, you inhibit my use of my beach; but when you
use my idea, you don't inhibit me from using my idea.  Patents may be
necessary, but it's silly to try to justify them on the same grounds as real
estate.

[pjt- for a very good explaination (as always :-)) of what a patent
	right gives you see Carl's article on 
		"patents upon patents"
	]

Path: gmdzi!zeus.ieee.org!dorm.rutgers.edu!rutgers!ub!zaphod.mps.ohio-state.edu!
think.com!mips!spool.mu.edu!munnari.oz.au!metro!cluster!egg...@twinsun.com
From: egg...@twinsun.com (Paul Eggert)
Newsgroups: comp.patents
Subject: patents are not like real estate
Message-ID: <3406@cluster.cs.su.oz.au>
Date: 11 Nov 91 01:59:41 GMT
Sender: n...@cluster.cs.su.oz.au
Lines: 10
Approved: pate...@cs.su.oz.au

Mike Waters writes:

>Patents are essentially a form of private property. It is convenient to
>trespass on and use someone else's beach, but if you can't do that you
>must go find one of your own.

Nonsense.  When you use my beach, you inhibit my use of my beach; but when you
use my idea, you don't inhibit me from using my idea.  Patents may be
necessary, but it's silly to try to justify them on the same grounds as real
estate.

Path: gmdzi!zeus.ieee.org!dorm.rutgers.edu!rutgers!uwm.edu!wupost!uunet!
munnari.oz.au!metro!cluster!egg...@twinsun.com
From: egg...@twinsun.com (Paul Eggert)
Newsgroups: comp.patents
Subject: Re: Examples of software patent benefit sought
Message-ID: <3407@cluster.cs.su.oz.au>
Date: 11 Nov 91 02:15:05 GMT
References: <3341@cluster.cs.su.oz.au>
Sender: n...@cluster.cs.su.oz.au
Lines: 25
Approved: pate...@cs.su.oz.au

To summarize this topic:

Kim Letkeman asked for examples of where software patents have hurt.
Several people responded with specific examples,
including public key cryptosystems, XOR cursors, and backing store.

I asked where software patents have helped.
No one has responded with a specific example.

The only responses have been:

  Jim Adcock wrote about a combined hardware/software device that has never
  been patented, but (from the sketch presented) would be patentable under the
  Diamond v Diehr principle allowing patents on hardware devices that contain
  some software.

  Mike Waters misquoted the law and gave no examples.

  Waters and Letkeman both replied, ``Why should software be any different from
  hardware?''.  There are several good answers to their counter-question,
  starting with ``Why should software be any different from books?'',
  but they're evading my original question.

Surely software patent proponents can come up with _some_ example.
It's hard to believe that they base their position solely on faith.

Path: gmdzi!unido!mcsun!uunet!munnari.oz.au!metro!cluster!...@Software.Mitel.COM
From: k...@Software.Mitel.COM (Kim Letkeman)
Newsgroups: comp.patents
Subject: Re: Examples of software patent benefit sought
Message-ID: <3416@cluster.cs.su.oz.au>
Date: 11 Nov 91 18:11:13 GMT
References: <3341@cluster.cs.su.oz.au> <3407@cluster.cs.su.oz.au>
Sender: n...@cluster.cs.su.oz.au
Organization: MITEL Public Switching, Kanata, Ontario, Canada
Lines: 139
Approved: pate...@cs.su.oz.au
In-Reply-To: eggert@twinsun.com's message of 11 Nov 91 02:15:05 GMT


In article <3...@cluster.cs.su.oz.au> egg...@twinsun.com (Paul Eggert) writes:

| To summarize this topic:
| 
| Kim Letkeman asked for examples of where software patents have hurt.
| Several people responded with specific examples,
| including public key cryptosystems, XOR cursors, and backing store.

It appears that it's time to clear up some problems with terms of
reference.

1. Hurt

   I mean really hurt. Not inconvenienced. I got lots of answers like
   "if software patents did not exist all mail systems would be
   secure"; "some nasty company is running around fleecing all users
   of the XOR cursor method out of a few bucks"; and "no one can
   implement a window system because backing store is the obvious way
   to go".

   First, I don't believe that we'd all be using secure mail by now.
   Second, I doubt we need it that badly yet since the world is still
   turning and lots of EMAIL is zooming around.

   Third, the company running around taking its share of the XOR money
   is a pain in the butt, but people are paying because it lets them
   play the game for a fairly low entry fee.

   And finally, not everyone on the planet wants to write a windowing
   system. The biggies are already there: X, Windows 3, PM, Openlook,
   etc. Anyone who wants to join this league can either afford the
   license or will cross-license. If you're writing one for in house
   use, then don't worry about the patent. Who the heck is going to
   bother you?

   I guess I'm annoyed that the anti-patent people seem to want
   software to be exempt from the forces of the market. People pay to
   have software developed (BIG BUCKS) so they deserve some protection
   for a novel implementation.

2. Software

   Software is not a book. A computer program is totally useless
   unless you define its purpose and context (e.g.  hardware).
   Programs are not patented. Simple algorithms are not patented. But
   implementations are.

   Books don't make toasters go. There are no books inside your
   microwave oven. A telephone switch isn't controlled by a book.

   My point here is that I think a lot of people have an incredibly
   naive view of software. I'm not talking about that little matrix
   multiplication program you did in first year. I'm talking about the
   multi-million lines of code in a telephone switch, or an OS, or a
   big windowing system. There are lots of opportunities for patents
   in these areas that could never be documented as just "a program".

   I would appreciate it if Mr. Eggert et al would stop assuming that
   the pro patent people are talking about computer programs and/or
   algorithms as written in someone's notebook. We're talking about
   something that happens to contain software or perhaps be completely
   embodied as aoftware but that has been implemented and actually
   does something useful and, of course, is potentially money making.

   Software is not a book. Please say that 10 times. Or better yet,
   write some software to print it in a loop on a bit of hardware just
   to clarify the point.

3. Good

   Software patents are not necessarily good. They are here.  Probably
   to stay. I haven't even said that I thought they were good. But I
   have said two things:

   - that I feel that the general benefits of a patent system accrue
     to software as well; and

   - that I thought most of what people call "bad" about them is
     simply whining about the proper operation of a market that
     happens to include protection for the inventor of technology.

   A few people have had to shell out a few bucks. Some of it may not
   have been necessary. So what? I manage software projects and I see
   what it costs .... to anyone that might have to worry about
   patents, this is a drop in the bucket.

   As has been stated numerous times, the typical one man show has
   nothing to fear because he is too small to bother with.

   To the academics in the crowd that are upset that they can't give
   away someone else's algorithm I can only say "do you really think
   it would be fair?" Patents don't stop research into an area, they
   just eliminate the possibility of profiting from the area of
   research covered by the claims of the patent in question (as Mr.
   Oppedahl has been laboring to point out repeatedly.)

| I asked where software patents have helped.
| No one has responded with a specific example.

There have been several very well written (and well balanced IMO)
articles with respect to the patent system and the service it provides
to the technology community. I personally feel that the benefits cover
software patents as well (see my definition of software above before
getting excited at this.)

The authors were Terry Ritter and Carl Oppedahl should you choose to
reread their articles in light of my above definitions.

| The only responses have been:
| 
|   Jim Adcock wrote about a combined hardware/software device that has never
|   been patented, but (from the sketch presented) would be patentable under the
|   Diamond v Diehr principle allowing patents on hardware devices that contain
|   some software.

I know for a fact that most software patents are presented as a
machine performing a specified process under the control of software.
I think that's all Mr. Adcock was pointing out.

|   Mike Waters misquoted the law and gave no examples.

I assume this was merely a dig.

|   Waters and Letkeman both replied, ``Why should software be any different from
|   hardware?''.  There are several good answers to their counter-question,
|   starting with ``Why should software be any different from books?'',
|   but they're evading my original question.

As I pointed out above, there are a couple of good articles that
explain the good the patent system does. If you don't believe this can
accrue to software as well, then you can never agree.

| Surely software patent proponents can come up with _some_ example.
| It's hard to believe that they base their position solely on faith.

Again, reread Mr. Ritter and Mr. Oppedahl. If you are not convinced
(given the above more restrictive definition of what I consider to be
patentable software) then we shall have to agree to disagree.
--
Kim Letkeman	k...@Software.Mitel.COM

Path: gmdzi!ieee.org!dorm.rutgers.edu!rutgers!cs.utexas.edu!wupost!uunet!
munnari.oz.au!metro!cluster!0001811...@mcimail.com
From: 0001811...@mcimail.com (Carl Oppedahl)
Newsgroups: comp.patents
Subject: 1.2 million
Message-ID: <3424@cluster.cs.su.oz.au>
Date: 12 Nov 91 23:05:00 GMT
Sender: n...@cluster.cs.su.oz.au
Lines: 69
Approved: pate...@cs.su.oz.au


 
egg...@twinsun.com (Paul Eggert) writes:
 
>To summarize this topic:
 
>I asked where software patents have helped.
>No one has responded with a specific example.
 
At risk of sounding like a smart aleck, how about this.  It must be that
software patents have helped, because clients are willing to pay patent 
lawyers to help them apply for them.  
 
In fairness, I suppose the argument is more like, Carl Oppedahl has software 
patent clients, so it must be that his clients believe that software patents
help.  And of course clients' beliefs might not be based in reality.
 
As long as we are (apparently) trying to compare the ratio of helpful to 
hurtful software patents, I offer some information that may illuminate.
Recently I asked for people in this group to provide the patent numbers for 
the software patents they thought were bad.  A number of people sent me
patent numbers, including a couple of earlier postings to the group that I had
missed due to (a) my recent joining of the group and (b) my inability, through
MCIMail, to get access to older postings.  (I cannot FTP, for example.)
 
About sixty distinct U.S. patent numbers were provided to me.  Of those, not
all were said to be "bad".  Perhaps thirty were said to be bad.
 
Consider, for perspective, how many patents have issued during the same 
period of time.  I used the Lexis database Patent Util, which provides the full
text of patents from January 1, 1975 to the present.  The patent number on 
January 1, 1975 was about 3,859,635, and the patent number on November 5, 1991
was about 5,063,607.  So during that approximately 17 years about 1.2 million
patents have issued.  (For these to issue, several million patent applications
had to have been filed.  About half never resulted in an issued patent.)
 
Of those 1.2 million patents it is interesting to wonder how many are 
software patents.  Of course, for those on this group the definition of that
term is not settled.  And if we could settle on a definition, someone would
have to review each of the 1.2 million patents, one by one, to determine 
whether it is one.  Less time-consuming than the human genome project, but 
still lengthy.
 
I quite arbitrarily decided not to count software patents, but instead to 
count patents containing the words "ALGORITHM" or "SOFTWARE" or "SOURCE CODE" 
or "FLOWCHART".  That test surely misses some patents which some would say
are software patents.  And it surely counts some patents which some would 
say are not software patents.  But it's just orders of magnitude I'm trying to 
approach, here.
 
About 40,000 patents were found.  Their time grouping was as follows:
 
        Issued 1975-1979:  2,763
        Issued 1980-1985:  6,202
        Issued 1985-1989: 19,492
        Issued 1990-now : 11,886
 
        Total           : 40,343
 
So if we assume (which is not really justified) that these are software
patents then of the 40,000 software patents, maybe thirty have been 
discovered to be bad.  A thousand-to-one ratio.  The suggestion is that
maybe the Patent Office is not doing so badly at distinguishing things that
should be patentable from things that should not.
 
Also note that these 40,000 patents represent many thousands of distinct 
clients, each of whom apparently thought the patent system would help
them.

Path: gmdzi!ieee.org!dorm.rutgers.edu!rutgers!cs.utexas.edu!qt.cs.utexas.edu!
yale.edu!spool.mu.edu!munnari.oz.au!metro!cluster!...@kithrup.COM
From: s...@kithrup.COM (Sean Eric Fagan)
Newsgroups: comp.patents
Subject: Re: Examples of software patent benefit sought
Message-ID: <1991Nov13.065335.7592@kithrup.COM>
Date: 13 Nov 91 06:53:35 GMT
References: <3341@cluster.cs.su.oz.au> <3407@cluster.cs.su.oz.au> 
<3416@cluster.cs.su.oz.au>
Sender: n...@cluster.cs.su.oz.au
Organization: Kithrup Enterprises, Ltd.
Lines: 59
Approved: pate...@cs.su.oz.au

In article <3...@cluster.cs.su.oz.au> k...@Software.Mitel.COM (Kim Letkeman) writes:
>First, I don't believe that we'd all be using secure mail by now.

I would.  News, at least.

>And finally, not everyone on the planet wants to write a windowing
>system. The biggies are already there: X, Windows 3, PM, Openlook,
>etc. Anyone who wants to join this league can either afford the
>license or will cross-license. If you're writing one for in house
>use, then don't worry about the patent. Who the heck is going to
>bother you?

Yeah.  Let's ignore the fact that X was written, not by a commercial
organization, but by a university, and is available *free* for anyone who
wants it.  Let's ignore the fact that this is exactly the type of thing that
patents are supposed to encourage, and resulted in something that
"benefited" the public (let's not get into a discussion about whether or
not X is good for you 8-)).  Let's also ignore the apparantly little-known
MGR window manager, also freely available, written by a small number of
people.  Let's ignore the fact that if you're infringing on a patent, EVEN
IN HOUSE, you can be forced to stop using the patented thing (algorithm, in
this case).

In other words, let's ignore facts.  Right?

>People pay to
>have software developed (BIG BUCKS) so they deserve some protection
>for a novel implementation.

Small software houses also develop software.  There are quite a few people
who make a living doing things like porting X windows.  How many of these
would be forced out of business if they had to have a patent search done on
every single algorithm in their program?  And, of course, a patent search
doesn't cover pending patents, meaning that they can be bitten years down
the line and forced to pay royalties, or discontinue use.

>Programs are not patented. Simple algorithms are not patented. But
>implementations are.

Uhm, what the hell do you call using XOR for cursor manipulation?  That's
about as simple an algorithm as you can get!

>Books don't make toasters go. There are no books inside your
>microwave oven. A telephone switch isn't controlled by a book.

So?  Songs aren't books either, but they're not patentable.  And books
aren't anything without the paper they're printed on, so I guess books
should be patentable, as well.

And the crowning piece of idiocy:

>   As has been stated numerous times, the typical one man show has
>   nothing to fear because he is too small to bother with.

-- 
Sean Eric Fagan  | "I made the universe, but please don't blame me for it;
s...@kithrup.COM  |  I had a bellyache at the time."
-----------------+           -- The Turtle (Stephen King, _It_)
Any opinions expressed are my own, and generally unpopular with others.

Path: gmdzi!ieee.org!dorm.rutgers.edu!rutgers!cs.utexas.edu!qt.cs.utexas.edu!
yale.edu!spool.mu.edu!munnari.oz.au!metro!cluster!brns...@KRAMDEN.ACF.NYU.EDU
From: brns...@KRAMDEN.ACF.NYU.EDU (Dan Bernstein)
Newsgroups: comp.patents
Subject: Re: Examples of software patent benefit sought
Message-ID: <3494@cluster.cs.su.oz.au>
Date: 13 Nov 91 19:02:17 GMT
References: <3341@cluster.cs.su.oz.au> <3407@cluster.cs.su.oz.au>
Sender: n...@cluster.cs.su.oz.au
Organization: IR
Lines: 47
Approved: pate...@cs.su.oz.au

Kim Letkeman writes:
>    I mean really hurt. Not inconvenienced.

Is it an ``inconvenience'' that nobody can use public-key encryption,
or some of the best compression methods, for free?

>    I guess I'm annoyed that the anti-patent people seem to want
>    software to be exempt from the forces of the market.

I want software to be *subject* to the forces of the market. Monopolies
always work against a free market. They are only tolerated when they
benefit society. The courts have always taken this view.

>    I would appreciate it if Mr. Eggert et al would stop assuming that
>    the pro patent people are talking about computer programs and/or
>    algorithms as written in someone's notebook.

But that's exactly what the most harmful patents cover. I don't know
what point you were trying to make in talking about multi-million-line
programs; RSA can be described in two lines, and LZW in three.

>    Software patents are not necessarily good. They are here.  Probably
>    to stay.

I doubt that.

>    To the academics in the crowd that are upset that they can't give
>    away someone else's algorithm I can only say "do you really think
>    it would be fair?"

James Storer and I discovered a compression algorithm independently
within the space of a few years. The idea behind that algorithm was an
idea whose time had come. Storer patented it. So you can't use it. You
think that's fair? I discovered the algorithm too; why shouldn't I have
any rights to it? Who benefits from Storer's monopoly?

> | I asked where software patents have helped.
> | No one has responded with a specific example.
> There have been several very well written (and well balanced IMO)
> articles with respect to the patent system and the service it provides
> to the technology community.

Stop hedging. Give an example. I suspect that a number of readers would
like you to stop beating around the bush. Name one software patent which
has shown a benefit to society. Well?

---Dan

Path: gmdzi!ieee.org!dorm.rutgers.edu!rutgers!gatech!asuvax!cs.utexas.edu!
swrinde!mips!think.com!rpi!batcomputer!munnari.oz.au!metro!cluster!ji...@microsoft.com
From: ji...@microsoft.com (Jim ADCOCK)
Newsgroups: comp.patents
Subject: Re: Examples of software patent benefit sought
Message-ID: <1991Nov13.211136.20400@microsoft.com>
Date: 13 Nov 91 21:11:36 GMT
References: <3341@cluster.cs.su.oz.au> <3407@cluster.cs.su.oz.au>
Sender: n...@cluster.cs.su.oz.au
Organization: Microsoft Corp.
Lines: 29
Approved: pate...@cs.su.oz.au

In article <3...@cluster.cs.su.oz.au> egg...@twinsun.com (Paul Eggert) writes:
|The only responses have been:
|
|  Jim Adcock wrote about a combined hardware/software device that has never
|  been patented, but (from the sketch presented) would be patentable under the
|  Diamond v Diehr principle allowing patents on hardware devices that contain
|  some software.

Would have been, if the principle had been established.  

Also:

At the same time I described software inventions at HP that were successfully
patented, and which were necessary to be patented, or our Japanese
competitors would have copied our inventions within a year.  So, 
until we had assurances from our lawyers that we could probably
successfully patent our ideas, we did not have management support
to develop them.  Once management had assurance that the inventions --
if they could be successfully developed -- would probably be patentable,
then management was willing to commit to the several years and several
hundreds of thousands of dollars necessary to develop this particularly
extremely complicated and narrowly applied invention.

So, in this case the availability of patents was the deciding factor
for society getting the algorithm *at all.*  If we couldn't have 
patented it, we would have never had developed it, and it is sufficiently
extremely non-obvious that no one else would ever come up with anything
similar.

But, of course, you neglected to quote these patents, because they
don't support your position.

Path: gmdzi!ieee.org!dorm.rutgers.edu!rutgers!gatech!psuvax1!wupost!
uwm.edu!rpi!batcomputer!munnari.oz.au!metro!cluster!egg...@twinsun.com
From: egg...@twinsun.com (Paul Eggert)
Newsgroups: comp.patents
Subject: Re: 1.2 million
Message-ID: <3441@cluster.cs.su.oz.au>
Date: 14 Nov 91 08:31:40 GMT
References: <3424@cluster.cs.su.oz.au>
Sender: n...@cluster.cs.su.oz.au
Lines: 29
Approved: pate...@cs.su.oz.au

0001811...@mcimail.com (Carl Oppedahl) writes:

>Carl Oppedahl has software patent clients,
>so it must be that his clients believe that software patents help.

Obviously there are some winners in the software patent business, or
else people wouldn't apply for them.  The biggest winners are the
software patent lawyers and the companies that can afford them.  But
the question is, do software patents benefit us as a whole?  Since
software patents are government-enforced monopolies that deliberately
infringe on our individual rights, the burden of proof is on software
patent proponents to show why we should put up with them.  So far in
this thread, we've seen several specific examples where they hurt
society, and not a single specific example where they help.


>So if we assume (which is not really justified) that these are software
>patents then of the 40,000 software patents, maybe thirty have been 
>discovered to be bad.  A thousand-to-one ratio.

You're misusing your (admittedly rough) statistics.  By your own
numbers, so far we've discovered thirty harmful software patents and
zero helpful ones.  You assume without foundation that the rest of the
40,000 software patents are helpful.  It's far more likely that the
vast majority of them are fairly obvious to those versed in the field,
and are thus a large net harm to society.  I can't think of a single
software patent that has had a positive influence on the field.  For
this we've spent hundreds of millions of dollars in patent fees and
litigation?

Path: gmdzi!ieee.org!dorm.rutgers.edu!rutgers!uwm.edu!rpi!batcomputer!
munnari.oz.au!metro!cluster!bar...@think.com
From: bar...@think.com (Barry Margolin)
Newsgroups: comp.patents
Subject: Re: 1.2 million
Message-ID: <ki4c2oINNg57@early-bird.think.com>
Date: 14 Nov 91 08:14:16 GMT
References: <3424@cluster.cs.su.oz.au>
Sender: n...@cluster.cs.su.oz.au
Organization: Thinking Machines Corporation, Cambridge MA, USA
Lines: 52
Approved: pate...@cs.su.oz.au
NNTP-Posting-Host: godot.think.com

In article <3...@cluster.cs.su.oz.au> 0001811...@mcimail.com (Carl Oppedahl) writes:
>I quite arbitrarily decided not to count software patents, but instead to 
>count patents containing the words "ALGORITHM" or "SOFTWARE" or "SOURCE CODE" 
>or "FLOWCHART".  That test surely misses some patents which some would say
>are software patents.  And it surely counts some patents which some would 
>say are not software patents.  But it's just orders of magnitude I'm trying to 
>approach, here.

>About 40,000 patents were found.

I think this is an overly broad definition of software patents, and could
be off by an order of magnitude or more.  The Supreme Court decision that
was taken as allowing software patents simply said that patents could not
be rejected simply because software is included as part of the the device.
I'll bet most of those 40,000 patents are for physical devices that include
programmed devices (e.g. anti-lock braking systems and electronic
fuel-injection systems), rather than being patents on the programs or
algorithms themselves; I don't think most of us would consider these to be
software patents.

"Flowchart" is an extremely bad choice, since flowcharts are often used for
describing hardware devices.  For instance, I've seen computer manuals that
use flowcharts to describe the execution of instructions, although the
actual implementation is in silicon.  Here's an example of the flowchart of
a rifle's firing mechanism:

	<Trigger depressed?> -- No --> [Do nothing]
	       Yes		    /
	<Safety disabled?> -- No --/
	       Yes		   |
	<Shell in chamber?> -- No -/
	       Yes
	     [Fire]

>So if we assume (which is not really justified) that these are software
>patents then of the 40,000 software patents, maybe thirty have been 
>discovered to be bad.  A thousand-to-one ratio.  The suggestion is that
>maybe the Patent Office is not doing so badly at distinguishing things that
>should be patentable from things that should not.

The main problem with this conclusion is that it assumes that the people
who judge patents "good" or "bad" are familiar with these 40,000 patents.
As you pointed out, the list of well-known software patents only has 60
items on it.

My conslusion is that most software patents aren't well publicized, not
that most software patents are (in some sense) OK.

-- 
Barry Margolin, Thinking Machines Corp.

bar...@think.com
{uunet,harvard}!think!barmar

Path: gmdzi!ieee.org!dorm.rutgers.edu!rutgers!ub!zaphod.mps.ohio-state.edu!
rpi!batcomputer!munnari.oz.au!metro!cluster!...@software.mitel.com
From: k...@software.mitel.com (Kim Letkeman)
Newsgroups: comp.patents
Subject: Re: Examples of software patent benefit sought
Message-ID: <3461@cluster.cs.su.oz.au>
Date: 14 Nov 91 14:09:58 GMT
References: <3341@cluster.cs.su.oz.au> <3407@cluster.cs.su.oz.au>
	<3416@cluster.cs.su.oz.au> <1991Nov13.065335.7592@kithrup.COM>
Sender: n...@cluster.cs.su.oz.au
Organization: MITEL Public Switching, Kanata, Ontario, Canada
Lines: 123
Approved: pate...@cs.su.oz.au


In article <1991Nov13.065335.7...@kithrup.COM> s...@kithrup.COM 
(Sean Eric Fagan) writes:

| In article <3...@cluster.cs.su.oz.au> k...@Software.Mitel.COM 
(Kim Letkeman) writes:
| >First, I don't believe that we'd all be using secure mail by now.
| 
| I would.  News, at least.

Just curious, but why would secure news benefit you?

| >And finally, not everyone on the planet wants to write a windowing
| >system. The biggies are already there: X, Windows 3, PM, Openlook,
| >etc. Anyone who wants to join this league can either afford the
| >license or will cross-license. If you're writing one for in house
| >use, then don't worry about the patent. Who the heck is going to
| >bother you?
| 
| Yeah.  Let's ignore the fact that X was written, not by a commercial
| organization, but by a university, and is available *free* for
| anyone who wants it.

I wasn't ignoring this fact. The patent wars X is going through are
no different from those that Microsoft, Digital Research, IBM, Apple
and Xerox (to name a few) have to go through to get a major
undertaking to market.

| Let's ignore the fact that this is exactly the type of thing that
| patents are supposed to encourage, and resulted in something that
| "benefited" the public (let's not get into a discussion about
| whether or not X is good for you 8-)).

I hardly think the patent system was invented so that University's
could go out and give stuff away that the rest of the world has to
charge money for. I don't know about the US educational system, but
the Canadian one is funded a great deal by the taxpayers, so it seems
a bit unfair for (effectively) the Government to be giving something
like X away when the rest of the world has to charge money for similar
systems.

By the way, I'm not saying that it would not be a good thing for X to
eventually dominate the world's window systems and give us a
relatively good standard, I'm merely addressing your point w.r.t. what
the patent system should or should not encourage IMO.

| Let's also ignore the apparantly little-known MGR window manager,
| also freely available, written by a small number of people.

How can I be accused of ignoring something few people have ever heard
of?

| Let's ignore the fact that if you're infringing on a patent, EVEN IN
| HOUSE, you can be forced to stop using the patented thing
| (algorithm, in this case).

Yes, you can. I said that you almost certainly would not, since noone
can look inside your organization to see how you do your in-house
stuff.

| In other words, let's ignore facts.  Right?

Ok. If you insist.

| >People pay to >have software developed (BIG BUCKS) so they deserve
| some protection >for a novel implementation.
| 
| Small software houses also develop software.  There are quite a few people
| who make a living doing things like porting X windows.  How many of these
| would be forced out of business if they had to have a patent search done on
| every single algorithm in their program?  And, of course, a patent search
| doesn't cover pending patents, meaning that they can be bitten years down
| the line and forced to pay royalties, or discontinue use.

Your specific example is covered by the fact that MIT is handling the
X window problem. If X survives, so will the little guys who make a
living porting it. But if X dies (does anyone actually believe it
could?), then these guys go out of business. Unless, of course, MIT
decides to stop them itself. Not likely, I suppose.

By the way, I don't believe in the "entitlement mentality". People are
not owed a living porting X. If they make a good living, good for
them. If, for any reason, the world changes and they can not support
themselves on X (you can substitute any equivalent situation here),
then they must react and find something else. It's the same for
everyone.

| >Programs are not patented. Simple algorithms are not patented. But
| >implementations are.
| 
| Uhm, what the hell do you call using XOR for cursor manipulation?  That's
| about as simple an algorithm as you can get!

Yes. Probably a bad patent. Everyone agrees.

| >Books don't make toasters go. There are no books inside your
| >microwave oven. A telephone switch isn't controlled by a book.
| 
| So?  Songs aren't books either, but they're not patentable.  And books
| aren't anything without the paper they're printed on, so I guess books
| should be patentable, as well.

I won't comment on whether I can actually understand this paragraph.

My point was that a book might be part of a novel invention once in a
millenia. Software, on the other hand, does novel things all the time.
Machines do different things when software is changed. This makes
software a key part of inventions every day. Patents are going to be
awarded for that.

| And the crowning piece of idiocy:

Thank you. Too eloquent for any response to do it justice.

| >   As has been stated numerous times, the typical one man show has
| >   nothing to fear because he is too small to bother with.

This may not have been brilliantly stated, but only the totally
paranoid would stop working on a project out of simple fear.

Also, this comes back to the intitlement thing again. Just because
software is in the hands of millions, does that mean we should let
everyone who wants to, take a hack at a new and improved windowing
system? Will that benefit society? I doubt it. But then, I'm sure I've
been wrong before.
--
Kim Letkeman	k...@Software.Mitel.COM

Path: gmdzi!ieee.org!dorm.rutgers.edu!rutgers!cs.utexas.edu!swrinde!
zaphod.mps.ohio-state.edu!rpi!batcomputer!munnari.oz.au!metro!cluster!
0001811...@mcimail.com
From: 0001811...@mcimail.com (Carl Oppedahl)
Newsgroups: comp.patents
Subject: 1.2 million
Message-ID: <3466@cluster.cs.su.oz.au>
Date: 14 Nov 91 18:18:00 GMT
Sender: n...@cluster.cs.su.oz.au
Lines: 72
Approved: pate...@cs.su.oz.au

Barry Margolin (bar...@think.com) writes:
 
>I think this [patents containing the words "ALGORITHM" or "SOFTWARE" or 
>"SOURCE CODE" or "FLOWCHART"] is an overly broad definition of software 
>patents, and could
>be off by an order of magnitude or more.  The Supreme Court decision that
>was taken as allowing software patents simply said that patents could not
>be rejected simply because software is included as part of the the device.
>I'll bet most of those 40,000 patents are for physical devices that include
>programmed devices (e.g. anti-lock braking systems and electronic
>fuel-injection systems), rather than being patents on the programs or
>algorithms themselves; I don't think most of us would consider these to be
>software patents.
 
I am very interested to read this comment.  Do other commenters, including
those who characterize themselves as being against software patents, agree
with Mr. Margolin?  Changing what he said a little, if a patent claims a 
physical device that includes programmed devices rather than being a patent 
in the program or algorithm itself, then it is not a software patent [and 
we will not quibble with the Patent Office and say that its software content 
should have rendered it unpatentable]?
 
Understand the consequences of saying yes.  If you say yes, then you have
perhaps unwillingly put yourself in agreement with me, a patent lawyer.
(I assume that for at least a few of our commenters this is undesirable.)
If you say yes, then you will find yourself committed to a point of view
that says that vanishingly few of the patents issued by the U.S. Patent Office
are software patents.  I have seen only one patent ever issued by the U.S. 
Patent Office where the claim says "I claim a computer program comprising ..."
and it is quite likely the Patent Office will not let it happen again.  So
you will find yourself in the possibly uncomfortable position of saying that 
the U.S. Patent Office, but for a tiny handful of lapses, is doing the right 
thing with software patent applications.
 
(One partner in my law firm has been getting patents issued of the type 
described by Mr. Margolin for many years, indeed long before Diamond v. Diehr, 
the oft-quoted Supreme Court case that supposedly opened the floodgates of 
software patent applications.)
 
Understand the further consequences of your saying yes.  An issued patent
with claims covering a physical device will cover that device and its 
equivalents.  Where the claimed device is a processor running a stored program, 
and the accused device is a different make of processor running a written-
from-scratch different stored program, the accused device might nonetheless 
be judged to infringe if it accomplishes substantially the same result 
in substantially the same way by substantially the same means.  So while 
pure programmers in the abstract need not worry much about patents, those
programmers who happen to program for physical devices that are directed to
physical ends will still have to worry about patents.  This puts them in the 
same position as makers of mousetraps, who also have to worry they might be
infringing a patent.
 
Notice I said "an issued patent", and before you get too worried recall that
the patent will be issued only if a patent examiner determines that the
claims (1) cover something novel (no one did it before);  (2) cover something
unobvious;  (3) are not merely a thinly-veiled attempt to in fact lock up all
ways of performing a previously known algorithm or computational process.
And before you get too worried recall also that the patent's existence does
not mean third parties are forbidden to do anything and everything mentioned
in the patent text.  Third parties are merely forbidden to do what is claimed
in the claims.
 
But if you say yes, I submit you are in fact saying that the present patent
system is doing everything just right regarding software patents.  (Except, 
perhaps, the pesky problem that there are not enough examiners, and that 
better mechanisms are needed to search the prior art, and so on.)
 
Comments, anyone?
 
Carl Oppedahl
30 Rockefeller Plaza
New York, NY  10112-0228

Path: gmdzi!ieee.org!dorm.rutgers.edu!rutgers!cs.utexas.edu!uunet!munnari.oz.au!
metro!cluster!...@Software.Mitel.COM
From: k...@Software.Mitel.COM (Kim Letkeman)
Newsgroups: comp.patents
Subject: Re: 1.2 million
Message-ID: <3471@cluster.cs.su.oz.au>
Date: 14 Nov 91 19:41:29 GMT
References: <3424@cluster.cs.su.oz.au> <3441@cluster.cs.su.oz.au>
Sender: n...@cluster.cs.su.oz.au
Organization: MITEL Public Switching, Kanata, Ontario, Canada
Lines: 38
Approved: pate...@cs.su.oz.au


In article <3...@cluster.cs.su.oz.au> egg...@twinsun.com (Paul Eggert) writes:

| So far in this thread, we've seen several specific examples where
| they hurt society, and not a single specific example where they
| help.

Now, this looks to me like proof by repeated assertion. A couple of
people have posted examples in the last couple of days which I thought
qualified.

| >So if we assume (which is not really justified) that these are software
| >patents then of the 40,000 software patents, maybe thirty have been 
| >discovered to be bad.  A thousand-to-one ratio.
| 
| You're misusing your (admittedly rough) statistics.  By your own
| numbers, so far we've discovered thirty harmful software patents and
| zero helpful ones.

I would suggest that assuming that there are 30 bad patents, 39,970
patents that make no difference, and zero good ones is a bit of an
abuse of statistics.

| You assume without foundation that the rest of the 40,000 software
| patents are helpful.  It's far more likely that the vast majority of
| them are fairly obvious to those versed in the field, and are thus a
| large net harm to society.  I can't think of a single software
| patent that has had a positive influence on the field.  For this
| we've spent hundreds of millions of dollars in patent fees and
| litigation?

I've stated in a previous article that I assume that the vast majority
of software patents are for very specialized mechanisms within large
and/or embedded systems. The kind, if you will, that give a company an
edge over the competition. That's why you never hear about them. I say
this because that's the kind that the software group I belong to gets
involved with and we seem to be doing pretty much what most "hi-tech"
(I know, yuk) companies are doing to get product out the door.
--
Kim Letkeman	k...@Software.Mitel.COM

Path: gmdzi!ieee.org!dorm.rutgers.edu!rutgers!cs.utexas.edu!uunet!
munnari.oz.au!metro!cluster!0001811...@mcimail.com
From: 0001811...@mcimail.com (Carl Oppedahl)
Newsgroups: comp.patents
Subject: Benefits of Patents (was 1.2 million)
Message-ID: <3469@cluster.cs.su.oz.au>
Date: 14 Nov 91 18:26:00 GMT
Sender: n...@cluster.cs.su.oz.au
Lines: 92
Approved: pate...@cs.su.oz.au

Paul Eggert (egg...@twinsun.com) writes:
 
>Obviously there are some winners in the software patent business, or
>else people wouldn't apply for them.  The biggest winners are the
>software patent lawyers and the companies that can afford them.  But
>the question is, do software patents benefit us as a whole?
 
**
Thomas M. Breuel (t...@ai.mit.edu) writes:
 
>Granting anybody a monopoly on anything can only benefit them
>personally.  I'm not surprised that businesses are lining up for
>software patents.  
 
>The question is whether the industry and society as a whole benefits
>from such practices, in particular, the granting patents on software.
>The purpose of the patent system is to encourage the creation and
>dissemination of information ...
 
At risk of sounding like I am repeating myself from a previous posting, let
me briefly go over the benefits the patent system offers to Mr. Eggert and 
to Mr. Breuel and to the rest of us as a whole.
 
Patents are publications, the information content of which benefits everybody.
To get a patent, the applicant must provide all the information required to
enable one skilled in the art to practice the invention.  That means the 
inventor is revealing all.  Think of the wealth of information in those 
patents.
 
Trivial example.  I wonder how Medeco keys work.  I read the patent number off
the key, order the patent for $1.50, and read it.  All in one place, everything
I need to know to know how they work.
 
More potent example.  Someone comes up with something really neat, and 
commercializes it only in a way that keeps the workings concealed from view.
This could be a mechanism in a housing that self-destructs if you open it.
Or a product (e.g. a Michelin radial tire) made using a technique that Michelin
keeps to itself as a trade secret.  Or a piece of software distributed in 
executable code only with a positive obligation on end users not to disassemble
or reverse compile it.  All that inhibits the free flow of ideas.  All sorts of
people who might improve on it are denied the information that might help them
do so.  And yet, if the mechanism, or technique, or software, is patented, then
its innermost workings are revealed to everyone, at a cost of $1.50.  All of
society benefits.
 
Another example of how everyone benefits.  "Everyone" includes independent
programmers and small companies, not just "the companies that can afford" 
patents.  Those programmers and small companies can protect their market 
position against the Goliaths of the world (see my posting of this date
regarding the comments of Mr. Adcock).  That promotes innovation, competition,
creativity, and open disclosure.
 
Mr. Eggert says:
 
>Since
>software patents are government-enforced monopolies that deliberately
>infringe on our individual rights, the burden of proof is on software
>patent proponents to show why we should put up with them.  
 
If David invents some novel, unobvious solution B to a problem on day A, then
prior to day A no one, anywhere, had B.  So if David gets a 
patent on B, no one is worse off by the grant of the patent, because no one 
even knew to try using B to solve the problem.  David gave the world B, which
the world did not have before.  No one's "individual rights" were infringed,
deliberately or otherwise, by the grant of the patent on B.
 
Mr. Eggert is using rather biased language here.  Ownership of a car is a
government-enforced monopoly too.  My ability to stop someone else from 
using my car, and my ability to turn to the government (the police, the 
court system, the prison system) to enforce my wishes regarding the car
counts as government-enforced.  Does that deliberately infringe on the rights
of the would-be car thief?  Surely not.
 
One might respond that a car is not an idea.  If someone else uses my idea I
am not denied the ability to keep using it myself, the argument goes, 
and that is different from the situation with a stolen car.  
 
To see why that argument is no good, realize that if someone else uses my idea
I do lose out, in simple, easy to quantify terms.  If I write a song and 
someone else records it and sells the records, it is little consolation to me
that I could continue to perform the song in the privacy of my home.  I want
the profits from that record.  That's why we have the government-enforced
monopoly that is called copyright.  Does that government-enforced monopoly
infringe on the individual rights of the pirate who runs off copies of my
songs in his basement filled with tape duplicators?
 
I trust the thoughts mentioned above show that patents, even software patents,
help all of us.
 
Carl Oppedahl
30 Rockefeller Plaza
New York, NY  10112-0228

Path: gmdzi!ieee.org!dorm.rutgers.edu!rutgers!cs.utexas.edu!wupost!
spool.mu.edu!munnari.oz.au!metro!cluster!bar...@think.com
From: bar...@think.com (Barry Margolin)
Newsgroups: comp.patents
Subject: Re: Benefits of Patents (was 1.2 million)
Message-ID: <ki83clINN3qp@early-bird.think.com>
Date: 15 Nov 91 18:10:29 GMT
References: <3469@cluster.cs.su.oz.au>
Sender: n...@cluster.cs.su.oz.au
Organization: Thinking Machines Corporation, Cambridge MA, USA
Lines: 32
Approved: pate...@cs.su.oz.au
NNTP-Posting-Host: gandalf.think.com

In article <3...@cluster.cs.su.oz.au> 0001811...@mcimail.com (Carl Oppedahl) 
writes:
>If David invents some novel, unobvious solution B to a problem on day A, then
>prior to day A no one, anywhere, had B.

That's the crux.  Most of the s/w patents we've seen are either not novel,
or are obvious.

Consider the XOR cursor patent (sorry, I don't know the number).  Suppose I
buy a terminal, and the documentation says that it uses a patented cursor
technology (stating that something is patented is often intended to impress
the reader -- perhaps it's the (often untrue) implication that the patented
device was invented by the vendor, suggesting his technical expertise).  I
look at the cursor, and say to myself, "interesting, I wonder what they're
using instead of XOR to make the pixels under the cursor alternate colors"
(since every decent programmer knows that XOR is a trivial way to invert
bits, and anyone familiar with raster graphics knows that pixels are
represented as memory bits).  I send my $1.50 to the patent office and find
out that they're doing precisely that.  Not only that, but they didn't even
invent it, they licensed it from someone else; I have no idea whether this
means that they developed it independently and then found out that it
required licensing, or their programmers are so dumb that they had to look
up how to make a blinking cursor.

Pike's backing store patent might be considered unobvious, but it was
hardly novel.  There were window systems being developed in labs all over
the country at the time, and they all used variants of this.

-- 
Barry Margolin, Thinking Machines Corp.

bar...@think.com
{uunet,harvard}!think!barmar

[pjt- note novelty and obviousness take on particular meaning in
	reference to patent law...
	]

Path: gmdzi!unido!fauern!Sirius.dfn.de!darwin.sura.net!haven.umd.edu!ames!
agate!spool.mu.edu!munnari.oz.au!metro!cluster!pate...@cs.su.oz.au
From: pate...@cs.su.oz.au (Peter Treloar)
Newsgroups: comp.patents
Subject: Re: Examples of software patent benefit sought
Message-ID: <3495@cluster.cs.su.oz.au>
Date: 18 Nov 91 07:44:01 GMT
References: <3341@cluster.cs.su.oz.au> <3407@cluster.cs.su.oz.au>
Sender: n...@cluster.cs.su.oz.au
Organization: IR
Lines: 54
Approved: pate...@cs.su.oz.au

In article <3...@cluster.cs.su.oz.au> Dan Bernstein writes:

>Stop hedging. Give an example. I suspect that a number of readers would
>like you to stop beating around the bush. Name one software patent which
>has shown a benefit to society. Well?

I would give the example of  US4744028 which is (I think ?) Karmarkar's 
linear programming patent titled: " Methods and Apparatus For Efficent
Resource Allocation". 

This patent shows an improved method for linear programming that is
apparently far superior to previous methods...

I quote from page 2:
	"The best known prior art approach to solving allocation problems
	posed as linear programming models is known as the simplex method,
	invented by George B. Danzig in 1947...."

Hence this is the first substantial improvement for approximately 50
years.

Now, I think one of the major arguements for software patents is to 
look at the alternatives... Here we have a method that, I assume, is
totally isolated from its outputs and inputs. The improvement made is one of
time. Hence if this patent was not published you would, presumably
have no guaranteed way of finding out how they work their magicial
method no matter how much money/labour you throw at it. So assume that
nobody could do it without seeing the document.

Hence the patent owner (guess who - AT&T :-)), if they seriously wanted
to obtain the maximum financial benefit from their discovery (which IMHO can
just as easily be called a mathematical theorem) could keep the method
(essentially an algorithm) secret for the next, say, 200 years and
allow its usage only to the highest bidder.... The user only being
allowed to send them a tape and gets back the results on tape and that
is all....

Note, dont quote me on this, but I think this was one of the very
early (ie Middle Ages) reasons for patents or copyright. The story goes that
when printing was invented in Germany, they tried to keep it secret
much to the consternation of surrounding Kingdoms. Anyway apparently
the English KIDNAPPED one of the German printers sworn to secrecy and
by means of threats and promises of various types were able to coerce
him/her into revealing the secret. Of course even this may not be an
option given the above situation.

This to me is the main reason why software patents are necessary, for
I ask how can I best take advantage of my invention given that I am
not allowed to patent it... And this was what I would suggest could
occur.

====================================================================
Peter Treloar - comp.patents Moderator
pate...@cs.su.oz.au

Path: gmdzi!ieee.org!dorm.rutgers.edu!rutgers!micro-heart-of-gold.mit.edu!
xn.ll.mit.edu!ames!agate!spool.mu.edu!munnari.oz.au!metro!cluster!
ji...@microsoft.com
From: ji...@microsoft.com (Jim ADCOCK)
Newsgroups: comp.patents
Subject: Re: 1.2 million
Message-ID: <1991Nov18.223517.16795@microsoft.com>
Date: 18 Nov 91 22:35:17 GMT
References: <3466@cluster.cs.su.oz.au>
Sender: n...@cluster.cs.su.oz.au
Organization: Microsoft Corp.
Lines: 55
Approved: pate...@cs.su.oz.au

In article <3...@cluster.cs.su.oz.au> 0001811...@mcimail.com (Carl Oppedahl) 
writes asking rhetorically is a software patent an invention that is
*only* software:

To which I answer in agreement that an invention that is only software
is not patentable, because such cannot *do* anything.  Someone could
read the program listing and execute the steps specified there in
one's head, I suppose, but clearly then that would be a mental process
that is not patentable.

When that software is executed in a machine for some useful business
purpose, then that software becomes a business process, of which
the software is the controlling element.  It is that business process
that is patentable.

In the early days of patenting software to make this point clear the 
people writing the patents had to laboriously write the patent in such
a manner as to make it clear that the software was only to be considered
a controlling element in the overall process -- albeit that controlling
element might be what is truly novel in the whole process.

But, once it becomes clear what we are talking about when we are patenting
software, then it becomes no longer necessary for people drafting software
patents to go through all these laborious push-ups to demonstrate the
hardware involvement in the overall process -- rather, it becomes obvious
to anyone skilled in the art that when we talk of a "software invention"
we are referring to a complete process executing on an actual physical
machine -- typically called a "computer" -- of which the novel aspect --
the software -- is simply a controlling element.

Thus people who are writing "software patents" are simply stating that
it is obvious to anyone skilled in the arts -- including hopefully
the patent examiner -- that the software is executed in a physical
machine -- and therefore we should not have to continually belabor the
obvious whenever we submit a "software patent" to the patent office.
Stating or not stating the blatently obvious in a patent submission
should not have any impact in the patent's allowability.  Rather,
the patent application should only have to state the new and non-obvious,
placing that in the context of the established.

Thus said, the implication is that merely distributing a program,
either in listing form or in executable form, does not in itself
violate anyone's patent.  You have to actually execute the program
to infringe.  However, a court might well find that a software company
is distributing software with the intent that the purchaser use the
software without obtaining a license from the inventor first, such
distribution then being prohibited.  Conversely, the software might
be distributed under the requirement that the end user obtain license
from the patent holder before use. [Such restriction might be acceptable
for professional software, but hardly would be acceptable when 
distributing software to the non-professional user.  You can't expect
to get away with saying in the shrink-wrap "Oh by the way, make sure
you obtain all applicable liceness]

This also implies that the static form of the software -- say a source
listing -- can be copyright, while simultaneous the dynamic form --
the software process -- can be patented.

Path: gmdzi!ieee.org!dorm.rutgers.edu!rutgers!micro-heart-of-gold.mit.edu!
xn.ll.mit.edu!ames!agate!spool.mu.edu!munnari.oz.au!metro!cluster!
mcgre...@atherton.com
From: mcgre...@atherton.com (Scott McGregor)
Newsgroups: comp.patents
Subject: Re: Examples of software patent benefit sought
Message-ID: <3513@cluster.cs.su.oz.au>
Date: 19 Nov 91 00:36:08 GMT
Sender: n...@cluster.cs.su.oz.au
Lines: 72
Approved: pate...@cs.su.oz.au

In article <3...@cluster.cs.su.oz.au>, brns...@KRAMDEN.ACF.NYU.EDU (Dan Bernstein) 
writes:

> Is it an ``inconvenience'' that nobody can use public-key encryption,
> or some of the best compression methods, for free?

It is merely an inconvenience for me if I can otherwise use them 
at a fair cost. It would have been a greater inconvenience for me 
if I could not use them at all because no one had published them 
or commercialized them.  A polonomial time solution for a large class 
of real traveling salesman problems would be commercially valuable, 
even if it was not free, but it hasn't been developed yet.  
If an economic incentive to a developer causes it to be developed 
sooner, that would benefit many people.

> James Storer and I discovered a compression algorithm independently
> within the space of a few years. The idea behind that algorithm was an
> idea whose time had come. Storer patented it. So you can't use it. You
> think that's fair? I discovered the algorithm too; why shouldn't I have
> any rights to it? Who benefits from Storer's monopoly?

Who benefits from Storer's monopoly? Anyone who uses his algorithm.  
Bernstein says we can't use Storer's algorithm but this is untrue--we 
can by paying a license. Before it was commercialized we were all 
the poorer because we didn't have such a good way 
to compress information. Now it is commercialized it is a form of 
competative advantage to those who would pay the cost of its use. The 
disadvantage is that it costs more than the competing (nonpatented) 
alternatives which may not be quite so good. The market pressures force 
licensing costs to be low enough that it will be more economical to use 
than not use, and so society as a whole benefits.  Admittedly they may 
benefit less than they would have if it were free.  But surely they 
benefit more than if it were not commercialized at all.  

Is it fair that Storer benefited and Bernstein did not? I think this 
is not a question of benefit to society from the creation of wealth, 
but merely a question of redistribution of wealth.  The creation of the 
new algorithm enables some new products to exist, employing additional 
people, and providing the additional value that the new products provide.  
There are lower costs (in storage and/or transmission) for those who 
buy them.  That's the true wealth to society. But what did Storer do? 
Storer attempted to commercialize that discovery and now I can buy it.  
Bernstein did not.  Bernstein did not publish it before Storer attempted 
to patent it (otherwise it would be prior art)--Bernstein may have been 
richer but society was poorer as it remained a secret
.  Bernstein did not distribute it publicly before Storer attemtped to patent 
it (again prior art), so again we couldn't buy it and benefit from it. 
Storer invented it and commercialized it, so that thousands of users wouldn't 
have to discover it for themselves, and in fact may now go on to innovate 
in other areas.  Surely it is fair that Storer receive some 
benefit from having made this available for
us to use and benefit from. But what did Bernstein offer us--only the 
opportunity to invest our time and effort to rediscover these things 
for ourselves. Of course, we can rediscover this for ourselves if 
we want. But why bother when we could be working 
on something new?

> Stop hedging. Give an example... Name one software patent which
> has shown a benefit to society. Well?

How about the Storer patent then. A patented compression algorithm 
exists now.  You don't need to reinvent one.  You can license one.  
And once you license it, you can compress data and save on transmission 
and/or storage costs. With the money you save on
 storage costs, and the time saved not reinventing one, perhaps you 
can invent and commercialize your own labor saving device. 
Publish it, distribute or commercialize it, but don't keep it a secret 
for others to have to solve for themselves.

"The solution is left to the reader" is not the most economic means of 
advancing commerce; a good solutions to problems book would be better.

Scott McGregor

Path: gmdzi!unido!mcsun!uunet!munnari.oz.au!metro!cluster!...@Software.Mitel.COM
From: k...@Software.Mitel.COM (Kim Letkeman)
Newsgroups: comp.patents
Subject: Re: Examples of software patent benefit sought
Message-ID: <3523@cluster.cs.su.oz.au>
Date: 19 Nov 91 01:47:10 GMT
References: <3341@cluster.cs.su.oz.au> <3407@cluster.cs.su.oz.au>
	<3494@cluster.cs.su.oz.au>
Sender: n...@cluster.cs.su.oz.au
Organization: MITEL Public Switching, Kanata, Ontario, Canada
Lines: 66
Approved: pate...@cs.su.oz.au


In article <3...@cluster.cs.su.oz.au> brns...@KRAMDEN.ACF.NYU.EDU 
(Dan Bernstein) writes:

| I want software to be *subject* to the forces of the market.
| Monopolies always work against a free market. They are only
| tolerated when they benefit society. The courts have always taken
| this view.

Yes, but a patent is not a monopoly. Properly exercised, it provides
control of the technology. Several people have pointed out that abuse
of that control can result in the loss of the patent.

| But that's exactly what the most harmful patents cover. I don't know
| what point you were trying to make in talking about multi-million-line
| programs; RSA can be described in two lines, and LZW in three.

I was illustrating (perhaps unclearly) my own opinion that the vast
majority of patents in software cover specialized areas of large
systems, not algorithms so trivial that they can be expressed in a
couple of lines.

| >    Software patents are not necessarily good. They are here.  Probably
| >    to stay.
| 
| I doubt that.

Ok. Another data point.

| James Storer and I discovered a compression algorithm independently
| within the space of a few years. The idea behind that algorithm was an
| idea whose time had come. Storer patented it. So you can't use it. You
| think that's fair? I discovered the algorithm too; why shouldn't I have
| any rights to it? Who benefits from Storer's monopoly?

You do have rights to it. License the algorithm from him if you can
make use of it. If you merely want to continue researching the area,
you don't even need the license.

Think about it, your name will be forgotten just as surely as the
"other" guy who invented the telephone. He probably didn't think it
was fair either.

| Stop hedging. Give an example. I suspect that a number of readers would
| like you to stop beating around the bush. Name one software patent which
| has shown a benefit to society.

The nature of this discussion gives a distinct advantage to the
anti-patent camp. They can merely trot out a few spectacular blunders
and shout "bad bad bad!!!!"

The other side of the issue is harder to argue. Can anyone say for
sure whether any individual patent (excluding the aforementioned) in
any field was "good", "bad" or "indifferent"?

If you (and others) feel that this admission loses the argument for
the pro-patent camp, then I can only say that I wish the world were as
binary for me. Good. Evil. Simple.

I believe that the stated goals of the patent system are "good" and
that the results so far are acceptable. I don't think they are
screwing up as often as others do, but I agree that when they do it's
a beauty.

| Well?

Yes, I'm well thanks. And you?
--
Kim Letkeman	k...@Software.Mitel.COM

Path: gmdzi!unido!mcsun!uunet!munnari.oz.au!metro!cluster!egg...@twinsun.com
From: egg...@twinsun.com (Paul Eggert)
Newsgroups: comp.patents
Subject: Re: Benefits of Patents (was 1.2 million)
Message-ID: <3519@cluster.cs.su.oz.au>
Date: 19 Nov 91 03:51:41 GMT
References: <3469@cluster.cs.su.oz.au>
Sender: n...@cluster.cs.su.oz.au
Lines: 27
Approved: pate...@cs.su.oz.au

0001811...@mcimail.com (Carl Oppedahl) writes:


>Patents are publications, the information content of which benefits everybody.

Patents are also monopolies, the enforcement of which hurts everybody
(except the monopolies' owners).  In deciding whether we should grant
patents, we should compare the benefits to the harms.

In the case of software, the benefits of patent publications are few.
Few fundamental software discoveries are published in patent form.
There is no evidence that allowing software patents has encouraged
publication.  Few textbooks contain any material based on software
patents.  Few programmers consult software patents except to avoid the
techniques described therein.  In short, as a means of publishing
software techniques, patents are a waste of time and money.


>If I write a song and 
>someone else records it and sells the records, it is little consolation to me
>that I could continue to perform the song in the privacy of my home.  I want
>the profits from that record.  That's why we have the government-enforced
>monopoly that is called copyright.

Exactly.  Copyright monopoly is also sufficient to protect software.
Therefore, we don't need to grant software patents, which have few benefits
and demonstrated harms.

Path: gmdzi!unido!mcsun!uunet!munnari.oz.au!metro!cluster!brns...@KRAMDEN.ACF.NYU.EDU
From: brns...@KRAMDEN.ACF.NYU.EDU (Dan Bernstein)
Newsgroups: comp.patents
Subject: Re: Examples of software patent benefit sought
Message-ID: <3524@cluster.cs.su.oz.au>
Date: 19 Nov 91 05:52:30 GMT
References: <3341@cluster.cs.su.oz.au> <3407@cluster.cs.su.oz.au>
Sender: n...@cluster.cs.su.oz.au
Organization: IR
Lines: 47
Approved: pate...@cs.su.oz.au

In response to my challenge, Peter suggests that Karmarkar's algorithm
(USP 4744028) has shown a benefit to society. In support of this, he
says that the algorithm is a ``substantial improvement'' over the
simplex method, and that the patent owner would have kept the algorithm
secret for many years if a patent had not been available.

Just for fun I'll demolish this argument piece by piece. First of all,
Karmarkar's algorithm is an incredibly poor algorithm in practice. It is
far more difficult to implement than the simplex method. It runs much
more slowly except on very large, contrived examples. It's of some
theoretical interest because it's guaranteed to run in polynomial time,
and subsequent refinements have made it reasonably fast, but I don't
know anyone who would willingly use it in place of simplex. About the
only exceptions are problems where it's important to get an answer in a
fixed amount of time.

So there's no chance that the patent owner would have been able to make
money off the method if a patent hadn't been available. (The point of
getting a patent for such a poor method, in case anyone's wondering, is
that the patent can still apply to later improvements which make the
method truly useful. That hasn't happened for Karmarkar's algorithm---
yet.) Would Karmarkar have kept his algorithm secret, as Peter suggests?
Of course not. He would have published somewhere, and the news would
have been of some interest to the computer science community. That's
what people do with inventions in mathematics and computer science.
That's what AT&T, IBM, and other big companies were doing for years
before anyone had heard of software patents.

As is, the patent is still out there, and if someone does come up with a
variation that's better than simplex, we'll all have to pay the patent
owner for the privilege of using what would otherwise have been free.
That's a benefit?

Finally, is anyone willing to claim that Karmarkar's algorithm is not a
procedure for solving a mathematical problem? I say it is. As such, it
is not subject to a patent. There isn't any question here. The courts
have defined ``mathematical algorithm'' as a procedure for solving a
mathematical problem, and it's well established that mathematical
algorithms are not subject to patents. Now if only patent examiners were
competent to recognize mathematics, the software patent problem would go
away.

My challenge still stands. Name one software patent which has benefited
society. We've all seen dozens of examples of software patents which
have hurt. If you think any software patents have helped, name them!

---Dan

Path: gmdzi!unido!mcsun!uunet!munnari.oz.au!metro!cluster!egg...@twinsun.com
From: egg...@twinsun.com (Paul Eggert)
Newsgroups: comp.patents
Subject: Re: Examples of software patent benefit sought
Message-ID: <3525@cluster.cs.su.oz.au>
Date: 19 Nov 91 06:25:25 GMT
Sender: n...@cluster.cs.su.oz.au
Lines: 46
Approved: pate...@cs.su.oz.au

A while ago I asked for specific examples where software patents have helped,
given their demonstrated harm in specific examples like the patents for public
key cryptosystems, XOR cursors, and backing store.  Responses included:

  A.  Peter Treloar suggested the linear programming patent developed at AT&T
  Bell Labs by Karmarkar.  This was an advance over previous technique, but it
  is a weak example of software patent benefits, for three reasons.

    1.  The patent does not disclose crucial details of AT&T's method,
    thus removing the principal claimed benefit to society of software patents.

    2.  AT&T's method was based on an earlier idea developed in the Soviet
    Union by Khachian, which had recently attracted wide attention; if AT&T
    had not invented the patented algorithm, someone else was likely to
    anyway.  (Needless to say, Khachian's idea was _not_ patented.)

    3.  AT&T would most likely have published this algorithm anyway even if
    software patents were not allowed.  It did publish many other software
    ideas developed at Bell Labs before software patents were allowed.  Since
    academic competition in the area was fierce, since the result was
    prestigious for AT&T, and since further necessary improvements were far
    more likely if it were published than if it were kept secret, there was
    strong pressure to publish.

  B.  Scott McGregor suggested the Storer compression algorithm.  This is an
  example of harm, not benefit.  The algorithm was independently and nearly
  simultanously reinvented by Dan Bernstein, who promptly published it (before
  Storer's patent was granted, but after it was filed).  So it is a net loss
  to society to have granted this particular monopoly.

  C.  Jim Adcock wrote about two devices (one music, one measurement).  But
  they both apparently fall under the Diamond v Diehr principle allowing
  patents on hardware devices or processes that contain some software.  I was
  asking for software patents, so these examples don't apply.

  D.  Mike Waters claimed to know several examples, but wouldn't cite any.

Surely software patent proponents can come up with better examples than this.


PS.  Many people also gave general arguments about software patents.
But I was asking for specific examples, not general arguments or analogies.
Since there are theoretical arguments both for and against software patents,
specific examples carry considerable weight.  So far, we have seen many
examples of harmful software patents, and not one good, specific example of a
beneficial one.|

Path: gmdzi!ieee.org!dorm.rutgers.edu!rutgers!dimacs.rutgers.edu!seismo!
uunet!munnari.oz.au!metro!cluster!...@Software.Mitel.COM
From: k...@Software.Mitel.COM (Kim Letkeman)
Newsgroups: comp.patents
Subject: Re: Benefits of Patents (was 1.2 million)
Message-ID: <3507@cluster.cs.su.oz.au>
Date: 18 Nov 91 13:36:28 GMT
References: <3469@cluster.cs.su.oz.au> <ki83clINN3qp@early-bird.think.com>
Sender: n...@cluster.cs.su.oz.au
Organization: MITEL Public Switching, Kanata, Ontario, Canada
Lines: 59
Approved: pate...@cs.su.oz.au


In article <ki83clINN...@early-bird.think.com> bar...@think.com (Barry Margolin) 
writes:

| In article <3...@cluster.cs.su.oz.au> 0001811...@mcimail.com (Carl Oppedahl) 
| writes:
| >If David invents some novel, unobvious solution B to a problem on day A, then
| >prior to day A no one, anywhere, had B.
| 
| That's the crux.  Most of the s/w patents we've seen are either not novel,
| or are obvious.

Actually, I think you have found the crux. A lot of
anti-software-patent argument focuses on the XOR and backing store
patents as being neither novel nor unobvious. There are two points
worth arguing against the stated assumptions:

1. These are two patents out of many thousand, and they serve only to
   underline the fact that patent examiners are human. I know, there
   are a few more patents that people dislike, but face it, these are
   the ones that everyone loves to hate.

2. Obviousness is hard to define. It has already been pointed out many
   times that ideas become extremely obvious once documented. We all
   share an unfortunate tendency to believe that we would come up with
   the same solution to the problem covered by these controversial
   patents. It's just human nature.

| Consider the XOR cursor patent (sorry, I don't know the number).  Suppose I
| buy a terminal, and the documentation says that it uses a patented cursor
| technology (stating that something is patented is often intended to impress
| the reader -- perhaps it's the (often untrue) implication that the patented
| device was invented by the vendor, suggesting his technical expertise).  I
| look at the cursor, and say to myself, "interesting, I wonder what they're
| using instead of XOR to make the pixels under the cursor alternate colors"
| (since every decent programmer knows that XOR is a trivial way to invert
| bits, and anyone familiar with raster graphics knows that pixels are
| represented as memory bits).  I send my $1.50 to the patent office and find
| out that they're doing precisely that.  Not only that, but they didn't even
| invent it, they licensed it from someone else; I have no idea whether this
| means that they developed it independently and then found out that it
| required licensing, or their programmers are so dumb that they had to look
| up how to make a blinking cursor.

Language like "every decent programmer" and "are so dumb" simply
underscores the natural tendency to believe in one's own superior
ability. The two patents discussed in the referenced article are the
two easiest targets on the planet. I don't think these assertions work
nearly as well as we get into the more specialized software patents
that are the norm.

| Pike's backing store patent might be considered unobvious, but it
| was hardly novel.  There were window systems being developed in labs
| all over the country at the time, and they all used variants of
| this.

Again, an easy target by now. Especially when one considers the vast
number of articles arguing exactly this point.



--
Kim Letkeman	k...@Software.Mitel.COM

Path: gmdzi!ieee.org!dorm.rutgers.edu!rutgers!news.cs.indiana.edu!spool.mu.edu!
munnari.oz.au!metro!cluster!0001811...@mcimail.com
From: 0001811...@mcimail.com (Carl Oppedahl)
Newsgroups: comp.patents
Subject: benefits of patents
Message-ID: <3563@cluster.cs.su.oz.au>
Date: 19 Nov 91 21:00:00 GMT
Sender: n...@cluster.cs.su.oz.au
Lines: 147
Approved: pate...@cs.su.oz.au

In response to:
>>Patents are publications, the information content of which benefits everybody.
 
Paul Eggert (egg...@twinsun.com) writes:
 
>Patents are also monopolies, the enforcement of which hurts everybody
>(except the monopolies' owners).  In deciding whether we should grant
>patents, we should compare the benefits to the harms.
 
>In the case of software, the benefits of patent publications are few.
>Few fundamental software discoveries are published in patent form.
>There is no evidence that allowing software patents has encouraged
>publication.  Few textbooks contain any material based on software
>patents.  Few programmers consult software patents except to avoid the
>techniques described therein.  In short, as a means of publishing
>software techniques, patents are a waste of time and money.
 
>>If I write a song and 
>>someone else records it and sells the records, it is little consolation to me
>>that I could continue to perform the song in the privacy of my home.  I want
>>the profits from that record.  That's why we have the government-enforced
>>monopoly that is called copyright.
 
>Exactly.  Copyright monopoly is also sufficient to protect software.
>Therefore, we don't need to grant software patents, which have few benefits
>and demonstrated harms.
 
Let me respond item by item.
 
>Few fundamental software discoveries are published in patent form.
 
You can't have it both ways. If fundamental software discovery
X (call it "FSD X") was not published in patent form, that is, the inventor
did not filed a patent application and/or it did not issue as a patent, then
in either case there is no patent so the harm-to-the-world which he sees in
a patent does not exist with respect to FSD X.  If FSD Y _was_ published
in patent form, then this counts as evidence tending to disprove Mr. Eggert's
assertion. 
 
>There is no evidence that allowing software patents has encouraged
>publication.
 
There is loads and loads of evidence.  I suspect if you would interview 
Profs. Rivest et al, they would tell you that they were tempted to keep secret
their methods for authenticating and encrypting messages, so as to 
commercialize them in a way that reduced the prospects of someone running off
with their ideas.  I suspect they would tell you that the ability to file
patent applications made it easy for them to decide to disclose quite openly
their methods (after having filed patent applications).  And so on and so on
through almost every single patent owned by any university or hospital or
research center.  Universities, especially, in recent years are trying to get
money to survive from every possible source, including commercializing
inventions that were created in their facilities.  They want their professors
to be able to publish but want to protect their ability to make money.
Patents (and that subset of them that have to do with software) let 
universities do that.
 
Are any of our academically based commenters (with .edu addresses, perhaps?) 
willing to share the experiences of their faculty patent committees or
technology licensing programs in this area?
 
>Few textbooks contain any material based on software patents.
 
Wait a minute.  
One would have to have gone out and read most textbooks, and then 
checked on an idea-by-idea basis to make sure that as for each idea it can 
be said that no patent application was filed?  I have read dozens of 
engineering texts where the things taught are protected by patents (and 
almost certainly would not have been disclosed to the textbook publisher
had it not been possible to patent it).  Many of these are pure hardware
or a mix of hardware and software.  Networking books.  Books on cellular
telephone technology.  Books on network security.  In many cases the author
of the text probably does not even know that what he is writing (drawn, say,
from someone else's lecture or publication) was only possible because that
someone else was willing to lecture or publish _only_ because it was possible 
to apply for a patent on it.
 
A few illustrations.  Yee's book on cellular telephone systems engineering.
In this excellent book, Mr. Yee describes frequency usage plans for adjacent 
cells.  (That and everything else, from top to bottom, about how one
engineers a system.)  Some of the frequency reuse planning techniques are 
the subjects of patents, yet it would not surprise me a bit if Mr. Yee 
has not particularly sought out or kept track of which are patented and which 
are not.  The valuable service the author provides is to survey some of the
many techniques, and those who openly described their techniques to him might
not have done so if they had not been able to get patents.
 
Or consider any of several excellent books on the Intel 80X86 family of
processors.  They describe how, for example, the math co-processor is 
internally configured.  One consequence is that programmers can write 
better code.  Neither the authors of the books, nor the readers of the books,
are likely to have any particular reason to be aware of Intel's patents
directed to the co-processors.  (The co-processors, by the way, contain all
sorts of software.)  Yet I feel it likely Intel would tell 
you, if asked, that it would not have been so forthcoming about the 
inner workings of the co-processors had it not been able to apply for 
patents first.
 
>Few programmers consult software patents except to avoid the techniques 
>described therein.  
 
That's a pity, frankly.  People who make mousetraps and computer hardware
and modems and cranes and ski lifts and antilock brake systems all choose
to look at patents.  It helps them figure out if they'll get sued, it helps
them keep on top of what their competitors are up to, it stimulates their
creativity.  It helps them understand the inner workings of their 
competitors' products.  A great opportunity, and yet if Mr. Eggert is correct,
few programmers take advantage if the opportunity.
 
>In short, as a means of publishing software techniques, patents are a waste 
>of time and money.
 
The patents themselves are not the _entirety_ of the publication benefit 
that comes from their being a patent system.  
While I do urge that the patents themselves are an
invaluable source of published information, I need to remind everyone that
I also urge that the availability of the patent system also frees people to 
publish elsewhere (technical journals, books, magazines) when they might 
otherwise feel constrained to clam up and not publish at all so as to keep 
things secret.
 
Those who do not believe me should browse a while in Inspec, a database on 
Dialog that abstracts and cross-indexes articles from technical journals, 
from magazines, from patents, and other sources, in the areas of physics, 
electronics, and computing.  In the computing subfile the areas covered 
include computational mathematics, computer applications, computer hardware,
computer software, control applications, control systems, information science,
and systems and control theory.  There are many tens of thousands of patent
records in that large and very useful database.  If you were to talk to the
authors of the over a million articles abstracted there, I believe you 
would find that in at least ten percent of the cases, the writer was willing
to publish _only_ because it was possible to file a patent application first.
 
>[Just as with music, copyright] monopoly is also sufficient to protect 
>software.  Therefore, we don't need to grant software patents, which have 
>few benefits and demonstrated harms.
 
Mr. Eggert misses that there are ideas and there are expressions, and 
copyright only protects expressions.  If someone sees something clever 
in my software and writes her own software to do that, copyright is unlikely
to be of any help.  But if that something clever happens to be patentable and
patented, then I will be able to protect it but only through patent.  So the
proposition that "copyright is sufficient to protect software" is false.
 
Carl Oppedahl
30 Rockefeller Plaza
New York, NY  10112-0228

Path: gmdzi!ieee.org!dorm.rutgers.edu!rutgers!news.cs.indiana.edu!spool.mu.edu!
munnari.oz.au!metro!cluster!0001811...@mcimail.com
From: 0001811...@mcimail.com (Carl Oppedahl)
Newsgroups: comp.patents
Subject: benefits of patents
Message-ID: <3566@cluster.cs.su.oz.au>
Date: 19 Nov 91 21:03:00 GMT
Sender: n...@cluster.cs.su.oz.au
Lines: 51
Approved: pate...@cs.su.oz.au

Commenters were challenged to name "one good specific example of a beneficial
software patent".  

In response to some replies, Paul Eggert (egg...@twinsun.com) writes:
>Jim Adcock wrote about two devices (one music, one measurement).  But
>they both apparently fall under the Diamond v Diehr principle allowing
>patents on hardware devices or processes that contain some software.  I was
>asking for software patents, so these examples don't apply.
 
Once again I am reminded that commenters, especially those to advocate highly
polarized positions, are not doing us a service unless they define their 
terms.  Would it be an undue burden to ask each commenter who says there
should be no software patents to state what his/her definition of a software
patent is?
 
I ask this because I see Mr. Eggert's comment.  He seems to be saying that
if a patent covers a hardware device or process that contains some software, 
then that patent is not a software patent.  It is convenient to say 
this because he can disallow Mr. Adcock's candidate beneficial software 
patents from discussion.  But one can't have it both ways.
If patent covering a hardware device or process that contains some software 
is not a software patent, then I take it he would not object to its having 
been issued.
 
That's fine, except that the last dozen or so patents secured by my law firm
for our clients which _we_ would call software patents fall squarely within
Mr. Eggert's apparently non-objectional category.  And I believe it would
turn out that all but a small handful (a few dozen, perhaps) of the patents
issued by the U.S. patent office would fall squarely within Mr. Eggert's 
apparently non-objectional category.
 
Next thing you know, we are not even in disagreement about whether there 
should be software patents.
 
>Many people also gave general arguments about software patents.
>But I was asking for specific examples, not general arguments or analogies.
>Since there are theoretical arguments both for and against software patents,
>specific examples carry considerable weight.  So far, we have seen many
>examples of harmful software patents, and not one good, specific example of a
>beneficial one.
 
Now that I have seen what happens when commenters here proffer patents as 
being possibly beneficial software patents, you can imagine why I hesitate
to introduce any of the patents I have prosecuted (obtained for clients).
I'm afraid all I would do is unnecessarily expose my clients to accusations
of having stolen it from elsewhere, or withheld the good parts, or otherwise
having done bad things.
 
Carl Oppedahl
30 Rockefeller Plaza
New York, NY  10112-0228

			        About USENET

USENET (Users’ Network) was a bulletin board shared among many computer
systems around the world. USENET was a logical network, sitting on top
of several physical networks, among them UUCP, BLICN, BERKNET, X.25, and
the ARPANET. Sites on USENET included many universities, private companies
and research organizations. See USENET Archives.

		       SCO Files Lawsuit Against IBM

March 7, 2003 - The SCO Group filed legal action against IBM in the State 
Court of Utah for trade secrets misappropriation, tortious interference, 
unfair competition and breach of contract. The complaint alleges that IBM 
made concentrated efforts to improperly destroy the economic value of 
UNIX, particularly UNIX on Intel, to benefit IBM's Linux services 
business. See SCO v IBM.

The materials and information included in this website may only be used
for purposes such as criticism, review, private study, scholarship, or
research.

Electronic mail:			       WorldWideWeb:
   tech-insider@outlook.com			  http://tech-insider.org/