🚀 go-pugleaf

In article <3988@well.UUCP> shibumi@well.UUCP (Kenton A. Hoover) writes:
>>You can patent anything that you can convince the patent office to
>>issue you a patent for.

In article <7335@ism780c.UUCP> tim@ism780c.UUCP (Tim Smith) writes:
>The patent office will issue for almost anything that hasnt yet been invented
>and that meets the 'what can be patented' standard.  The strength of your
>patent is almost always only determined through litigation.

That was my point.  The patent is no good unless the courts will
enforce it, and I suspect that the RSA patent will fall as soon as an
organization with sufficient legal resources decides to challenge it.
But don't try this at home if you don't want to spend time in court.


-- 
- Joe Buck  {uunet,ucbvax,sun,decwrl,<smart-site>}!epimass.epi.com!jbuck
	    Old internet mailers: jbuck%epimass.epi.com@uunet.uu.net

    The book "Advanced Legal Strategies for Software Protection" is a good
way to get an idea of how software patents work in practice.  It is definitely
possible to patent an algorithm that can be expressed in custom hardware.

    Patents are taken much more seriously today than a few years ago.  One
reason is the Kodak-Polaroid legislation, where Kodak was held to be 
infringing on Polaroid's instant color film patents.  The judgement was
that Kodak was 1) ordered to get out of the instant photography
business within thirty days 2) ordered to offer to buy back every Kodak
instant camera ever sold 3) ordered to pay damages to Polaroid in the 8-figure
range.  All these things were done.  This shook up a lot of businesses that
assumed that if they were found to be in infringement they would just pay
some damages and be done with it.  When you have to factor in the possibility
of being kicked out of a business and writing off a few hundred million in
now-useless manufacturing facilities, patent infringement looks very risky.
Being responsible for a debacle like that can result in a severe career
setback.

    More recently, jurisdiction over patent, trademark, and copyright cases 
has been centralized; all such cases now are heard by a Federal court
specializing in such matters, rather than by the regional district court.
Turnaround time has improved; the queue is separate from the general
docket of Federal civil cases.  The new court is considered pro-patent-holder.

    Being a free-lance inventor, I think that all this is just great.

					John Nacs!u

I believe the best method is to take a description to a notary (or perhaps
two young ones, in case one dies) and have your signature notarized on each
page of the description.  Leave space for the notary's seal at the bottom,
so that no essential details are obscured.  Have the notary keep a copy
(they usually do anyway), and specify how long you think the information
might be important.  It might be worth having the notary keep a copy
in a sealed envelope to counter claims of fiddling notarial files.

My Uncle Paul, a fairly prolific inventor (of teflon-coated bullet fame),
makes regular trips to the notary with sheafs of paper.

IBM uses an interesting system: you are required to keep notes of your work
in special IBM notebooks.  At the end of the day, you sign and date all
the pages you used that day.  These notebooks have been important evidence
at patent infringement trials.  If you use such a system, get bound books
of lined paper, like the green books the government issues.  Spiral and
loose-leaf books are much less persuasive, for obvious reasons.

In article <934@bakerst.UUCP> kathy@bakerst.UUCP (Kathy Vincent) writes:
-In article <1329@se-sd.sandiego.NCR.COM> dennis@se-sd.sandiego.NCR.COM writes:
->In article <2411@calmasd.GE.COM> jnp@calmasd.GE.COM (John Pantone) writes:
->>One, commonly used, method for "proving" the date of invention is to
->>document the invention and mail it to yourself (postmark).
->Gosh, I think I'll mail myself some envelopes (unsealed of course) to
->get them postmarked today.  Then, when something comes along that I
->want to be the inventor of, I'll write a description of it and put it
->into one of the envelopes and seal it.
-...  The envelope would
-have to have be sealed.   And the mail has to be registered, and they
-stamp the heck out of the envelope - including the seal, as proof that
-the contents were in the envelope on the date the envelope was sealed,
-registered, and mailed.

During wartime (and perhaps at other times), censors regularly extract
the contents of envelopes sealed like this, examine them, and reinsert
them, without leaving noticeable evidence of their tampering.  You could
use similar techniques (which I won't describe, but they're simple) to
fake time-stamped registry via postal means; just insert the document
you want to be thought of as "old", like Dennis suggested.  There are
tests that could possibly detect the relative age of the paper and its
contents, but it's unlikely anyone would go that far in a challenge.

If there's a lot of money involved, and you can't trust a guy's word,
why should you trust his evidence?

I also am looking for a public key program. I have been working on one
in turbo pascal and have had some problems with large prime numbers on
the system. Does anyone have something written or know where I can start
looking for the information? All the info here at school covers the
techinical aspects, but doesn't get down into the actual hard code.
Unfortunatly, I don't have the time to keep writing from scratch...
 
If need be, I can take a C program also...

-- 
...!lll-crg --> !csustan --\                         | Whatever I said doesn't
                            >->!polyslo!caus-dp!root | mean diddly as I forgot
  ...!ihnp4 --> !csun -----/                         | it even before finishing
             ...!dmsd ----/                          | typing it all out!!!

>I also am looking for a public key program. I have been working on one
>in turbo pascal and have had some problems with large prime numbers on
>the system. Does anyone have something written or know where I can start
>looking for the information?

I take it you want something that will run on a PC.  I have seen (and
have a copy of) a public domain public key system called PKCRYPT for
the PC.  I'm not sure about it though:  It uses the RSA scheme which
I seem to remember is under copyright (wasn't there a big discussion
about this?).  If there is some way around the copyright, then I have
a bunch of "large number" arithmetic routines that could be thrown
together for a public key system fairly quickly.  They are written
in C.

-- 
Steve Tate			UUCP: ..!{ihnp4,decvax}!duke!srt
				CSNET: srt@duke
				ARPA:  srt@cs.duke.edu
"There ain't nothin' in the world that a T-Bone Shuffle won't cure."

  Since several people have asked for PKS code, here are some
places to find ideas and/or code.  I have no knowledge if any
of the articles worthwhile; I haven't had an opportunity to
look closely at them.  These articles appeared in easily-obtained
journals or magazines, so happy hunting.

1. This gives some good suggestions for writing RSA
	"A Proposed Standard for RSA Cryptosystems"
	Philip Zimmermann
	IEEE Computer, Sept 86 pp21-34

2. (full?) RSA code in RatFOR
	"RSA: A Public Key Cryptography System"
	C. E. Burton
	Dr Dobbs Journal, March pp 16-43 (part 1) and June 84 pp 32-59

3. Article with some Pascal code (not a complete implementation)
	"Public-key Encryption Techniques"
	T. F. Elbert
	Journal of Pascal and Ada, Sept/Oct 83 pp14-20

4. Some silliness about using Basic for cryptography
	"Implementing Cryptographic Algorithms in Microcomputer Basic"
	Byte, Oct 84  pp126-?

Bob Devine

> [...public key encryption...prime numbers...]

Sounds like  you'r talking about RSA public key encryption. This technique
is patented by Rivest, Shamir, and Adelman (? did I get these names right?).

Therefore there is not now nor will there ever be (until the 17 or so years
is up, anyway) a public domain implementation of this.

If anybody knows anything about any of the non-patented public key systems,
(knapsack?), please post a followup.
-- 
-- Peter da Silva `-_-' ...!seismo!soma!uhnix1!sugar!peter
--                 'U`  <-- Public domain wolf.

In article <226@vianet.UUCP> devine@vianet.UUCP (Bob Devine) writes:
>
>  Since several people have asked for PKS code, here are some
>places to find ideas and/or code.  I have no knowledge if any
>of the articles worthwhile; I haven't had an opportunity to
>look closely at them. 
>
>1. This gives some good suggestions for writing RSA
>	"A Proposed Standard for RSA Cryptosystems"
>	Philip Zimmermann
>	IEEE Computer, Sept 86 pp21-34
>  ...
>Bob Devine

It is, maybe, worthwhile to compare the section ``Shortcut for RSA decryption''
(attributed to Charles Merritt, 1984, 1985) of this paper (there is also
errata in another issue of IEEE Computer) and the paragraph
``Fast deciphering algorithm'' of the following paper:

J.-J. Quisquater and C. Couvreur, ``Fast decipherment algorithm for RSA
public-key cryptosystem'', Electronics Letters, 14th Oct. 1982, Vol.18, 
No. 21, pp. 905-907.

Any way ISO (International Standard Organisation) is studying the use
of public-key systems for signature (only) under the working group
TC97/SC20/WG2. I know that because I am the editor of the (draft) 
document (accepted and presented with Louis Guillou -- CCETT, Rennes, France --
at CRYPTO'87, Santa Barbara, Calif., 16-20 Aug.).

 Jean-Jacques Quisquater,

SMAIL:
 Philips Research Laboratory Brussels
 Avenue Van Becelaere, 2; Box 8; B-1170 Brussels, Belgium.
UUCP: 
 World:   ...{philabs,uunet,munnari,mcvax}!prlb2!jjq
INTERNET: 
 na.quisquater@score.stanford.edu

In article <686@sugar.UUCP> peter@sugar.UUCP (Peter da Silva) writes:
>> [...public key encryption...prime numbers...]
>
>Sounds like  you'r talking about RSA public key encryption. This technique
>is patented by Rivest, Shamir, and Adelman (? did I get these names right?).
>
>Therefore there is not now nor will there ever be (until the 17 or so years
>is up, anyway) a public domain implementation of this.

There is public-domain code (Dr. Dobbs' Journal and other places);
whether it's legal to use is another matter.

There are other number-theoretic public-key algorithms out there.
Also, it's not clear to me that the RSA patent could withstand a
well-financed challenge in court, and I suspect it will be thrown out
eventually.  The way the patent laws were set up, you can't patent a
discovery or a mathematical truth.  Until recently algorithms were
considered unpatentable since they were seen as mathematical laws.
There's a lot more than that to real software.  But RSA is a
straightforward application of Euler's Theorem.  Also, R, S, and A
didn't come up with the core concept of public-key encryption -- a
"one way" transformation.  Their contribution was to recognize a new
way in which to use 200 (or more)-year-old number theory.  Unless the
license fees are kept very reasonable and low, the patent will be
challenged and ruled invalid in the near future.  

A patent gives the patent holder grounds to sue someone who infringes
on it.  The government will not enforce the patent.  It does not give
a guarantee that the patent holder will emerge victorious; frequently
the patent holder emerges without the patent (because it is ruled
invalid).  

>If anybody knows anything about any of the non-patented public key systems,
>(knapsack?), please post a followup.

The knapsack approach has been broken; fast ways of solving knapsack
problems are now known.


-- 
- Joe Buck    jbuck@epimass.epi.com
	      {uunet,ucbvax,sun,decwrl,<smart-site>}!epimass.epi.com!jbuck
	      Old arpa mailers: jbuck%epimass.epi.com@uunet.uu.net

I agree with jbuck@epimass's assessment of the situation.  As I
remember it, you can't patent something that can be done (in principle)
by a person using pencil and paper.

Several months ago I posted a request asking just what about RSA the
MIT patent covers.  I got no answers.  It seems to me that they can be
patenting any of a number of things:

    o their special purpose hardware to perform RSA

    o their method of selecting primes (again probably only
      as it's imbedded in ROM in some hardware).

    o their specific approach to choosing primes and using
      exponentiation (their specific algorithm)

    o the general notion of using C = M^E mod pq

The first two would hold up, because you can patent a system with
algorithms embedded in it, but then the algorithm itself is not
necessarily protected.  The last two seem to violate the pencil and
paper test.

Not wanting to spend the time or money to ask the patent office for a
copy of the patent, I hoped I could find out on the net.  Does anyone
know specifically what the Patent was granted for/what they claimed?

Michael L. Mauldin (Fuzzy)		Department of Computer Science
ARPA: Michael.Mauldin@NL.CS.CMU.EDU	Carnegie-Mellon University
Phone: (412) 268-3065			Pittsburgh, PA  15213-3890

In article <25@nl.cs.cmu.edu> mlm@nl.cs.cmu.edu (Michael Mauldin) writes:
>    o the general notion of using C = M^E mod pq

I believe that the patent covers the use of this equation for
a specific purpose.  Patents protect applications of phenomena,
not the phenomena.  The patent doesn't cover other uses of this
equation.

-andy

ps - The "pencil and paper" rule is bogus.  Copiers are patentable.
-- 
Andy Freeman
UUCP:  {arpa gateways, decwrl, sun, hplabs, rutgers}!sushi.stanford.edu!andy
ARPA:  andy@sushi.stanford.edu
(415) 329-1718/723-3088 home/cubicle

In article <25@nl.cs.cmu.edu> mlm@nl.cs.cmu.edu (Michael Mauldin) writes:
< I agree with jbuck@epimass's assessment of the situation.  As I
< remember it, you can't patent something that can be done (in principle)
< by a person using pencil and paper.

You can patent anything that you can convince the patent office to
issue you a patent for.

The book "Software Law - A Primer" ( or something like that ) has
some examples of patents that have been granted that are similar
to RSA.

For example, there was some manufacturing process that involved the
use of feedback ( if I remember correctly ).  Someone came up with
the idea of using a specific function to control the feedback.
It was a fairly simple function, too.  One that shows up in other
things.  Something like y=exp(-x).

Their patent was not an attempt to patent the exponential function.
Their patent was an attempt to patent the use of an exponential
function as a key step in making a specific type of item.

The patent was granted.  This did not mean that other people could
not use exponential functions.  It just meant that other people
could not use them in the same way when making the same type of
product as the patent holder.

I suspect that the RSA patent would be similar.  Anyone can use
the functions that RSA uses for things other than a public key
cryptography system.  It would only be the specific use that
would be covered by the patent.

Disclaimer: What do I know?  I am just a person who read a book
that covered this.  Consult a real lawyer before taking anything
I say seriously!
-- 
Tim Smith, Knowledgian		{sdcrdcf,uunet}!ism780c!tim
				tim@ism780c.isc.com
"Oh I wish I were Matthew Wiener, That is who I truly want to be,
 'Cause if I were Matthew Wiener, Tim Maroney would send flames to me"

In article <7335@ism780c.UUCP> tim@ism780c.UUCP (Tim Smith) writes:
>You can patent anything that you can convince the patent office to
>issue you a patent for.

The patent office will issue for almost anything that hasnt yet been invented
and that meets the 'what can be patented' standard.  The strength of your
patent is almost always only determined through litigation.


-- 
!   Kenton A. Hoover             {hoptoad,hplabs,lll-lcc,ptsfa}!well!shibumi ! 
!                                 SNAIL: 1748 Clement Street                 !
!   Prescriptive Technology              San Francisco, CA 94121             !
! "Evil will always triumph over good because good is dumb!"                 !

Andy Freeman writes:
>Michael Mauldin writes:
>>    o the general notion of using C = M^E mod pq

>I believe that the patent covers the use of this equation for
>a specific purpose.  Patents protect applications of phenomena,
>not the phenomena.  The patent doesn't cover other uses of this
>equation.

>ps - The "pencil and paper" rule is bogus.  Copiers are patentable.

  Specific copier technology is sometimes patentable.  The idea of
a copier is not.  The use of reflected light in a copier is not
patentable.  Etc.  You get the idea, I hope.

  Whether or not this rather ridiculous patent holds up in court, you
may be sure that lots of people (myself included) are already using

       o the general notion of using C = M^E mod pq

for our intellectual amusement, as well as occasional protection
of data, for signatures, or whatever.

-- 
  Robert Bickford                 {hplabs, ucbvax, lll-lcc, ptsfa}!well!rab
/-------------------------------------v-------------------------------------\
| Don't Blame Me: I Voted Libertarian | Ron Paul: Libertarian for President |
\-------------------------------------^-------------------------------------/

In article <25@nl.cs.cmu.edu> mlm@nl.cs.cmu.edu (Michael Mauldin) writes:
> 
> Several months ago I posted a request asking just what about RSA the
> MIT patent covers.  I got no answers.  

> Not wanting to spend the time or money to ask the patent office for a
> copy of the patent, I hoped I could find out on the net.  Does anyone
> know specifically what the Patent was granted for/what they claimed?
> 
> Michael L. Mauldin (Fuzzy)

Well, I am having cracker-jack Government patent lawyers take a look
at the RSA Patent.  The patent grants some rights to the U.S.
Government, so my answer may not be for everyone.  I have a copy
of the patent, but I can't really read it (you need a recursive
descent parser to wade through some of the clauses :-)).  They seem
to be patenting the "implementation" of an algorithm for a specific
purpose (cryptography).  The example posted earlier on the use
of exp() for feedback loops is probably very similar.
I'll let ya'll know what I find out.

Tim Tessin - Lawrence Livermore National Laboratory 
Phone: (415) 423-4560 / 422-8971
ARPA:  tjt@lll-tis.ARPA
UUCP:  {ihnp4,dual,sun}!lll-lcc!lll-tis!tjt

In article <1509@epimass.EPI.COM>, jbuck@epimass.EPI.COM (Joe Buck) writes:
> 
> That was my point.  The patent is no good unless the courts will
> enforce it, and I suspect that the RSA patent will fall as soon as an
> organization with sufficient legal resources decides to challenge it.
> But don't try this at home if you don't want to spend time in court.

	Unless, of course, IBM, AT&T, the NSA or someone with hordes of
lawyers decides to try to defend it. Hell, just anyone with enough money
who decides to make a deal with R&S&A. Or is it just that my perception
of reality is warped ??

-- 
If they think you're crude, go technical; if they think you're technical,
go crude. I'm a very technical boy. So I get as crude as possible. These
days, though, you have to be pretty technical before you can even aspire
to crudeness...			         -Johnny Mnemonic

>   Whether or not this rather ridiculous patent holds up in court...

Why is it ridiculous?  If one ignores the minutiae of US patent law for
the moment, RS&A clearly have produced an honest-to-God invention:  a
useful idea that is neither obvious nor in the domain of natural law.
(Nature may have supplied the properties of the integers, but using them
for encryption is strictly a human concept.)  They *should* be able to
patent it and have the patent hold up in court.  Now, it's not clear that
they *can* under current patent law, but that's another issue.  Is there
any reason for considering the idea "ridiculous" other than the
spoiled-brat desire to use their invention without paying royalties?
-- 
"There's a lot more to do in space   |  Henry Spencer @ U of Toronto Zoology
than sending people to Mars." --Bova | {allegra,ihnp4,decvax,utai}!utzoo!henry

Aside from the appropriateness of the RSA algorithm itself for patenting,
how about the manner in which it was distributed?  In a paper distributed
by MIT and in Communications of the ACM, the authors described RSA without
any mention of a patent or pending patent.

Now, I know zero about patent law, but shouldn't this place the algorithm
in the public domain?  Suppose, eg, that I hadn't read in BYTE or in this
newsgroup that it's patented, and built some large system based on RSA.
How can they claim I should have known it's theirs?

Anyone who knows more about patent law (ie, nearly everyone) care to comment?

 -- Mike Morton // P.O. Box 11378, Honolulu, HI  96878
    Local gurus assure me my addresses are:
      INTERNET: cs313s19@uhccux.uhcc.hawaii.edu
      UUCP:     {ihnp4,uunet,dcdwest,ucbvax}!sdcsvax!nosc!uhccux!cs313s19
      BITNET:   cs313s19%uhccux.uhcc.hawaii.edu@rutgers.edu

In article <17179@glacier.STANFORD.EDU> jbn@glacier.UUCP (John B. Nagle) writes:
>
>    Patents are taken much more seriously today than a few years ago.  One
>reason is the Kodak-Polaroid legislation
>
>    More recently, jurisdiction over patent, trademark, and copyright cases 
>has been centralized ...  The new court is considered pro-patent-holder.
>
>    Being a free-lance inventor, I think that all this is just great.
>
>					John Nagle


Seems to me the trick would be do you as an individual, free-lance
inventor have the money and the time to go up against the battery of
lawyers and words and funds that a corporation intent on "breaking"
(not being a lawyer, I don't claim to be using the correct word here)
your patent would bring to bear if it were really serious?  Polaroid
isn't exactly the guy next door with a lab in his garage and a computer
in his attic and $150 in his savings account ...



Kathy Vincent ------> Home: {ihnp4|mtune|codas|ptsfa}!bakerst!kathy
              ------> AT&T: {ihnp4|mtune|burl}!wrcola!kathy

In article <849@uhccux.UUCP>, cs313s19@uhccux.UUCP (Cs313s19) writes:
> In a paper distributed
> by MIT and in Communications of the ACM, the authors described RSA without
> any mention of a patent or pending patent.
> 
> Now, I know zero about patent law, but shouldn't this place the algorithm
> in the public domain?  Suppose, eg, that I hadn't read in BYTE or in this
> newsgroup that it's patented, and built some large system based on RSA.
> How can they claim I should have known it's theirs?

Repeat after me:

     Algorithms aren't patentable; HARDWARE IMPLEMENTATIONS are.

     Ignorance of a prior patent is no excuse.

Specific text (like a paper) or code is copyrightable, but that's
not what we're discussing here.  Algorithms are either known or
not known, much like laws of science.  Algorithms are neither
patentable nor copyrightable.  The only way to retain control of
an algorithm is to keep it a secret; and even then you lose if
someone discovers it independently.

Patent rights are enforceable as long as the patent is valid.
This means that if you want to market something, you're required
to perform a patent search first.  A patent search is an
examination of all the existing patents related to your design.
The onus is on you to look for existing patents, rather than on
the patent holders to tell you in advance that they have a patent
(which is obviously unworkable).  Failure to perform a patent
search is foolhardiness of the highest order, and you'd DESERVE to
get sued.
-- 
{hplabs,amdahl,ames}!bnrmtv!perkins         --Henry Perkins

It is better never to have been born.  But who among us has such luck?
One in a million, perhaps.

Mike Morton noted that RSA was described in a published paper, and wonders
whether it is thus in the public domain.

(I'm not a lawyer, so don't trust this.  My name is, however, on a patent.)

In the United States, you have one year after "disclosure" to file
for a patent.  You must, however, be able to show the actual date
of the invention.  In general, a bound notebook with a dated page
is sufficient.  Having a couple of witnesses sign the page would
be a good idea.  (This isn't needed for the patent, but for the
all-important defense against claims that someone else invented it
first.)

In other countries, you must file for the patent before disclosure.

Disclosure generally means public knowledge of the patentable idea.
You can tell others your idea if they have signed a "non-disclosure
agreement" without comprimising your situation.

Hope this helps ease the confusion.

Martin Minow
decvax!minow

The above does not represent the position of Digital Equipment Corporation.

In article <8640@utzoo.UUCP> henry@utzoo.UUCP (Henry Spencer) writes:
+ >   Whether or not this rather ridiculous patent holds up in court...
+ 
+ Why is it ridiculous?  If one ignores the minutiae of US patent law for
+ the moment, RS&A clearly have produced an honest-to-God invention:  a
+ useful idea that is neither obvious nor in the domain of natural law.
+ (Nature may have supplied the properties of the integers, but using them
+ for encryption is strictly a human concept.)  They *should* be able to
+ patent it and have the patent hold up in court.  Now, it's not clear that
+ they *can* under current patent law, but that's another issue.  Is there
+ any reason for considering the idea "ridiculous" other than the
+ spoiled-brat desire to use their invention without paying royalties?

  All rudeness aside (the temptation is great.....), the point of the
ongoing discussion here has been precisely the fact that the patent would
appear to cover not just a single defineable invention but in fact an
entire class of possible inventions, discovered or otherwise.  It is on
this basis that I described the patent as "ridiculous".  Waiving (for
the sake of discussion) the question of whether anyone can in fact patent
an algorithm, and assuming (again, for the sake of discussion) that this
patent did in fact cover a specific algorithm, then I for one would feel
honor bound to respect it.  (Not, by the way, because of any silly law,
but because of my libertarian belief in the primacy of individual rights,
which include the right(s) to intellectual property.  But that's another
subject.)  However, this patent would appear to attempt to include any
and all such algorithms, discovered or not, and for that reason I find it
not deserving of the respect one would normally give toilet paper.
  By the way, I have noticed you several times use the phrase "spoiled brat"
when arguing on the net.  (Do you have children?)  This is known as an
'ad hominem' attack.  Don't bother; if we want to stoop to that sort of
thing we can all go into conventional politics.

-- 
  Robert Bickford                 {hplabs, ucbvax, lll-lcc, ptsfa}!well!rab
/-------------------------------------v-------------------------------------\
| Don't Blame Me: I Voted Libertarian | Ron Paul: Libertarian for President |
\-------------------------------------^-------------------------------------/

In article <4023@well.UUCP> rab@well.UUCP (Bob Bickford) writes:
>  All rudeness aside (the temptation is great.....), the point of the
>ongoing discussion here has been precisely the fact that the patent would
>appear to cover not just a single defineable invention but in fact an
>entire class of possible inventions, discovered or otherwise.

Since no one has bothered to actually look up the RSA patent to find
out what is covered, various people on the net have assumed that what
it covers supports their position.  One possibility is that it is for
the use of a certain integer function to encrypt digital information.
Another possibility is that it is for one way to encrypt digital information
using integer functions with certain properties.  A third possibility is
that it covers encryption of digital information using integer functions
with certain properties.  Bob's objection suggests that he thinks that the
RSA patent is one of the latter two.

>		However, this patent would appear to attempt to include any
>and all such algorithms, discovered or not, and for that reason I find it
>not deserving of the respect one would normally give toilet paper.

True, the broad patents described above don't list the integer functions,
they merely describe them.  I fail to see why this is objectionable.  For
example, let's say that the US patent office was open when the pulley was
invented.  Bob's like-minded ancestor would have objected to a pulley patent
unless it covered only pulleys made out of materials the inventor mentioned.
Many of the rest of us would settle for a description of the properties that
the the pulley material must have.

So Bob, what do you think the RSA patent covers?  What do you think it
would be legitimate for it to cover?

-andy
-- 
Andy Freeman
UUCP:  {arpa gateways, decwrl, sun, hplabs, rutgers}!sushi.stanford.edu!andy
ARPA:  andy@sushi.stanford.edu
(415) 329-1718/723-3088 home/cubicle

Please move this non-technical discussion of patent law away from sci.crypt.

Thank you.

	Kurt F. Sauer
	Tulsa, OK

In article <104@lll-risky.arpa> tjt@lll-risky.arpa.arpa (Tim Tessin) writes:
>
>They seem
>to be patenting the "implementation" of an algorithm for a specific
>purpose (cryptography).  The example posted earlier on the use
>of exp() for feedback loops is probably very similar.
>I'll let ya'll know what I find out.
>
>Tim Tessin - Lawrence Livermore National Laboratory 

I mentioned before that I have a copy of a public key cryptosystem
using RSA that has been released in the public domain.  So how is
this effected?  One possible point is that the person who wrote it
lives in Canada.  Does this mean he doesn't have to abide by US
Patents?  How about importing such a thing into the US?  Maybe the
guy is just not aware that it is patented and should be notified?

-- 
Steve Tate			UUCP: ..!{ihnp4,decvax}!duke!srt
				CSNET: srt@duke
				ARPA:  srt@cs.duke.edu
"There ain't nothin' in the world that a T-Bone Shuffle won't cure."

> Aside from the appropriateness of the RSA algorithm itself for patenting,
> how about the manner in which it was distributed?  In a paper distributed
> by MIT and in Communications of the ACM, the authors described RSA without
> any mention of a patent or pending patent.
> 
> Now, I know zero about patent law, but shouldn't this place the algorithm
> in the public domain? ...

No.  Whether the publication mentions intent to patent is entirely irrelevant.
(A related issue:  it is permissible for something like 18 months to elapse
between publication and filing for a patent.)  It would have been nice for
them to have mentioned it, and perhaps wise, but it is not compulsory, as
I understand it.

Consider:  Popular Science does an article about RSA, based on interviews
etc.  They decide not to mention the pending patent.  Does this void the
patent?  Clearly not.  Does saying "well, I read about it in Popular Science,
and they didn't mention a patent, so I figured it must be public domain"
constitute a valid defense in court?  Obviously not.  It's entirely possible
that the original CACM paper, as *submitted*, mentioned the patent but the
editors deleted it for some silly policy reason.  Should this void the patent?
Nonsense.

> ... Suppose, eg, that I hadn't read in BYTE or in this
> newsgroup that it's patented, and built some large system based on RSA.
> How can they claim I should have known it's theirs?

They don't; what they claim is that you are using their invention, which
is all that matters.  Whether you knew about it or not DOES NOT MATTER
(although they might be more polite if your infringement did appear to be
ignorance rather than malice).  Even inventing it yourself, entirely
independently and with no knowledge of their work, is not a defense.
-- 
"There's a lot more to do in space   |  Henry Spencer @ U of Toronto Zoology
than sending people to Mars." --Bova | {allegra,ihnp4,decvax,utai}!utzoo!henry

I am not a lawyer, so take this with the "standard grain of salt":

One, commonly used, method for "proving" the date of invention is to
document the invention and mail it to yourself (postmark).  Leave the
envelope un-opened.
-- 
These opinions are solely mine and in no way reflect those of my employer.  
John M. Pantone @ GE/Calma R&D, Data Management Group, San Diego
...{ucbvax|decvax}!sdcsvax!calmasd!jnp          jnp@calmasd.GE.COM

In article <615@rocky.STANFORD.EDU>, andy@rocky.STANFORD.EDU (Andy Freeman) writes:
> Since no one has bothered to actually look up the RSA patent to find
> out what is covered, various people on the net have assumed that what
> it covers supports their position.

[ I've added misc.legal to the distribution -- Bob]

  Well, since I couldn't get my hands on a copy of the patent quick
enough to contribute to this discussion, I talked to someone who has
read it.

  Philip Zimmerman wrote the IEEE Computer article last year about RSA.
Since he lives within a local-call distance, I asked him about his take
on the patent.

His comments:

1. The patent does not cover an implementation (hardware nor software)
   of the RSA algorithm.
2. Many side issues are also listed in the patent.  (eg. using 3 primes,
   bootstrapping into another algorithm)
3. It seemed to him that patenting the algorithm is like patenting
   a mathematical idea.  [I agree.  Remember folks, the patent test of
   "novel and unobvious" is easily passed when dealing with judges who
   were Poli-Sci majors. :-)]
So, he concluded that a patent challenge would likely win.


  Now other points are with the patent process itself:  while it is not
totally clear, it looks like software is NOT patentable (but can be
copyrighted easily enough) because the idea, per se, is in the underlying
algorithm.  Be careful in distinquishing between an algorithm and a
process; the courts did but it seems rather artificial to me.  If you
read patents, many times they will show the idea manifested as a hardware
device.  I think even Ritchie's setuid idea used a hardware diagram.
I conclude that a hardware version of RSA is definitely patentable but
that pushing it to more than that is questionable.

  Now if RSA is patentable (patents are easily obtained) and stands up
to challenges (RSA probably wouldn't), it still has to guard against
infringement.  And, oh lordy, there's been a ton o' infringement.  It
seems that everyone and their uncle has come up with their own implementation
of RSA.  RSA Inc. has not attempted to stop infringment (I haven't heard of
any).  So, it may fall like Ritchie's setuid patent because of non-enforcement.

yet another non-lawyer,
Bob Devine

	You know what occurs to me as a really ha-ha thought !?!  Remember how
much sh** people had to go through befroe exporting anything with (breakable)
DES in it ? Well, the same country goes and lets people PUBLISH RSA...

	Isn't that a scream ? I suppose DES is still restricted, too. It's
a good thing that the spooks don't live in the REAL world - they'd really get
in our way....

--mjr();
-- 
If they think you're crude, go technical; if they think you're technical,
go crude. I'm a very technical boy. So I get as crude as possible. These
days, though, you have to be pretty technical before you can even aspire
to crudeness...			         -Johnny Mnemonic

In article <1398@osiris.UUCP> mjr@osiris.UUCP (Marcus J. Ranum) writes:
>	You know what occurs to me as a really ha-ha thought !?!  Remember how
>much sh** people had to go through befroe exporting anything with (breakable)
>DES in it ? Well, the same country goes and lets people PUBLISH RSA...
>
>	Isn't that a scream ? I suppose DES is still restricted, too. It's
>a good thing that the spooks don't live in the REAL world - they'd really get
>in our way....

The boys with guns (of/for/by the people) did try to supress RSA, I believe
that after the Scientific American article about it came out, giving
people all over the world everything they needed to implement the algorithm,
the State dept. tried to prevent reprints of RSA's article from getting
out of the country.  It took the better part of a year before I got my
copy in Canada.

DES is not classified, only the open secret (or widespread belief) that
the NSA has a backdoor to it is classified.  Implementations of DES
are subject to export controls, however, meaning that if you want to
market an implementation internationally, do it from a foreign country.
(The balance of trade, of course, does not affect national security,
so this is the only patriotic thing to do.)

		David Palmer
		palmer@tybalt.caltech.edu
		...rutgers!cit-vax!tybalt.caltech.edu!palmer
	The opinions expressed are those of an 8000 year old Atlantuan
	priestess named Mrla, and not necessarily those of her channel.

In article <2411@calmasd.GE.COM> jnp@calmasd.GE.COM (John Pantone) writes:
>
>One, commonly used, method for "proving" the date of invention is to
>document the invention and mail it to yourself (postmark).  Leave the
>envelope un-opened.
> 
Gosh, I think I'll mail myself some envelopes (unsealed of course) to
get them postmarked today.  Then, when something comes along that I
want to be the inventor of, I'll write a description of it and put it
into one of the envelopes and seal it.  Incontrovertible evidence that
the invention was mine!  :-) :-) :-)

In article <1329@se-sd.sandiego.NCR.COM> dennis@se-sd.sandiego.NCR.COM writes:
>In article <2411@calmasd.GE.COM> jnp@calmasd.GE.COM (John Pantone) writes:
>>One, commonly used, method for "proving" the date of invention is to
>>document the invention and mail it to yourself (postmark).  Leave the
>>envelope un-opened.
>> 
>Gosh, I think I'll mail myself some envelopes (unsealed of course) to
>get them postmarked today.  Then, when something comes along that I
>want to be the inventor of, I'll write a description of it and put it
>into one of the envelopes and seal it.  Incontrovertible evidence that
>the invention was mine!  :-) :-) :-)

While "common", this method is not "incontrovertible" proof, and the
courts know it.  It's as true with copyright law as with patent law.
Because copyright law requires registration with two copies when you
publish <insert legal definition of public distribution>, the only
time this trick is needed for copyrights is when an author delays
publication for some reason (I've got a whole mess of brass quartets
and quintets in the hopper, to be published "someday").

Here's a trick:  put a one-cent stamp across each (sealed) seam of
the envelope, wait in line at your local PO, and have them hand-cancel
each stamp.
--

As I think about this, though, the technical ways to break into the
envelope exist and, I assume, one could reclose and reglue the stamps
without it showing.  What are the options?  Have a couple of people
review copies of your design papers and sign them as witnesses?  Get
the signatures notarized so there's a permanent record?

...Kris
-- 
Kristopher Stephens, | (408-746-6047) |          {whatever}!amdahl!krs
Amdahl Corporation   |                |    -or-  krs@amdahl.amdahl.com
     [The opinions expressed above are mine, solely, and do not    ]
     [necessarily reflect the opinions or policies of Amdahl Corp. ]

In article <4081@cit-vax.Caltech.Edu> palmer@tybalt.caltech.edu.UUCP (David Palmer) writes:
>DES is not classified, only the open secret (or widespread belief) that
>the NSA has a backdoor to it is classified.

Excuse me, let's please try to be more precise in our terminology.
Classified information (in the U.S.) is that which has been specifically
flagged as CONFIDENTIAL, SECRET, or TOP SECRET, depending on the
probable degree of damage to the national interests if the information
were to be divulged to the wrong parties.  Rumors by their nature
are not classified, although they MAY be about material that is
classified.  Unless you know that to be the case (I personally don't
know one way or the other), the rumor you refer to is just a rumor.

In article <1329@se-sd.sandiego.NCR.COM> dennis@se-sd.sandiego.NCR.COM writes:
>In article <2411@calmasd.GE.COM> jnp@calmasd.GE.COM (John Pantone) writes:
>>
>>One, commonly used, method for "proving" the date of invention is to
>>document the invention and mail it to yourself (postmark).  Leave the
>>envelope un-opened.
>> 
>Gosh, I think I'll mail myself some envelopes (unsealed of course) to
>get them postmarked today.  Then, when something comes along that I
>want to be the inventor of, I'll write a description of it and put it
>into one of the envelopes and seal it.  Incontrovertible evidence that
>the invention was mine!  :-) :-) :-)


I don't think so, guys.  As I remember from having done this years ago
in an effort to protect some songs I was writing ...  The envelope would
have to have be sealed.   And the mail has to be registered, and they
stamp the heck out of the envelope - including the seal, as proof that
the contents were in the envelope on the date the envelope was sealed,
registered, and mailed.  Also, I was curious and asked a lawyer friend
just how valid that was, and the whole thing is very iffy - might not,
in fact, stand up in court.  

Oh.  One small consolation.  You don't actually have to MAIL the
thing, if you decide to go this route anyway.  Just go to the PO,
get your enveloped stomped, pay your money - and walk back out
with your envelopes.  :-)  Then take them to the bank and put them
in your safety deposit box until time to open them in court.

You DO have a safety deposit box for all these things, right? :-)


Kathy Vincent ------> Home: {ihnp4|mtune|codas|ptsfa}!bakerst!kathy
              ------> AT&T: {ihnp4|mtune|burl}!wrcola!kathy

In article <849@uhccux.UUCP> cs313s19@uhccux.UUCP (Cs313s19) writes:
>Aside from the appropriateness of the RSA algorithm itself for patenting,
>   ...
>Now, I know zero about patent law, but shouldn't this place the algorithm
>in the public domain?  Suppose, eg, that I hadn't read in BYTE or in this
>newsgroup that it's patented, and built some large system based on RSA.
>How can they claim I should have known it's theirs?
 
       It is reasonable to believe that the "claim" would have
       to be through the court.

       If the result of your work, is close enough to what has
       been patented, to prove that it "could have" been a "copy",
       even though you "claim" you didn't "copy", the law would
       turn in favor of the patent holder.  (please remember, I
       am not discussing the "fairness" of the deal, just how the
       patent rights can be interpreted).

       An example (although admittedly a crude one) may help clarify
       the point:

           If you build a structure on a land that you *THINK* is yours, 
           in all honesty, and another person happens to have the
           title, legal papers, etc. to that land, then who do
           you think *OWNS* the structure that you innocently 
           built on somebody else's land?

           If your answer is the "legal owner" of the land, you
           are right!  The law is very clear about that.  The only 
           exception is homesteading, in which case you could have
           owned the land and the building when certain conditions are 
           met.

       In dealing with the *LAW*, ignorance is **NOT** a bliss.
    
>
>Anyone who knows more about patent law (ie, nearly everyone) care to comment?
>
> -- Mike Morton // P.O. Box 11378, Honolulu, HI  96878
>    Local gurus assure me my addresses are:
>      INTERNET: cs313s19@uhccux.uhcc.hawaii.edu
>      UUCP:     {ihnp4,uunet,dcdwest,ucbvax}!sdcsvax!nosc!uhccux!cs313s19
>      BITNET:   cs313s19%uhccux.uhcc.hawaii.edu@rutgers.edu

> One, commonly used, method for "proving" the date of invention is to
> document the invention and mail it to yourself (postmark).  Leave the
> envelope un-opened.

If one looks into the matter, one quickly finds that this is valueless.
The preferred technique is to have someone else read and *understand*
the material and sign and date each page indicating this.
-- 
"There's a lot more to do in space   |  Henry Spencer @ U of Toronto Zoology
than sending people to Mars." --Bova | {allegra,ihnp4,decvax,utai}!utzoo!henry

In article <2411@calmasd.GE.COM>, John Pantone (jnp@calmasd.GE.COM) writes:

> I am not a lawyer, so take this with the "standard grain of salt":
> 
> One, commonly used, method for "proving" the date of invention is to
> document the invention and mail it to yourself (postmark).  Leave the
> envelope un-opened.

I am not a lawyer either.  However, I have been told by lawyers whose
specialty is patent law that the "letter" method is not acceptable.

Many years ago, when Varian (I think. . .) was in the process of
developing the klystron, they kept engineering notebooks in which
each page was two (or three) pages separated by carbons.  At the
end of each day, each page that had been written in was witnessed
and then one copy taken to the company's lawyer, who dated them
independently and stored them in a safe deposit box.  In later
legal battles, Varian prevailed.

For proper protection of inventions, it is my understanding (again,
from lawyers whose specialty is patent law) that a carefully-kept,
properly witnessed engineering notebook is sufficient to support
your claims.

					Steve Rice

-----------------------------------------------------------------------------
new: stever@videovax.tv.Tek.com
old: {decvax | hplabs | ihnp4 | uw-beaver | cae780}!tektronix!videovax!stever

I would think that several witnesses signing in the
presence of a notary public would be the best way
to do something like this.

James Moore

In article <4862@pbhya.UUCP> ba@pbhya.UUCP (Behzad Alavi) writes:
>In article <849@uhccux.UUCP> cs313s19@uhccux.UUCP (Cs313s19) writes:
> ...
>       An example (although admittedly a crude one) may help clarify
>       the point:
>
>           If you build a structure on a land that you *THINK* is yours, 
>           in all honesty, and another person happens to have the
>           title, legal papers, etc. to that land, then who do
>           you think *OWNS* the structure that you innocently 
>           built on somebody else's land?
>
>           If your answer is the "legal owner" of the land, you
>           are right!  The law is very clear about that.  The only 
>           exception is homesteading, in which case you could have
>           owned the land and the building when certain conditions are 
>           met.

There is the concept of "flagrant and adverse possession" in which you can
claim the land after 20 years of being on it in a flagrant manner without
being asked to leave by the owner.

I'm certainly NOT an attorney, but sat in on one day of Law School and
enjoy this opportunity (and that of the Bork hearings) to preach what
I can not practice. :-)

I might even be wrong!
-- 
Rob Lake
{decvax,ihnp4!cbosgd}!mandrill!nitrex!rbl

In article <1329@se-sd.sandiego.NCR.COM>, dennis@se-sd.sandiego.NCR.COM (Dennis Foster x5739) writes:
]In article <2411@calmasd.GE.COM> jnp@calmasd.GE.COM (John Pantone) writes:
]>
]>One, commonly used, method for "proving" the date of invention is to
]>document the invention and mail it to yourself (postmark).  Leave the
]>envelope un-opened.
]> 
]Gosh, I think I'll mail myself some envelopes (unsealed of course) to
]get them postmarked today.  Then, when something comes along that I
]want to be the inventor of, I'll write a description of it and put it
]into one of the envelopes and seal it.  Incontrovertible evidence that
]the invention was mine!  :-) :-) :-)

I think that the idea is to write the address on the *back* of the envelope
and put the stamp so that it covers the seal.
-- 
	doug foxvog	...allegra!lcuxlj!dougf	[Please use lcuxlj not lcuxlm]
			If only Bell Labs agreed with my opinion...
NSA: names of CIA agents in NRO working on TEMPEST encrypted above.  Drug
	dealing terrorists assassinated for planned hijacking.

> ]In article <2411@calmasd.GE.COM> jnp@calmasd.GE.COM (John Pantone) writes:
> ]>
> ]>One, commonly used, method for "proving" the date of invention is to
> ]>document the invention and mail it to yourself (postmark).  Leave the
> ]>envelope un-opened.
> ]> 

I had heard that this was the way to COPYRIGHT an item.  For some reason,
the date of first appearance seems to be important.

-- 

David B. Lewis    {ihnp4,allegra,ulysses}!homxc!lewisd
201-615-5306 EDT