UK Software Patents (was: File Encryption programs...)
Owen Lewis
da...@djwhome.demon.co.uk "David Woolley" writes:
>DJW>implementations. In the US, software is patentable, so the RSA
>DJW>patent applies to software implementations, there, but not here.
>
>>If I may say so, this advice is 100% wrong and does a general disservice.
>
>If this is true, it is a major change in UK law as far as the
>software industry is concerned, which has gone unpublicised.
>I think this would be better known, if it were the case.
It is known to some, particularly patent attorneys, patent offices and
patentees. It would seem that, had someone thought to do so, virtual reality
could have been patented.
>In the context of this demon.security, it almost certainly means that
>PGP is illegal in the UK,
Why? RSADSI have not registered their patent on the RSA algorithm outside
of the US. PGP is perfectly legal in the UK in specific regard to the RSA
patent.
> I can't see a patent not being taken on it if it
>were legally possible to do so.
That's as maybe. Fact is it wasn't.
>
>Software patents have major implications on software developers,
>as there is a tendency for people to patent common techniques,
>and for courts to accept these patents, even though the
>techniques are obvious and well known (there is apparently a
>patent on using XOR to make temporary changes to bit mapped
>displays).
A good patent cannot be obtained on a technique that has previously existed.
If, subsequent to the grant of a patent, 'prior art' can be demonstrated by
another party, then the patent falls.
> The implication is that software companies will have
>to spend a lot of money on tracking down the owners of patents
Why? A search at the Patent office is sufficient.
>and negotiating royalties, even though they have independently
>designed the algorithm which is subject to the patent.
If they got there first, well and good. If someone else got there first *and*
patented the idea then royalties are due.
>They will also have to spend a lot of money on registering
>patents,
If the commercial potential of an idea is sufficient, patenting it is a
sensible move.
> ..or continue the current practice of confidentiality
>agreements.
If the work is unpatentable or of so little commercial value as not to
warrant the expense of patenting, a confidentiality agreement is better than
nothing.
>>My patent attorney has successfully filed, on my behalf, a patent application
>>in the UK. This application is in respect of means for the post event
>>analysis of spectral data.
>If the essence of your invention is the software component (i.e.
>any hardware components could be replaced without violating the
>patent), I would suggest you find out exactly what piece of UK
>legislation, or case law, now makes it possible to patent software. If
>it is not too late, you may need to withdraw the patent, so that you can
>control the software by non-disclosure agreements.
Hmmm....
As I understand it. a patent is granted on an original idea. It is an
essential stipulation that a working example of the idea has been
completed. In addition to a description of the concept, a complete, detailed
description of a working example is required as a part of the patent
application.
It matters not at all that the implementation of the idea is a stream of code
to be read and acted on by a computer (not invented here). Nor does it matter
that the code is in fair part drawn (in properly licenced form) from other
software applications. What matters is that bits and pieces that had either
already existed or else have been purpose made are put together to give
expression to an entirely new idea or way of doing things. A simple example
of this would be as follows:
- A heavy load consisting of may disparate parts can be conveniently
moved over a distance by placing it in a container and then dragging or
pushing the container. This was once an original concept.
- Mounting the load container on axles and wheels is a major
improvement as heavier loads can be transported over longer distances, faster
and for less effort. However, the original concept of containing all parts of
the load in a convenient container remains incorporated. In modern times, the
inventor of the wheel might have patented it as the problem of the unwanted
limiting friction is largely eliminated while the essential static friction is
retained.
- Removing the contact of wheels and ground by containing them
caterpillar tracks enables much greater loads to be carried over soft ground
(because of the greatly increased loadbearing area). Also broken surfaces and
gradients can be negotiated that will defeat and conventionally wheeled
vehicle. Tracks rely on wheeled running gear but constitute an original
development in which the wheels are contained within an endless roadway that
is flexible in only one dimension. The load carrier still remains unchanged
in its original concept. Wheels are still wheels but, for certain conditions,
the invention of a very large wheel of flexible shape gives otherwise
unobtainable improvement in performance over land (alright, who said
"hovercraft" :-))
>>Ascom certainly do hold a valid patent that covers software implementations
>>of IDEA. They expect and require proper payment for the commercial use
>>of a product that has cost them a great deal to develop and to bring to
>>market.
>>If any one doubts this, why not ask them?
>
>If true, this would definitely take PGP23 and PGP26ui outside of
>the scope of the Gnu licence in the UK
What has "copyleft" to do with the existence of a valid patent on IDEA?
>(According to the PGP documentation, IDEA was actually developed
>by ETH, an academic body, although there may well have been tied
>funding.
I think you may mean that a contractual obligation existed between two or more
parties. Where and by whom the original work was done is, of course,
a matter of importance. However, it is largely irrelevant to the proper
ownership of a good patent.
> Also according to that documentation, ASCOM have waived
>royalties on the use of IDEA in PGP, for both commercial and
>non-commercial use.)
For non-commercial use, yes. For commercial use, no.
My copies of PGP 1.0, 2.0, 2.1, 2.2, 2.3 and 2.3a documentation do not read
as you claim. As I said, if you find this difficult to accept, why not ask
ASCOM? They won't bite :-).
Apologies to demon.security for continuance of an off group dialogue. Followups to alt.security.pgp.
--
-= Owen Lewis =-
@
Tele/fax +44-(0)794-301731 ELOKA Consultancy & Project Management
o...@eloka.demon.co.uk
pgp 2.3a public key on request
David Woolley
if it is an FAQ there), and alt.security.pgp, because of the serious
implications for PGP in Europe, Gnu also considered, as Gnu licence is
incompatible with patents. ]
Standard disclaimer: I am not legally qualified.
DJW>implementations. In the US, software is patentable, so the RSA
DJW>patent applies to software implementations, there, but not here.
>If I may say so, this advice is 100% wrong and does a general disservice.
If this is true, it is a major change in UK law as far as the
software industry is concerned, which has gone unpublicised.
I think this would be better known, if it were the case.
In the context of this demon.security, it almost certainly means that
PGP is illegal in the UK, as it seems unlikely that the RSA
algorithm would be licensed in the UK in violation of US export
restrictions. I can't see a patent not being taken on it if it
were legally possible to do so.
Software patents have major implications on software developers,
as there is a tendency for people to patent common techniques,
and for courts to accept these patents, even though the
techniques are obvious and well known (there is apparently a
patent on using XOR to make temporary changes to bit mapped
displays). The implication is that software companies will have
to spend a lot of money on tracking down the owners of patents
and negotiating royalties, even though they have independently
designed the algorithm which is subject to the patent.
They will also have to spend a lot of money on registering
patents, or continue the current practice of confidentiality
agreements.
>My patent attorney has successfully filed, on my behalf, a patent application
>in the UK. This application is in respect of means for the post event
If the essence of your invention is the software component (i.e.
any hardware components could be replaced without violating the
patent), I would suggest you find out exactly what piece of UK
legislation, or case law, now makes it possible to patent software. If
it is not too late, you may need to withdraw the patent, so that you can
control the software by non-disclosure agreements.
[ Full quote, and more comments ]
In article <787607...@eloka.demon.co.uk>
o...@eloka.demon.co.uk (Owen Lewis) writes:
>In article <787550...@djwhome.demon.co.uk>
> da...@djwhome.demon.co.uk "David Woolley" writes:
>>
>>It's [PGP] not covered by either in the UK, because you can't patent
>>software here. The IDEA patent applies to hardware
>>implementations. In the US, software is patentable, so the RSA
>>patent applies to software implementations, there, but not here.
>>It probably applies to hardware implementations here.
>
>If I may say so, this advice is 100% wrong and does a general disservice.
>
>My patent attorney has successfully filed, on my behalf, a patent application
>in the UK. This application is in respect of means for the post event
>processing of spectral data. This application is based on a successful software
>implementation.
>
>Furthermore, on the grant of this UK patent, the patent will be valid for
>registration throughout the EU, in the US and elsewhere, should I choose to
>so register it. For the meantime, it is internationally protected for 12 months.
>
>Ascom certainly do hold a valid patent that covers software implementations
>of IDEA. They expect and require proper payment for the commercial use
>of a product that has cost them a great deal to develop and to bring to market.
>If any one doubts this, why not ask them?
If true, this would definitely take PGP23 and PGP26ui outside of
the scope of the Gnu licence in the UK, and would have done so in the
US if it weren't for the enforceable patent on RSA already doing
so.
(According to the PGP documentation, IDEA was actually developed
by ETH, an academic body, although there may well have been tied
funding. Also according to that documentation, ASCOM have waived
royalties on the use of IDEA in PGP, for both commercial and
non-commercial use.)
--
David Woolley, London, England da...@djwhome.demon.co.uk
Demon supplies me with IP/SMTP/NNTP. *.demon hosts are independently managed.
Iolo Davidson
> > Ascom certainly do hold a valid patent that covers software
> > implementations of IDEA. They expect and require proper
> > payment for the commercial use of a product that has cost
> > them a great deal to develop and to bring to market. If any
> > one doubts this, why not ask them?
This has been thrashed through several times in alt.security.pgp.
To the best of my memory, the concensus goes something like this:
If you ask Ascom, they will say they have a patent. Switzerland
apparently allows software patents. Germany and Great Britain
*specifically* disallow patents on software. Software patents
are therefore not enforceable in Germany and Britain (and
possibly some other EEC countries). Switzerland is in Europe,
the sub-continent, but not in the European Community, or EEC.
It is however often possible for the same algorithm to be
implemented in both hardware and software. In such cases, a
hardware implementation may be patentable in Britain and/or
Germany. This only protects hardware implementations. An
all software implementation of the algorithm, running on a
general purpose computer, would not infringe.
> (According to the PGP documentation, IDEA was actually developed
> by ETH, an academic body, although there may well have been tied
> funding. Also according to that documentation, ASCOM have waived
> royalties on the use of IDEA in PGP, for both commercial and
> non-commercial use.)
This is a different deal. For one thing, the IDEA code in PGP is
copyright, quite separate from the patent issue. I believe that
at least one version of PGP had the wrong information about the
free licence for the IDEA code in the docs, and that Ascom have
not waived royalties for commercial use.
--
HALF A POUND COME ON SHAVERS
FOR YOU'RE IN LUCK
HALF A BUCK Burma Shave
Iolo Davidson
o...@eloka.demon.co.uk "Owen Lewis" writes:
> As I understand it. a patent is granted on an original idea.
Don't think so. You have to have more than an idea.
> It is an essential stipulation that a working example of the idea
> has been completed. In addition to a description of the concept,
> a complete, detailed description of a working example is required
> as a part of the patent application.
Yes. You can't patent an idea, per se. You patent the
application.
> It matters not at all that the implementation of the idea is a stream of code
> to be read and acted on by a computer (not invented here).
Yes it does, at least in Germany (some say also in the UK), where
a software implementation of any technique which runs on a
general purpose computer (rather than dedicated hardware) is
*specifically* excluded from elegibility for a patent.
> What has "copyleft" to do with the existence of a valid
> patent on IDEA?
The IDEA patent is Swiss, and is not valid in Germany for
software implementations running on general purpose computers,
and possibly not valid in other EEC countries in a similar way.
> > Also according to that documentation, ASCOM have waived
> >royalties on the use of IDEA in PGP, for both commercial and
> >non-commercial use.)
>
> For non-commercial use, yes. For commercial use, no.
This is my understanding also.
> My copies of PGP 1.0, 2.0, 2.1, 2.2, 2.3 and 2.3a documentation do
> not read as you claim. As I said, if you find this difficult to
> accept, why not ask ASCOM? They won't bite :-).
They might tell you that they have a valid patent when they
don't, though, due to the fact that their patent is valid where
they are but possibly not valid where you are.
Iolo Davidson
o...@eloka.demon.co.uk "Owen Lewis" writes:
> >This has been thrashed through several times in alt.security.pgp.
> >To the best of my memory, the concensus goes something like this:
> >
> >If you ask Ascom, they will say they have a patent. Switzerland
> >apparently allows software patents. Germany and Great Britain
> >*specifically* disallow patents on software.
>
> So you say. Evidence? As you say the disallowance is specific a quote,
> with reference from the UK statute or regulation will do nicely thanks.
Evidence of what? That it has been thrashed out thoroughly in
alt.security.pgp? That the position stated above was the
concensus? I haven't seen any supporting evidence, quotes from
legislation, etc., to support your position, so where do you get
off demanding them from me?
I am merely reporting what has happened when this subject came up
earlier this year. If you want to see all the stuff that the
participants cited, and there was a bunch, find an archive site
and dig. Otherwise I care not whether you accept my report or
not.
> Or is this something that you were once told and choose to believe?
This is something which raged for a month in alt.security.pgp.
Lot's of people, including natives of many EEC countries, took
part.
> The professional advice I paid for (expensively) says that you're
> absolutely and categorically wrong.
So this something that you were once told and choose to believe?
Lawyers never make mistakes of course, even when they can't tell
the difference between Europe and the EEC.
> Software patents
> >are therefore not enforceable in Germany and Britain (and
> >possibly some other EEC countries).
There is an important difference here between software running in
dedicated hardware and software running on a general purpose
computer. Someone *did* cite a German legislative reference on
this point. Better get your expensive lawyer to look again.
Owen Lewis
io...@mist.demon.co.uk "Iolo Davidson" writes:
>> > Ascom certainly do hold a valid patent that covers software
>> > implementations of IDEA. They expect and require proper
>> > payment for the commercial use of a product that has cost
>> > them a great deal to develop and to bring to market. If any
>> > one doubts this, why not ask them?
>
>This has been thrashed through several times in alt.security.pgp.
>To the best of my memory, the concensus goes something like this:
>
>If you ask Ascom, they will say they have a patent. Switzerland
>apparently allows software patents. Germany and Great Britain
>*specifically* disallow patents on software.
So you say. Evidence? As you say the disallowance is specific a quote,
with reference from the UK statute or regulation will do nicely thanks.
Or is this something that you were once told and choose to believe?
The professional advice I paid for (expensively) says that you're absolutely
and categorically wrong.
Software patents
>are therefore not enforceable in Germany and Britain (and
>possibly some other EEC countries).
Etc..... etc.....
--
-= Owen Lewis =-
@
Tele/fax +44-(0)794-301731 ELOKA Consultancy & Project Management
o...@eloka.demon.co.uk
pgp 2.x public key on request
David Swarbrick
>>If you ask Ascom, they will say they have a patent. Switzerland
>>apparently allows software patents. Germany and Great Britain
>>*specifically* disallow patents on software.
>
>So you say. Evidence? As you say the disallowance is specific a quote,
>with reference from the UK statute or regulation will do nicely thanks.
Hello Owen - how are you?
Patents Act 1977
---1.---
Patentable inventions
1.--(1) A patent may be granted only for an invention ...
2) It is hereby declared that the following (among other things)
re not inventions for the purposes of this Act, that is to say,
nything which consists of--
...
c) a scheme, rule or method for performing a mental act, playing
a game or doing business, or a program for a computer
There have I think been two cases where this has been tested in the way these
things are by optimistic patent agents.
Given the above, I think it falls to you to identify some patent which does
consist of a computer program.
--
David Swarbrick | Just Mooting UK Law On-line
Swarbrick & Co, Solicitors | +44 (0)484 401139 (24 hrs <=v.32bis)
22 Bradford Road Brighouse HD6 1RW|
Telephone [01484] 722531 | da...@swarb.demon.co.uk
John F Hall
> In article <787862...@mist.demon.co.uk>
> io...@mist.demon.co.uk "Iolo Davidson" writes:
>
> > ... ... Germany and Great Britain
> >*specifically* disallow patents on software.
>
> So you say. Evidence? As you say the disallowance is specific a quote,
> with reference from the UK statute or regulation will do nicely thanks.
Patents Act 1977, section 1(2)
(According to my book on Intellectual Property Law, published in 1990 and
which says it is accurate as at Sept 90. The 1988 Act modified patent law,
but didn't repeal the 1977 Act only amended it.)
The book details the following as being "not inventions" under that
section:
Discoveries.
Scientific theories.
Mathematical methods.
Literary, dramatic, musical or artistic works.
Other aesthetic creations.
A scheme, rule or method for performing a mental act.
A scheme, rule or method for playing a game or doing business.
Computer programs.
The presentation of information.
It is possible to patent a programmed computer where the computer program
is an essential part of the invention, but where the invention is more than
just the program.
> Or is this something that you were once told and choose to believe?
It's something I've been told several times by reliable sources.
> The professional advice I paid for (expensively) says that you're absolutely
> and categorically wrong.
There are actually two complications. One is the European Patent Office,
set up by the European Patent Convention in 1973. It can grant patents
binding in Austria, Belgium, Denmark, Greece, Spain, Switzerland, Germany,
France, Italy, Lichtenstein, Luxembourg, Netherlands, Sweden, the UK, and
possibly more. Secondly there is EU "harmonisation".
May I give a counter request. Do you know of significant changes in the
last four years?
Actually I feel quite strongly that it would be wrong to implement software
patents now. The UK has a very long history of computer programming. An
essential part of a patent is that it is innovative. You may recall the
chap on the Isle of Wight who was sued for substantial damages for refusing
to honour a patent granted to a US company for windsurfing. After a lot of
hassle his mother (I believe) found a photograph of him fooling around on a
surf board with a sail 10 years before the US patent was granted. He was
in fact the true inventor. The problem with old software is that it
doesn't leave such tangible traces - no physical form to photograph, no
purchase orders for parts, etc. I can think of several innovative
techniques I used in the 70s that I have never seen generally published.
If a company were to be granted a patent for them I would find it very
difficult to prove my prior use.
As you may gather my interest in, and knowledge of, software patents is as
a software developer rather than a patent specialist.
--
John F Hall jfh...@avondale.demon.co.uk
William Unruh
>The book details the following as being "not inventions" under that
>section:
>Discoveries.
>Scientific theories.
>Mathematical methods.
>Literary, dramatic, musical or artistic works.
>Other aesthetic creations.
>A scheme, rule or method for performing a mental act.
>A scheme, rule or method for playing a game or doing business.
>Computer programs.
>The presentation of information.
>It is possible to patent a programmed computer where the computer program
>is an essential part of the invention, but where the invention is more than
>just the program.
I believe that the typical patents are not for a computer program, but
for the use of some idea to accomplish a certain end. That that use is
embodied in a computer program is then irrelevant.
Eg, say RSA, the patent would be on the use of the difficulty of
factoring in order to impliment a public key crypto system. The patent
would be on that use of the idea of faxtoring, and not on the computer
program used to embody that idea. Thus you would not be sued for using
their computer program, you would be sued for implimenting crypto,
however you did it, by using the difficulty of factoring a number.
--
Bill Unruh
un...@physics.ubc.ca
J. R. Valverde (4423)
In article <3dbcrl$h...@nntp.ucs.ubc.ca>, un...@physics.ubc.ca (William Unruh) writes:
>From: un...@physics.ubc.ca (William Unruh)
>Newsgroups: demon.security,uk.legal,alt.security.pgp
>Subject: Re: UK Software Patents (was: File Encryption programs...)
>
>jfh...@avondale.demon.co.uk (John F Hall) writes:
>
>>The book details the following as being "not inventions" under that
>>section:
>
>>Discoveries.
>>Scientific theories.
>>Mathematical methods.
>>Literary, dramatic, musical or artistic works.
>>Other aesthetic creations.
>>A scheme, rule or method for performing a mental act.
>>A scheme, rule or method for playing a game or doing business.
>>Computer programs.
>>The presentation of information.
>
>>It is possible to patent a programmed computer where the computer program
>>is an essential part of the invention, but where the invention is more than
>>just the program.
>
>I believe that the typical patents are not for a computer program, but
>for the use of some idea to accomplish a certain end. That that use is
>embodied in a computer program is then irrelevant.
>Eg, say RSA, the patent would be on the use of the difficulty of
>factoring in order to impliment a public key crypto system. The patent
>would be on that use of the idea of faxtoring, and not on the computer
>program used to embody that idea. Thus you would not be sued for using
Why is it that this sounds like a "mathematical method" to me?
Why is it that encryption sounds like a different way for "presenting
information"?
Why, oh why...?
>their computer program, you would be sued for implimenting crypto,
>however you did it, by using the difficulty of factoring a number.
>--
>Bill Unruh
>un...@physics.ubc.ca
>
BTW, rules might have changed recently: As far as I know, the EU
decided to agree in a common set of rules for protecting software.
But, following the text of these rules, algorithms could not be
patented, nor methods nor bleh, bleh, bleh... You are even allowed
to disassemble a program to discover undocumented features if you
are going to make a new program that is not the same, but is going
to interface with the first one.
On these ground, you could dissasemble any program using RSA to
discover the underlying algorithm so as to be able to make a new
program that interfaces with the original. Say you have a Lotus
and want to make an add-on: there you have it.
Now, I know this has been updated in several EU countries, but can't
remember the list now and am not sure if it holds in UK. Maybe
someone knows better if these norms have been incorporated into
UK this year?
So, what? If they weren't you could always develop in any other EU
country, and let the market work.
jr
Owen Lewis
io...@mist.demon.co.uk "Iolo Davidson" writes:
>In article <787813...@eloka.demon.co.uk>
> o...@eloka.demon.co.uk "Owen Lewis" writes:
>
>> As I understand it. a patent is granted on an original idea.
>
>Don't think so. You have to have more than an idea.
Yes, you have to demonstrate that it is capable of practicl application.
This can be done using a software implementation.
>
>> It is an essential stipulation that a working example of the idea
>> has been completed. In addition to a description of the concept,
>> a complete, detailed description of a working example is required
>> as a part of the patent application.
>
>Yes. You can't patent an idea, per se. You patent the
>application.
Not as I am told. It is the idea that is proptected under patent. I.e. if
someone implements an idea using left handed rather than righthanded nuts,
they still require a licence from the patentee. Similarly, the recreation of a
patented concept by writing the code in C rather than in Pascal and painting
the screen blue rather than green will not aviod an obligation to licence.
>
>> It matters not at all that the implementation of the idea is a stream of code
>> to be read and acted on by a computer (not invented here).
>
>Yes it does, at least in Germany (some say also in the UK)
Yes, you said it.
>> What has "copyleft" to do with the existence of a valid
>> patent on IDEA?
>
>The IDEA patent is Swiss, and is not valid in Germany for
>software implementations running on general purpose computers,
>and possibly not valid in other EEC countries in a similar way.
I ask again, what has "copyleft" to do with the validity or not of a patent?
>> > Also according to that documentation, ASCOM have waived
>> >royalties on the use of IDEA in PGP, for both commercial and
>> >non-commercial use.)
>>
>> For non-commercial use, yes. For commercial use, no.
>
>This is my understanding also.
Good, then we are agreed.
Owen Lewis
>In article <787944...@eloka.demon.co.uk>
> o...@eloka.demon.co.uk "Owen Lewis" writes:
>
>> >This has been thrashed through several times in alt.security.pgp.
>> >To the best of my memory, the concensus goes something like this:
>> >
>> >If you ask Ascom, they will say they have a patent. Switzerland
>> >apparently allows software patents. Germany and Great Britain
>> >*specifically* disallow patents on software.
>>
>> So you say. Evidence? As you say the disallowance is specific a quote,
>> with reference from the UK statute or regulation will do nicely thanks.
>
>Evidence of what?
Please read the above. You stated that the law of Great Britain (by which
I take it you mean the law of England and Wales) specifically disallows patents
on software. I ask again, please, for your evidence.
>That it has been thrashed out thoroughly in
>alt.security.pgp? That the position stated above was the
>concensus? I haven't seen any supporting evidence, quotes from
>legislation, etc., to support your position, so where do you get
>off demanding them from me?
You made a specific claim. I just asked for the authority. I am quite content
with your statement that the matter had been discussed in this group but
have difficulty accepting that as an authority. See other posts re. the
UK statute law and its effect.
>This is something which raged for a month in alt.security.pgp.
>Lot's of people, including natives of many EEC countries, took
>part.
I believe you :-)
It's quite a common occurence :-)
>There is an important difference here between software running in
>dedicated hardware and software running on a general purpose
>computer.
In terms of patentability, what *is* the difference? My application runs on GP
platforms.
Someone *did* cite a German legislative reference on
>this point. Better get your expensive lawyer to look again.
Why don't we settle for the UK specifics and leave the German specifics to a
German?
Terry Ebdon
when programmed nor a disk containing a program). Generally, the
creator of computer software should look to copyright for his
protection."
From: A Legal Guide to Innovation, by Nabarro Nathanson
(Honorary Solicitors for the Prince of Wales Award for Innovation)
Price GBP 4, from 0171 493 9933. Proceeds go to Prince of Wales
Award for Innovation. Reviewed on p.57 Computing, 17 November 1994.
Terry
(No connection with the above publications or organisations)
Owen Lewis
da...@swarb.demon.co.uk "David Swarbrick" writes:
>Hello Owen - how are you?
Hello David! And how the devil are you? I'm fine but neglecting my friends
miserably. I'll e-mail more.
>Patents Act 1977
>---1.---
> Patentable inventions
>1.--(1) A patent may be granted only for an invention ...
>
>2) It is hereby declared that the following (among other things)
>re not inventions for the purposes of this Act, that is to say,
>nything which consists of--
>...
>
>c) a scheme, rule or method for performing a mental act, playing
> a game or doing business, or a program for a computer
>Given the above, I think it falls to you to identify some patent which does
>consist of a computer program.
See my lengthy post in reply to John Hall. Will the grant of a patent suffice?
Owen Lewis
Jfh...@avondale.demon.co.uk "John F Hall" writes:
>In article <787944...@eloka.demon.co.uk>
> o...@eloka.demon.co.uk "Owen Lewis" writes:
>Patents Act 1977, section 1(2)
>
>(According to my book on Intellectual Property Law, published in 1990 and
>which says it is accurate as at Sept 90. The 1988 Act modified patent law,
>but didn't repeal the 1977 Act only amended it.)
>
>The book details the following as being "not inventions" under that
>section:
>....etc.
>It is possible to patent a programmed computer where the computer program
>is an essential part of the invention, but where the invention is more than
>just the program.
My 'invention'is embodied in a program for the collection, collation and
analysis of spectral data in a manner that, across a theoretically unlimited
electromagnetic spectrum, reveals the presence of coherent activity that may
not be found by other means. It does this by taking several sets of data,
separated in time, space or both, from a suitable receiver or receivers
and manipulating the data to give information that neither the receiver(s) nor
the operator can, unaided, report as present nor easily determine the location
of the transmitting source. The invention is essentially a particular
application of inference analysis techniques to solve a class of problems.
It represents no great leap forward for mankind but it does make practical
some tasks that previously could not be attempted or, where the work was done
by other means, it could only be done at a cost prohibitive to most potential
users (of whom there are not too many in any event).
While custom built hardware could undertake the collation and analysis much
faster than the software implementatiion that was submitted in support of the
patent application, the gain in speed is made only at drastic increase in cost
and, to serve the present market, that speed gain is unnecessary. Future
special purpose applications could be implemented in custom built hardware.
It seems clear to both my patent attorney and the Patent Office that a
patent application for an invention implemented in software (that itself
is largely constructed with or draws on the facilities provided by other
patented and copyrighted software) is both unremarkable and allowable. As
I understand it, it is the unique manner in which a problem is solved that
earns a patent with a practical implementation of the innovation being
prerequisite for a valid impementation. There seems to be no bar on that
implementation being in software.
That said, am I right in saying that there is no major difference in our
positions?
>There are actually two complications. One is the European Patent Office,
>set up by the European Patent Convention in 1973. It can grant patents
>binding in Austria, Belgium, Denmark, Greece, Spain, Switzerland, Germany,
>France, Italy, Lichtenstein, Luxembourg, Netherlands, Sweden, the UK, and
>possibly more. Secondly there is EU "harmonisation".
Yes. My advice is that once a UK patent is granted, one may proceed to patent
at the European Patent Office and elsewhere. The process is expensive and it is
necessary to take a view as to whether the commercial value of the invention
warrants such expensive international protection. Either way, the filing of an
application for a UK patent is sufficient to give protection for twelve months
against the filing of a similar and competing application in any of the
signatory countries. I also understand that this is the position vis-a-vis
the US and some other countries.
> Do you know of significant changes in the last four years?
No. I'm not a lawyer :-) However, will not the grant of a patent be prima facie
proof?
>Actually I feel quite strongly that it would be wrong to implement software
>patents now. The UK has a very long history of computer programming.
That's a philosophical viewpoint to which you are, of course, entitled. Without
debating that point with you, let me wonder how else an investment of capital
and labour in an innovative process is to be protected? As I understand it, it
was an early recognition of this problem that led, internationally, to the
development of the patent system. Without some such protection, inventors are
(reasonably enough?) reluctant to share their knowledge and, without such
sharing of knowledge, the common wealth and common good are not best served.
I - and probably you - are aware of many instances of innovative machines or
software that passed into the hands of unscrupulous 'customers' for evaluation
is stripped, disassembled and copied with no sale ever being made nor licence
negotiated.
>An essential part of a patent is that it is innovative. You may recall the
>chap on the Isle of Wight who was sued for substantial damages for refusing
>to honour a patent granted to a US company for windsurfing. After a lot of
>hassle his mother (I believe) found a photograph of him fooling around on a
>surf board with a sail 10 years before the US patent was granted. He was
>in fact the true inventor. The problem with old software is that it
>doesn't leave such tangible traces - no physical form to photograph, no
>purchase orders for parts, etc. I can think of several innovative
>techniques I used in the 70s that I have never seen generally published.
>If a company were to be granted a patent for them I would find it very
>difficult to prove my prior use.
Yes, this is a major difficulty. Perhaps a certain amount of rough justice
is inevitable? The major risk, I believe, is borne by the would-be patentee
who, having made a substantial investment may subsequently find that he has
simply reinvented the wheel and has wasted his time and money. The only
protection I can see against this - and it is a long way from a complete
protection - is a good knowledge of the state of the art in the field of
innovation and through the searches required in the process of patenting.
Nevertheless, your example is a fair one of what may happen.
I suppose one might make a respectable argument that, outside of a purely
academic or hobby environment one should never invest resources in making
an innovation without first making a thorough search for what you seek from
available sources. A licence has got to be markedly cheaper in every case than
untertaking properly costed original research and development. Also - other
than for the dishonest - reinventing the wheel must be a pretty sterile
exercise. Ny sympathies are with Scott who, arriving at the South Pole on a
journey from which he never returned, had to record faithfully that Amundsen
had been there first.
>As you may gather my interest in, and knowledge of, software patents is as
>a software developer rather than a patent specialist.
Thank you for your thoughtful input. It will be interesting to see how things
develop.
Owen Lewis
un...@physics.ubc.ca "William Unruh" writes:
>I believe that the typical patents are not for a computer program, but
>for the use of some idea to accomplish a certain end. That that use is
>embodied in a computer program is then irrelevant.
This is exactly my understanding too. The secret is in the trick. Provided
that the trick can be practically demonstrated, the means by which the
demonstation is made is irrelevant.
>Eg, say RSA, the patent would be on the use of the difficulty of
>factoring in order to impliment a public key crypto system. The patent
>would be on that use of the idea of faxtoring, and not on the computer
>program used to embody that idea. Thus you would not be sued for using
>their computer program, you would be sued for implimenting crypto,
>however you did it, by using the difficulty of factoring a number.
Agreed, But only in countries where a patent is registered.
Richard Clayton
o...@eloka.demon.co.uk "Owen Lewis" writes:
>See my lengthy post in reply to John Hall. Will the grant of a patent suffice?
Years ago when I was but a junior in this trade my old and wise
Technical Director explained to me about patents. "Patents", he told me,
"are pretty much useless until you get to court. Patents which have
been upheld by that court. Ah. Now there's a thing of value."
This is not to deride the interest that we all have in the *granting*
of this patent. But when the House of Lords says that it's good ...
now that really WILL be amazing.
richard
-------------------------------------------------------------------------
Richard Clayton, T U R N P I K E Ltd tel: +44 1306 747747
Dorking Business Park, DORKING, Surrey, UK. RH4 1YL fax: +44 1306 747749
"Assembly of Japanese bicycle require great peace of mind" quoted in ZAMM
William Unruh
>"A computer program as such is not patentable (neither is a computer
>when programmed nor a disk containing a program). Generally, the
>creator of computer software should look to copyright for his
>protection."
>From: A Legal Guide to Innovation, by Nabarro Nathanson
> (Honorary Solicitors for the Prince of Wales Award for Innovation)
> Price GBP 4, from 0171 493 9933. Proceeds go to Prince of Wales
> Award for Innovation. Reviewed on p.57 Computing, 17 November 1994.
Note the crucial words "AS SUCH". As I read it this does not imply that
a patentable cannot be embodied in a computer program. It would then be
that patentable process that would be protected, and not the computer
program itself. It would be the use of that computer program to
accomplish that patented task that would be illegal, not the program
itself. If (by some chance) you could use that same computer program to
do someting else (eg use it to create pretty pictures on a computer
screen ratehr than to encypt communications) you would not be violating
the patent. However your use of that program to aencrypt data would
presumably violate the patent.
--
Bill Unruh
un...@physics.ubc.ca
Iolo Davidson
o...@eloka.demon.co.uk "Owen Lewis" writes:
> >Yes it does, at least in Germany (some say also in the UK)
>
> Yes, you said it.
Someone (not me) has posted the relevant reference from UK
legislation.
> >> What has "copyleft" to do with the existence of a valid
> >> patent on IDEA?
> >
> >The IDEA patent is Swiss, and is not valid in Germany for
> >software implementations running on general purpose computers,
> >and possibly not valid in other EEC countries in a similar way.
>
> I ask again, what has "copyleft" to do with the validity or not
> of a patent?
Nothing at all. Why do you think it does? Why have you mixed
"copyleft" up with the patent issue in the first place? The only
point *I* made about copyright was that the IDEA code used in PGP
is covered by copyright independently of any issue about patents.
There is no dispute about the applicability of copyright to
software.
> >> > Also according to that documentation, ASCOM have waived
> >> >royalties on the use of IDEA in PGP, for both commercial and
> >> >non-commercial use.)
> >>
> >> For non-commercial use, yes. For commercial use, no.
> >
> >This is my understanding also.
>
> Good, then we are agreed.
No we are not. I only agree that they have not waived their
claims, not that their claims are valid, at least in the UK and
Germany.
Iolo Davidson
o...@eloka.demon.co.uk "Owen Lewis" writes:
> >> >This has been thrashed through several times in alt.security.pgp.
> >> >To the best of my memory, the concensus goes something like this:
> >> >
> >> >If you ask Ascom, they will say they have a patent. Switzerland
> >> >apparently allows software patents. Germany and Great Britain
> >> >*specifically* disallow patents on software.
> >>
> >> So you say. Evidence? As you say the disallowance is specific a quote,
> >> with reference from the UK statute or regulation will do nicely thanks.
> >
> >Evidence of what?
>
> Please read the above. You stated that the law of Great Britain (by which
> I take it you mean the law of England and Wales) specifically disallows
> patents on software. I ask again, please, for your evidence.
Please read it yourself. I said "To the best of my memory, the
concensus goes something like this:" and then put the points you
refer to, which were the points in the concensus I was reporting.
If someone tells you "Fred says the moon is made of cheese" do
you demand he prove that the moon is made of cheese, or that Fred
said it?
In the meantime, I note that others have indeed quoted relevant
legislation for you.
> >That it has been thrashed out thoroughly in
> >alt.security.pgp? That the position stated above was the
> >concensus? I haven't seen any supporting evidence, quotes from
> >legislation, etc., to support your position, so where do you get
> >off demanding them from me?
>
> You made a specific claim. I just asked for the authority. I am quite content
> with your statement that the matter had been discussed in this group but
> have difficulty accepting that as an authority. See other posts re. the
> UK statute law and its effect.
I didn't say, "I saw it here, it must be true," I said "We have
had this discussion already, and this is how it came out last
time". Believe what you like.
> >There is an important difference here between software running in
> >dedicated hardware and software running on a general purpose
> >computer.
>
> In terms of patentability, what *is* the difference? My application
> runs on GP platforms.
The point was put that the issue of what constitutes a computer
program, and thus is not patentable under the specific
exceptions, hinges upon whether the program will run on a general
purpose computer or requires dedicated hardware.
Owen Lewis
io...@mist.demon.co.uk "Iolo Davidson" writes:
>In article <788269...@eloka.demon.co.uk>
> o...@eloka.demon.co.uk "Owen Lewis" writes:
>> I think we should stop now.
>
>You may stop whenever you like. I intend to carry on forever.
Have fun. Bye.
Graham King
From New Scientist 24/31 December 1994 (Vol 144 No.1957/1958) p.21:
QUOTE
Britain's Patent Office is calling on European countries to change
their laws to allow people to patent software instead of relying on
copyright. The office recently organised an open forum in which 150
legal experts and software producers agreed almost unanimously that the
system must change.
UNQUOTE
The very short article goes on to state that the UK Patent Act (1977)
prohibits patents on, inter alia, programs for computers, but that
the lack of a legal definition of what constitutes a program can
sometimes be exploited by skilful patent lawyers to obtain patents on
programs.
The article also notes that pending patents are published in Europe,
whereas their contents remain secret in the US.
Does anyone know anything about the "open forum" described in the
article. Specifically: which software producers were represented?
Do I smell a rat? Are the small-fry about to get stitched up?
Usual disclaimer applies; I'm not a lawyer.
--
Graham King (at home)
David Swarbrick
Jose.V...@ebi.ac.uk "J. R. Valverde (4423" writes:
>
>
>Why is it that this sounds like a "mathematical method" to me?
>Why is it that encryption sounds like a different way for "presenting
>information"?
>Why, oh why...?
Because it is
>BTW, rules might have changed recently: As far as I know, the EU
>decided to agree in a common set of rules for protecting software.
The Software _copyright_ directive is in effect. But copyright ain't
patent law.
>
>But, following the text of these rules, algorithms could not be
>patented, nor methods nor bleh, bleh, bleh...
Copyright ain't patent law.
You are even allowed
>to disassemble a program to discover undocumented features if you
>are going to make a new program that is not the same, but is going
>to interface with the first one.
>
>On these ground, you could dissasemble any program using RSA to
>discover the underlying algorithm so as to be able to make a new
>program that interfaces with the original. Say you have a Lotus
>and want to make an add-on: there you have it.
True - possibly
>
>Now, I know this has been updated in several EU countries, but can't
>remember the list now and am not sure if it holds in UK. Maybe
>someone knows better if these norms have been incorporated into
>UK this year?
Amendments have been made to the CDPA 1988
Richard Clayton
>>to disassemble a program to discover undocumented features if you
>>are going to make a new program that is not the same, but is going
>>to interface with the first one.
In article <788431...@swarb.demon.co.uk>
da...@swarb.demon.co.uk "David Swarbrick" writes:
>True - possibly
ISTR that the "right to disassemble" only exists if the original
manufacturer does not make information available for you to do the
interfacing.
This was a big sticking point at the time, since people wanted to be
able to build 3rd party add-ons to mainframes &c, but the mainframe
makers didn't want people disassembling their software, and ISTR
various threats being made about not shipping technology to Europe.
The compromise was that if technical info was made available then
the software remained protected.
usual disclaimers that this is just my memory, and not my expertise.
Owen Lewis
gra...@trio.demon.co.uk "Graham King" writes:
>Time to set another cat among the pigeons methinks...
>
>From New Scientist 24/31 December 1994 (Vol 144 No.1957/1958) p.21:
>QUOTE
>Britain's Patent Office is calling on European countries to change
>their laws to allow people to patent software instead of relying on
>copyright. The office recently organised an open forum in which 150
>legal experts and software producers agreed almost unanimously that the
>system must change.
>UNQUOTE
Yes, I read the article and wondered how long it would take to surface in this
discussion :-)
>The very short article goes on to state that the UK Patent Act (1977)
>prohibits patents on, inter alia, programs for computers, but that
>the lack of a legal definition of what constitutes a program can
>sometimes be exploited by skilful patent lawyers to obtain patents on
>programs.
As several who have posted here have pointed out, at least under the present,
law, that is what happens. Our patent application was for an innovative method
of producing useful information from data otherwise useless or of very
limited value. That the innovation's practical implementation is expressed in
software is immaterial. Patent law protection is claimed on the innovation
and copyright protection claimed on the software program that is licenced to
users.
Clearly, there must be occasions when the innovation and program
cannot be pulled apart and I presume that it is this aspect that concerns
the Patent Office and others.
>The article also notes that pending patents are published in Europe,
>whereas their contents remain secret in the US.
>
> Do I smell a rat? Are the small-fry about to get stitched up?
Well, as small fry, all I can say is that I don't feel stitched up, only
presented with hard choices. Life was ever thus.
In a perfect world, where all honour the spirit as well as the letter of
agreements, patents might not be required. However, in the world we live in,
copyright evasion is a daily exercise by many and a source of income to some.
In the popular mind it is a victimless crime, if indeed it is perceived as a
crime at all.
A major software supplier might be philosophical about a percentage of n^6
copies of his popular software being used unlicenced. He may even come to
accept that its spread in that manner serves to further popularise his
software and therebye to increase licence sales beyond the level that they
might otherwise have achieved. Hence, surely, the practice of giving licences
away for non-commercial use.
Not so for specialist software where the world market
may be no more than, say, 5,000 copies. The economics of recovering the R&D
and marketing globally mean that the unit price is high and every copy pirated
is a palpable loss. If the innovation in the product were to be 'lifted' by a
major company, who then write their own software implementation to avoid any
copyright, the innovative small fry is up the creek without a paddle.
At best, the small fry is faced will uncertain litigation where a major
opponent can easily extend the litigative costs to a point at which the
small fry cannot continue. Most likely, the prudent small fry faced with such
circumstances will simply limp off and lick his wounds, especially if offered
the inducement of 'get lost money'.
Small fry or not, imperfect system or not, the protection of an innovative
process under patent does seem the best way. The hard part for real small fry
is the cost of obtaining and maintaining full international patent protection.
As mentioned previously, this means that there must be some minimum potential
value of an innovation before a full process of patent protection becomes
a practical consideration. If that potential is present, then a venture
capital equity holder may become an inevitability, even when not gladly sought.
>Usual disclaimer applies; I'm not a lawyer.
Neither am I :-)
David Wadsworth
o...@eloka.demon.co.uk "Owen Lewis" writes:
>
> Small fry or not, imperfect system or not, the protection of an innovative
> process under patent does seem the best way. The hard part for real small fry
> is the cost of obtaining and maintaining full international patent protection.
I dispute that obtaining a patent is always the best way . I know of large
companies who have operated their own innovative processes under wraps
for tens of years. In that time they have had to pay no patent fees and
have retained the process secrets.
If they had patented the process, then they would have had to reveal it
in the patent, making it available to theft by other people. The current
costs of a patent in this country are high, and produce few benefits for
an individual. If your patent is stolen, firstly you have to discover the
fact, rather difficult in a world-wide setting. Then you have to become
involved in an expensive legal case, possibly against a large corporation.
Even if you win you could spend several years in non-productive legal work.
------
-David Wadsworth | -- stat rosa pristina nomine,
| nomina nuda tenemus. (Umberto Eco)
Mike Convey
>The book details the following as being "not inventions" under that
>section:
>
>Discoveries.
>Scientific theories.
>Mathematical methods.
>Literary, dramatic, musical or artistic works.
>Other aesthetic creations.
>A scheme, rule or method for performing a mental act.
>A scheme, rule or method for playing a game or doing business.
>Computer programs.
>The presentation of information.
>
>It is possible to patent a programmed computer where the computer program
>is an essential part of the invention, but where the invention is more than
>just the program.
The New Scientist (Dec 24) reports that Britain's Patent Office is calling
on European countries to change their laws to allow people to patent
software instead of relying on copyright. (After holding a conference of
legal experts and software producers).
So it looks like it may be possible to patent software in the future...
Mike
Anthony Naggs
>Time to set another cat among the pigeons methinks...
and what a cat!
>From New Scientist 24/31 December 1994 (Vol 144 No.1957/1958) p.21:
>QUOTE
>Britain's Patent Office is calling on European countries to change
>their laws to allow people to patent software instead of relying on
>copyright. The office recently organised an open forum in which 150
>legal experts and software producers agreed almost unanimously that the
>system must change.
>UNQUOTE
I haven't seen this NS yet, I would really like to know what the rationale
for this suggestion is. (Being a Brit who has had some contact with US
software patents).
>The very short article goes on to state that the UK Patent Act (1977)
>prohibits patents on, inter alia, programs for computers, but that
>the lack of a legal definition of what constitutes a program can
>sometimes be exploited by skilful patent lawyers to obtain patents on
>programs.
Strange. My recollection is that the Copyright, Design and Patent Act 1988
has a similar clause, it is surely the current legislation? Also my
recollection is that the European Directive on Computer Software Copyright
(1992?) stipulates that software cannot be patented. Does anyone have these
documents available to check?
>The article also notes that pending patents are published in Europe,
>whereas their contents remain secret in the US.
The US is always different: it allows publication of an invention's details
upto 1 year before filing the patent claim, and is a lot more flexible about
revising a patent application after it is filed.
>Does anyone know anything about the "open forum" described in the
>article. Specifically: which software producers were represented?
My bets are on DEC, IBM & HP.
>Do I smell a rat? Are the small-fry about to get stitched up?
There's certainly something strange going on.
Anthony Naggs
Owen Lewis
dwa...@etna.demon.co.uk "David Wadsworth" writes:
>In article <788537...@eloka.demon.co.uk>
> o...@eloka.demon.co.uk "Owen Lewis" writes:
>
>>
>> Small fry or not, imperfect system or not, the protection of an innovative
>> process under patent does seem the best way. The hard part for real small
>>fry is the cost of obtaining and maintaining full international patent
>>protection.
>I dispute that obtaining a patent is always the best way.
Specifically, I said that it was not. But, from where I sit, provided the
commercial worth is sufficient and the innovation patentable - two big ifs -
patenting is probably the best option.
> I know of large
>companies who have operated their own innovative processes under wraps
>for tens of years. In that time they have had to pay no patent fees and
>have retained the process secrets.
So do many of us. And there may be good and sufficent reasons for so doing.
There are so many variables. If it is bespoke software customed for a single
client, for example:
a. It quite possibly isn't patentable.
b. The economics of doing so might make little sense.
c. A confidentiality agreement will often do nicely.
However, many of us also know of work which contains no real innovation but
just a rearrangement or even a straight application of work procured from
elswhere. Where this is the case, the fact is quite often concealed from the
client. This phenomenon is not particular to software. Electronic circuits are
frequently potted to prevent casual inspection determining their origin of
easy copying of the circuit. The security industry is rife with this practice.
One might add that many security oriented programs and devices might not
withstand full disclosure as weaknesses can no longer be hidden.
>If they had patented the process, then they would have had to reveal it
>in the patent, making it available to theft by other people. The current
>costs of a patent in this country are high, and produce few benefits for
>an individual. If your patent is stolen, firstly you have to discover the
>fact, rather difficult in a world-wide setting. Then you have to become
>involved in an expensive legal case, possibly against a large corporation.
>Even if you win you could spend several years in non-productive legal work.
You pays your money and you takes your choice.
The truth is that with many innovations they cannot be concealed once they
are sold. Only the method of implementation can be obscured. Sell a zip
fastener and it is quite obvious, firstly, what a good idea it is and,
secondly, the innovative principle on which it works. Hence the desirability
of patenting the innovation. Without a patent, an innovation is there for
anyone to copy.
Did any one say RSA?
Owen Lewis
>So it looks like it may be possible to patent software in the future...
It is possible now, in the strict sense that what is patented is an
innovative process implemented in software. I.E. the process is patented,
the source code that contains the instructions for effecting the process is not.
The code may be copyrighted.
David Swarbrick
ric...@turnpike.com "Richard Clayton" writes:
>> You are even allowed
>>>to disassemble a program to discover undocumented features if you
>>>are going to make a new program that is not the same, but is going
>>>to interface with the first one.
>
>In article <788431...@swarb.demon.co.uk>
> da...@swarb.demon.co.uk "David Swarbrick" writes:
>
>>True - possibly
>
>ISTR that the "right to disassemble" only exists if the original
>manufacturer does not make information available for you to do the
>interfacing.
You are strictly correct. But it is an entirely moot point as to what
constitutes an interface. The Directive does not specify this, and I
doubt that the original writer of the software is able to specify where
a potential interfacer must interface.
One of the problems found when moving from MSDOS3 to MSDOS 4 was that
software houses had found all sorts of 'better' places to interface
with DOS.
>
>This was a big sticking point at the time, since people wanted to be
>able to build 3rd party add-ons to mainframes &c, but the mainframe
>makers didn't want people disassembling their software, and ISTR
>various threats being made about not shipping technology to Europe.
>The compromise was that if technical info was made available then
>the software remained protected.
>
But how much technical info has to be given out? The software house cannot
know what information is wanted until the other software houses (interfacers)
have said. The interfacers cannot say until they have the technical info.
Unless _all_ the info is released, potential users may be perfectly justified
in reverse engineering all the software to see where to interface and where
not. It is arguable that by interfacing at every point wit ha computer
program, you can destroy _all_ the copyright in it.
Cheers
John F Hall
I hope this isn't just "empire building" by the Patent Office at the
expense of the small developers. As I have said in other articles there
are good reasons for preferring copyright to patents for software
protection.
Perhaps we should patent some plots for novels. :-)
John F Hall
The "unique manner" must involve more than just thoughts, it must involve
new hardware. Provided it does do so it does not matter if the
implementation uses some software. However an idea expressed solely in
software running on a general purpose computer cannot be patented in the UK.
> That said, am I right in saying that there is no major difference in our
> positions?
Not quite.
> >Actually I feel quite strongly that it would be wrong to implement software
> >patents now. The UK has a very long history of computer programming.
>
> That's a philosophical viewpoint to which you are, of course, entitled. Without
> debating that point with you, let me wonder how else an investment of capital
> and labour in an innovative process is to be protected? As I understand it, it
> was an early recognition of this problem that led, internationally, to the
> development of the patent system. Without some such protection, inventors are
> (reasonably enough?) reluctant to share their knowledge and, without such
> sharing of knowledge, the common wealth and common good are not best served.
>
> I - and probably you - are aware of many instances of innovative machines or
> software that passed into the hands of unscrupulous 'customers' for evaluation
> is stripped, disassembled and copied with no sale ever being made nor licence
> negotiated.
Not philosophical at all. I have strong, and in my opinion valid, reasons
for disliking software patents. Software is adequately protected by
copyright, which covers the examples you state. There is the problem, of
course, of evidence and enforcement, but that exists for patents also.
> >An essential part of a patent is that it is innovative. You may recall the
> >chap on the Isle of Wight who was sued for substantial damages for refusing
> >to honour a patent granted to a US company for windsurfing. After a lot of
> >hassle his mother (I believe) found a photograph of him fooling around on a
> >surf board with a sail 10 years before the US patent was granted. He was
> >in fact the true inventor. The problem with old software is that it
> >doesn't leave such tangible traces - no physical form to photograph, no
> >purchase orders for parts, etc. I can think of several innovative
> >techniques I used in the 70s that I have never seen generally published.
> >If a company were to be granted a patent for them I would find it very
> >difficult to prove my prior use.
>
> Yes, this is a major difficulty. Perhaps a certain amount of rough justice
> is inevitable? The major risk, I believe, is borne by the would-be patentee
> who, having made a substantial investment may subsequently find that he has
> simply reinvented the wheel and has wasted his time and money. The only
> protection I can see against this - and it is a long way from a complete
> protection - is a good knowledge of the state of the art in the field of
> innovation and through the searches required in the process of patenting.
> Nevertheless, your example is a fair one of what may happen.
No. The major risk is borne by the ordinary software developer who is told
that an idea that he has been using on and off for ten years has been
patented and he now has to prove "prior use" and suffer the possible
disruption of litigation, injunctions, customers being scared off, etc.
You have missed the key point of my example, not that it happens, but that
hardware leaves footprints - software doesn't.
> I suppose one might make a respectable argument that, outside of a purely
> academic or hobby environment one should never invest resources in making
> an innovation without first making a thorough search for what you seek from
> available sources. A licence has got to be markedly cheaper in every case than
> untertaking properly costed original research and development. Also - other
> than for the dishonest - reinventing the wheel must be a pretty sterile
> exercise. Ny sympathies are with Scott who, arriving at the South Pole on a
> journey from which he never returned, had to record faithfully that Amundsen
> had been there first.
You seem to be speaking more like an engineer or inventor than like a
software developer. *All* software development is innovation; OK, where
possible one reuses existing modules and ideas, but in a new way; one also
from time to time has brand new ideas. It is totally unrealistic to
suggest that each idea has to be checked for originality - there's nowhere
to check amongst from other reasons.
Let me give an example: About 25 years ago as development manager for a
computer bureau I was responsible, inter alia, for their subroutine
library. I happened to be working on the hashing routine when I went to
lunch with a former employee of the company and we discussed what I was
doing. (Hashing works by producing a pseudo-random mapping of a key onto
an array of cells, with remapping to alternate cells if that cell is
occupied by a different key. The technique itself was well known; the
problems were the remapping technique, dealing with full arrays, etc. Most
existing books of the time recommended using arrays with prime numbers of
elements, and our existing routine had done so.) He pointed out that it
was unnecessary to squeeze the last drop of space out of the array (which
involve searching every cell), it would be sufficient to make 10 attempts,
if an empty cell wasn't found after 10 attempts the array was about 90%
full which was a reasonable packing, and the "array full" reply can be
given. In fact my implementation, after some trial and error, made 25
attempts. I overcame the requirement for a prime array by using a modified
quadratic hashing. Quadratic hashing says take two "random" transforms of
the key, "A" and "B", and trying A, A+B, A+2B, A+3B, etc, (modulo array
size). I used A, A+B, A+B+(B+1), A+B+(B+1)+(B+2), etc. That avoids the
problem of testing only a subset of the array when B and the array size
have a common factor.
The point is that I have *never* seen those two innovations, one mine, one
my colleague's, documented, and have never documented them outside the
company myself. (Why should I have given away our "edge" in using these
and similar ideas?) They were documented internally, but that
documentation has dissipated - the computer used, Elliot 803, now exists
only in the Science Museum. If someone were to produce a patent a couple
of years old detailing these, I would have great difficulty in proving that
they were our ideas. Sure I can produce program lists with old dates on
the top - and I can print up more with even older dates if you want, what
date would you like. I can produce old mag tapes, with the routine
recorded on - but has anybody got a 7 track mag tape handler about?
That's just one example. There was a *lot* of software innovation in the
60s and early 70s, both in our company and in others, and much of it is in
the same state - still useful, not earth-shattering, not worth protecting
by their users, but a really major pain if someone else were to patent them
in the belief that they'd thought of them (or even just to be awkward).
PS: Sorry for the delay in catching up after Xmas.
Clive D.W. Feather
> I can produce old mag tapes, with the routine
> recorded on - but has anybody got a 7 track mag tape handler about?
Huh ? The 803 used 35mm Magnetic Film.
Ah, nostalgia. 761026:770.
--
Clive D.W. Feather | Santa Cruz Operation | If you lie to the compiler,
cl...@sco.com | Croxley Centre | it will get its revenge.
Phone: +44 1923 813541 | Hatters Lane, Watford | - Henry Spencer
Fax: +44 1923 813811 | WD1 8YN, United Kingdom |
Peter G. Strangman
Jfh...@avondale.demon.co.uk "John F Hall" writes:
> expense of the small developers. As I have said in other articles there
> are good reasons for preferring copyright to patents for software
> protection.
For a start, it lasts longer.
--
Peter G. Strangman | Wer weiss was die wohl glauben,
Pe...@adelheid.demon.co.uk | Die uns zum Glauben schrauben?
http://134.220.198.66:8000 | (Friedrich von Logau)
Owen Lewis
>> >It is possible to patent a programmed computer where the computer program
>> >is an essential part of the invention, but where the invention is more than
>> >just the program.
>>
>> The New Scientist (Dec 24) reports that Britain's Patent Office is calling
>> on European countries to change their laws to allow people to patent
>> software instead of relying on copyright. (After holding a conference of
>> legal experts and software producers).
>
>I hope this isn't just "empire building" by the Patent Office at the
>expense of the small developers. As I have said in other articles there
>are good reasons for preferring copyright to patents for software
>protection.
Well, you have your view and I have mine - which is that there may be good
reasons to opt for both patent and copyright implementation.
Once more, with feeling :-) An original idea, demonstrated as having
practical application is patentable. Indeed, a patent is the only strong
protection on commercial rights to the idea. For the purposes of patent
application, the idea may be demonstrated as practical; this may be by
implememting it in code to run run on a computer. The coding,
for the moment at least, is not patentable but may be copyright.
>Perhaps we should patent some plots for novels. :-)
The magic words are "original" and demonstated as having practical application.
Owen Lewis
Sorry, no. New hardware is not required (in my case, not even envisaged).
>However an idea expressed solely in
>software running on a general purpose computer cannot be patented in the UK.
See the above.
>I have strong, and in my opinion valid, reasons
>for disliking software patents. Software is adequately protected by
>copyright, which covers the examples you state. As discussed in another post,
patent and copyright do not do - cannot do the same job. Where there is
original invention, they are complementary.
>> >I can think of several innovative
>> >techniques I used in the 70s that I have never seen generally published.
>> >If a company were to be granted a patent for them I would find it very
>> >difficult to prove my prior use.
>>
>> Yes, this is a major difficulty. Perhaps a certain amount of rough justice
>> is inevitable? The major risk, I believe, is borne by the would-be patentee
>> who, having made a substantial investment may subsequently find that he has
>> simply reinvented the wheel and has wasted his time and money. The only
>> protection I can see against this - and it is a long way from a complete
>> protection - is a good knowledge of the state of the art in the field of
>> innovation and through the searches required in the process of patenting.
>> Nevertheless, your example is a fair one of what may happen.
>
>No. The major risk is borne by the ordinary software developer who is told
>that an idea that he has been using on and off for ten years has been
>patented and he now has to prove "prior use" and suffer the possible
>disruption of litigation, injunctions, customers being scared off, etc.
Hmmmm. Prior art should be capable of proof without too much trouble.
Disruption is then non-existent (except to the patent aspirant).
Prior art = no patent.
>You have missed the key point of my example, not that it happens, but that
>hardware leaves footprints - software doesn't.
But an inventive idea (if ever used) does.
>> I suppose one might make a respectable argument that, outside of a purely
>> academic or hobby environment one should never invest resources in making
>> an innovation without first making a thorough search for what you seek from
>> available sources. A licence has got to be markedly cheaper in every case
than
>> untertaking properly costed original research and development. Also - other
>> than for the dishonest - reinventing the wheel must be a pretty sterile
>> exercise. Ny sympathies are with Scott who, arriving at the South Pole on a
>> journey from which he never returned, had to record faithfully that Amundsen
>> had been there first.
>
>You seem to be speaking more like an engineer or inventor than like a
>software developer.
Perhaps that's the key. My background is largely one of telecommunications
engineering. My invention is intimately bound up with radio engineering. That
it is implemented in software (only) matters not. It is coincidental.
I make no claim to be a software developer; only to have invented a process
that is implemented in software.
>*All* software development is innovation; OK, where
>possible one reuses existing modules and ideas, but in a new way......
Perhaps we could settle on the term 'creative'. All invention
is creative; however, not all creative work is inventive.
>... one also from time to time has brand new ideas.
And such are inventive and therefore might be patentable.
Perhaps if you felt the ideas were of value, it would have been sensible to
document them?
This is essentially the patenting process. But, as already discussed, it does
get expensive and is only worth it if the commercial value of the idea is
assessed as making the patenting process worthwhile.
Who actually owned the ideas, was it you or your employer?
>They were documented internally, but that
>documentation has dissipated - the computer used, Elliot 803, now exists
>only in the Science Museum. I
As an irrelevant aside, my introduction to programming was Elliot Autocode,
with routines running in Murray code on punched paper tape at 50 baud. It
almost put me off computing for life!
>If someone were to produce a patent a couple
>of years old detailing these, I would have great difficulty in proving that
>they were our ideas. Sure I can produce program lists with old dates on
>the top - and I can print up more with even older dates if you want, what
>date would you like. I can produce old mag tapes, with the routine
>recorded on - but has anybody got a 7 track mag tape handler about?
But you would have an equal difficulty to prove copyright, would you not?
No records = no rights. Ripples in a pond is an image that springs to mind.
>..... software innovation in the
>60s and early 70s, both in our company and in others, and much of it is in
>the same state - still useful, not earth-shattering, not worth protecting
>by their users
If you did not think it worth protecting - or even permanently recording - but
someone else did, that's tough luck (rough justice if you prefer).
>PS: Sorry for the delay in catching up after Xmas.
A happy New Year to you.
John F Hall
cl...@sco.com "Clive D.W. Feather" writes:
> In article <789738...@avondale.demon.co.uk>,
> John F Hall <Jfh...@avondale.demon.co.uk> wrote:
> > the computer used, Elliot 803, now exists only in the Science Museum
> [...]
> > I can produce old mag tapes, with the routine
> > recorded on - but has anybody got a 7 track mag tape handler about?
>
> Huh ? The 803 used 35mm Magnetic Film.
>
> Ah, nostalgia. 761026:770.
The next computer we used had 7 track tape.
John F Hall
o...@eloka.demon.co.uk "Owen Lewis" writes:
> In article <789744...@avondale.demon.co.uk>
> Jfh...@avondale.demon.co.uk "John F Hall" writes:
>
> >... ... As I have said in other articles there
> >are good reasons for preferring copyright to patents for software
> >protection.
>
> Well, you have your view and I have mine - which is that there may be good
> reasons to opt for both patent and copyright implementation.
>
> Once more, with feeling :-) An original idea, demonstrated as having
> practical application is patentable. Indeed, a patent is the only strong
> protection on commercial rights to the idea. For the purposes of patent
> application, the idea may be demonstrated as practical; this may be by
> implememting it in code to run run on a computer. The coding,
> for the moment at least, is not patentable but may be copyright.
>
> >Perhaps we should patent some plots for novels. :-)
> The magic words are "original" and demonstated as having practical
> application.
The problem, as you say, is proving "originality". Except my viewpoint is
not that of someone trying to protect his original idea, but of not having
someone else saying that I can't use it because he says it was his original
idea. I see a major difference between hardware where "prior art" leaves
physical traces, and software where there are no such traces that couldn't
have been forged. Sure there are well known techniques for recording dates
when one knows one wants to capitalise on it later, but these are not
applicable when one suddenly has to prove one was using an idea before a 5
year old patent that is produced.
I see the major difference between patents and copyright as being that
copyright give rights to each person commensurate with their efforts,
whereas patents give rights to the first person who has the idea. I see
the last as a recipe for futile, and expensive, argument when the ideas
only have an abstract form.
John F Hall
o...@eloka.demon.co.uk "Owen Lewis" writes:
> In article <789738...@avondale.demon.co.uk>
> Jfh...@avondale.demon.co.uk "John F Hall" writes:
>
> >No. The major risk is borne by the ordinary software developer who is told
> >that an idea that he has been using on and off for ten years has been
> >patented and he now has to prove "prior use" and suffer the possible
> >disruption of litigation, injunctions, customers being scared off, etc.
>
> Hmmmm. Prior art should be capable of proof without too much trouble.
How?
> Disruption is then non-existent (except to the patent aspirant).
> Prior art = no patent.
>
> >You have missed the key point of my example, not that it happens, but that
> >hardware leaves footprints - software doesn't.
>
> But an inventive idea (if ever used) does.
Not if they're nothing physical involved; nothing to buy, nothing to
photograph.
> >[I won't bother to repeat the example]
> >
> >The point is that I have *never* seen those two innovations, one mine, one
> >my colleague's, documented, and have never documented them outside the
> >company myself. (Why should I have given away our "edge" in using these
> >and similar ideas?)
>
> Perhaps if you felt the ideas were of value, it would have been sensible to
> document them?
They were documented at the time, to satisfy the requirements of the time,
which were that there was no question of a software patents, little idea of
copyright. In fact the general ethos was that we were in a young and
burgeoning industry and shared ideas. The "commercial" side of the
industry used confidentiality as much as anything for protection.
> This is essentially the patenting process. But, as already discussed, it does
> get expensive and is only worth it if the commercial value of the idea is
> assessed as making the patenting process worthwhile.
>
> Who actually owned the ideas, was it you or your employer?
Effectively both. Legally they would belong to the employer as they were
developed in the course of employment. But there was no way I could forget
them.
> >If someone were to produce a patent a couple
> >of years old detailing these, I would have great difficulty in proving that
> >they were our ideas. Sure I can produce program lists with old dates on
> >the top - and I can print up more with even older dates if you want, what
> >date would you like. I can produce old mag tapes, with the routine
> >recorded on - but has anybody got a 7 track mag tape handler about?
>
> But you would have an equal difficulty to prove copyright, would you not?
I have no desire to enforce copyright. My concern is only that someone
doesn't produce a 5 year old patent, and tell me that I can't use my 25
year old ideas.
> No records = no rights. Ripples in a pond is an image that springs to mind.
I'm not looking for rights against others, merely the right to continue to
use my own work (which is essentially the difference between patent and
copyright protection).
> >..... software innovation in the
> >60s and early 70s, both in our company and in others, and much of it is in
> >the same state - still useful, not earth-shattering, not worth protecting
> >by their users
>
> If you did not think it worth protecting - or even permanently recording - but
> someone else did, that's tough luck (rough justice if you prefer).
Again I ask you to think carefully about the US company that sued the
inventor of windsurfing for breach of a US patent, and my position if
someone were to have patented the "quadratic plus one" hashing algorithm.
The inventor of the windsurf board was just fooling about. But he was seen
and copied and many years later it was an industry. He had no thoughts of
having an exploitable invention. However he was very put out when 10 years
or so later he was told he would have to pay royalties to continue to use
his own invention. Fortunately his mother had a photo of him enjoying
himself on the Solent in her family album. (Somehow I feel a photograph of
me writing a program wouldn't have the same evidential effect.)
Jeff A Licquia
For a slightly different perspective on the issue of algorithm patents,
check out the rhetoric being spilled in comp.graphics and
comp.infosystems.www.??? concerning the recent GIF/LZW controversy, as well
as some of the reactions expressed on GNN (http://gnn.com/) and the League
for Programming Freedom (http://www.lpf.org/).
--
----------------------------------------------------------------------
Jeff Licquia (lame .sig, huh?) | Finger for PGP 2.6 public key
jali...@prairienet.org | Me? Speak for whom? You've got
lic...@cei.com (work) | to be kidding!
William Unruh
>>
....
>> Hmmmm. Prior art should be capable of proof without too much trouble.
...
>> >The point is that I have *never* seen those two innovations, one mine, one
>> >my colleague's, documented, and have never documented them outside the
>> >company myself. (Why should I have given away our "edge" in using these
>> >and similar ideas?)
>>
>> Perhaps if you felt the ideas were of value, it would have been sensible to
>> document them?
>They were documented at the time, to satisfy the requirements of the time,
>which were that there was no question of a software patents, little idea of
>copyright. In fact the general ethos was that we were in a young and
>burgeoning industry and shared ideas. The "commercial" side of the
>industry used confidentiality as much as anything for protection.
I think that a "trade secret" is not a bar to a patent. Ie, if you 20 years ago
developed something but kept it secret, and now someone patents it, you cannot haul
out your secret to invalicate the patent. However I am not clear on this (and may
depend on the country having to do with "first to file" vs "first to invent" etc.
--
Bill Unruh
un...@physics.ubc.ca
Russ Evans
>Those were the days, my friend. We thought they'd never end...... BTW,
>whatever *did* become of Sandie Shaw?
Mary Hopkin, actually. Sandie's still on the box, but last I heard
Mary was a happy housewife!
Russ
Bringing a Welsh thread back to s.c.c .... and the originator bears
a good Welsh name, too!
Owen Lewis
Jfh...@avondale.demon.co.uk "John F Hall" writes:
>In article <790016...@eloka.demon.co.uk>
>> Also, if a snapshot from the family album is sufficient to prove prior art,
>> then surely existence of your code on seven-track tape, now a museum item,
>> is equally good prima facie evidence?
>
>And you'll take my word for what's on the mag tape?
Its either there or its not. If its there any competent forensic lab will
recover it. Sledgehammer to crack a nut, I know: wouldn't it be easier to keep
a fully documented record of work of value?
>> >They were documented at the time, to satisfy the requirements of the time,
>> >which were that there was no question of a software patents, little idea of
>> >copyright.
>>
>> That's the rub, isn't it? Your work was, effectively, written on water. Had it>> either been registered or published, then 'prior art' would be as plain as a
>> pikestaff. While this would not, in itself, bring you any reward it would
>> absolutely defend against someone else subsequently claiming to have first
>> thought of your idea.
>
>There was no way, nor any need at the time, to register work; and the last
>thing we wanted to do was publish it.
In which case, you are in difficulty if someone reinvents your wheel,
registers it and makes a bundle from it. That may be hard but you had your
chance, didn't you? :-)
>> >I have no desire to enforce copyright. My concern is only that someone
>> >doesn't produce a 5 year old patent, and tell me that I can't use my 25
>> >year old ideas.
That's exactly what may happen if you cannot show prior art. How would you
be any better off in this regard under copyright law.
>> >I'm not looking for rights against others, merely the right to continue to
>> >use my own work (which is essentially the difference between patent and
>> >copyright protection).
>>
>> I don't follow you. Patent protects rights in an *idea* that has been given
>> practical expression. Copyright protects rights in the *expression* of an
>> idea, whether the idea is practical or not and whether the idea is original
>> or not, but confers no rights in the idea itself.
>
>Software is a special case. For literary works copying may involve various
>transformations: resetting in a different typeface, translating into a
>foreign language, etc. Because of the special nature of software and of
>computer processing, it is accepted that the (barred) transformations are
>quite wide: compiling, recoding in a different language, even decompiling
>and rewriting a similar program in a different language for a different
>computer. (On the other hand decompling may be permitted if it's the only
>way of making a legally owned program work to a different interface.)
And so what? Copyright still protects a method of implementation, with whtever
tools and materials. It does not protect an original idea. I can create a
painting of the Mona Lisa as many times as I like and sell them if I can.
However, I may not (were it protected by copyright:-)) make a facsimile copy
of Da Vinci's work and sell it. PRZ could implement the RSA algorithim in code,
ab initio but, had he - by any of the methods you suggest - copied a RSADSI
implementation........
>> Isn't this exactly how PRZ was able
>> to claim that an implementation of the RSA algorithm that he wrote from
>> scratch did not infringe RSADSI copyright on their implementation? And isn't
>> that how PGP up to 2.3a can be used outside of North America without
>> infringement of RSADSI copyright which is international as well as without
>> infringement of their patent which gives them North American rights only in
>> the RSA cipher system?
>
>No. Firstly it is US not UK law that applies. Secondly all sorts of other
>considerations arise for PGP: the patent's late filing, the US "munitions"
>law, the Swiss IDEA patent, etc.
Since we are discussing copyright here, the UK/US argument seems immaterial.
Copyright is subject to an international convention to which both countries
are signatory. Are you eliding copyright and patent? If so, why?
Owen Lewis
Jfh...@avondale.demon.co.uk "John F Hall" writes:
>In article <790017...@eloka.demon.co.uk>
> o...@eloka.demon.co.uk "Owen Lewis" writes:
>
>> In article <789962...@avondale.demon.co.uk>
>> Jfh...@avondale.demon.co.uk "John F Hall" writes:
>> This is no sudden change. Twenty years ago an old boss of mine, jointly with
>> the then Professor of Electronics at Bath, proved that the accepted
>> limitation of intelligibility in time division multiplexing communication
>> through a given bandwidth - the Nyquist rate - was wrong and could be
>> improved by a factor of > 7. They patented their idea.
>
>And this was implemented in comms technology to increase performance. I'm
>sure he (or a related person) delivered little boxes which had to be used
>to get the effect, and that these boxes had no other purpose.
Essentially, no. It consisted of a program for a unique manipulation
of a digital datastream.
>I still feel you have not grasped the ephemeral nature of software. I key
>it into a computer, without making any change to its appearance except
>temporarily to its screen; I compile it and write a floppy disk or tape
>cartridge, which looks like any other floppy disk or tape cartridge. The
>recipient reads it into his computer, which now behaves differently, though
>he doesn't know why, and it looks the same. Later he may delete the
>program from his disk, and it's as though it never was there; except that
>there are results on his disk or in his filing cabinet with no trace of how
>they got there; it's not possible to work back and say they *had* to come
>from running the program, there are many obvious alternatives.
I think I grasp it.
It's like writing on paper with a pencil that has a rubber ('eraser' for our US
readers) on the blunt end.
Or it's like making a machine with all the working parts carved from solid CO2.
Or altering an electromagnetic state in a way that passes information between
people in remote locations.
So what? Its easy to destroy the work leaving no trace. However, its just as
easy to preserve and record its creation.
I don't think you the case for software, that being ephemeral (when it runs) it
is unique or a special case.
John F Hall
o...@eloka.demon.co.uk "Owen Lewis" writes:
No. A CO2 machine still has a physical existence and can be seen and
photographed. A pencil with a built-in eraser still leaves a mark (for a
time). It's more like making drawings (and their masters) in invisible
ink, with a machine that can "read" them and produce an article, and asking
a clerk with a filing cabinet of these, with no other indication of what's
on them, to keep the important ones and throw away the rest. There is
perhaps a serial number in small print at the corner of each drawings, but
not a key to say what each is. (One of the drawings has "index" pencilled
in after the serial number.)
Perhaps my analogy isn't perfect, the point is that computer programs are
*never* seen. No one can come along 10 years later and say "I saw John
Hall writing that routine" or "I saw a prototype of that idea 10 years
ago" or "here's a photograph of the computer when it was running it" or
"this purchase invoice for magnetic tapes clearly includes the one it was
recorded on". They may be able to say "I remember John talking about
something similar to that", but I don't think that's sufficient in the case
of hardware products; the windsurfer had to produce *physical* evidence.
I remember the lunch I had with Harry Lawrence and what we discussed;
there's no guarantee he does.
John F Hall
o...@eloka.demon.co.uk "Owen Lewis" writes:
> In article <790228...@avondale.demon.co.uk>
> Jfh...@avondale.demon.co.uk "John F Hall" writes:
>
> >In article <790016...@eloka.demon.co.uk>
>
> >> Also, if a snapshot from the family album is sufficient to prove prior art,
> >> then surely existence of your code on seven-track tape, now a museum item,
> >> is equally good prima facie evidence?
> >
> >And you'll take my word for what's on the mag tape?
>
> Its either there or its not. If its there any competent forensic lab will
> recover it. Sledgehammer to crack a nut, I know: wouldn't it be easier to keep
> a fully documented record of work of value?
I don't believe it, unless they can find an actual mechanism in existence
somewhere. There are two many stacked encoding schemes for it to be done
by hand.
> >> >They were documented at the time, to satisfy the requirements of the time,
> >> >which were that there was no question of a software patents, little idea of
> >> >copyright.
> >>
> >> That's the rub, isn't it? Your work was, effectively, written on water. Had
> it>> either been registered or published, then 'prior art' would be as plain
> as a
> >> pikestaff. While this would not, in itself, bring you any reward it would
> >> absolutely defend against someone else subsequently claiming to have first
> >> thought of your idea.
> >
> >There was no way, nor any need at the time, to register work; and the last
> >thing we wanted to do was publish it.
>
> In which case, you are in difficulty if someone reinvents your wheel,
> registers it and makes a bundle from it. That may be hard but you had your
> chance, didn't you? :-)
If someone sees commercial value where I didn't and exploits it, good look
to him. However I would object strongly to having to pay him to use my
ideas.
> >> >I have no desire to enforce copyright. My concern is only that someone
> >> >doesn't produce a 5 year old patent, and tell me that I can't use my 25
> >> >year old ideas.
>
> That's exactly what may happen if you cannot show prior art. How would you
> be any better off in this regard under copyright law.
Copyright doesn't require me to pay royalties if I develop the idea
independently. I don't have to prove I developed it *first*.
> >> >I'm not looking for rights against others, merely the right to continue to
> >> >use my own work (which is essentially the difference between patent and
> >> >copyright protection).
> >>
> >> I don't follow you. Patent protects rights in an *idea* that has been given
> >> practical expression. Copyright protects rights in the *expression* of an
> >> idea, whether the idea is practical or not and whether the idea is original
> >> or not, but confers no rights in the idea itself.
> >
> >Software is a special case. For literary works copying may involve various
> >transformations: resetting in a different typeface, translating into a
> >foreign language, etc. Because of the special nature of software and of
> >computer processing, it is accepted that the (barred) transformations are
> >quite wide: compiling, recoding in a different language, even decompiling
> >and rewriting a similar program in a different language for a different
> >computer. (On the other hand decompling may be permitted if it's the only
> >way of making a legally owned program work to a different interface.)
>
> And so what? Copyright still protects a method of implementation, with whtever
> tools and materials. It does not protect an original idea. I can create a
> painting of the Mona Lisa as many times as I like and sell them if I can.
> However, I may not (were it protected by copyright:-)) make a facsimile copy
> of Da Vinci's work and sell it. PRZ could implement the RSA algorithim in code,
> ab initio but, had he - by any of the methods you suggest - copied a RSADSI
> implementation........
No. "Method of implementation" is inventor's jargon. Copyright of
literary works, which includes computer programs, gives protection against them
being copied or adapted; and "adaption" is interpreted widely in the case
of programs, and bars the uses I detailed. In the words of the Copyright
Act (paragraph 21(4)): "In relation to a computer program a 'translation'
includes a version of the program in which it is converted into or out of a
computer language or code or into a different computer language or code,
otherwise than incidentally in the course of running the program". You
will note it bars *all* such translations, nor merely mechanical ones. If
the Mona Lisa were like a computer program you could not repaint it after
having seen it, even if you were to do so in the style of Picasso.
> >> Isn't this exactly how PRZ was able
> >> to claim that an implementation of the RSA algorithm that he wrote from
> >> scratch did not infringe RSADSI copyright on their implementation? And isn't > >> that how PGP up to 2.3a can be used outside of North America without
> >> infringement of RSADSI copyright which is international as well as without
> >> infringement of their patent which gives them North American rights only in
> >> the RSA cipher system?
> >
> >No. Firstly it is US not UK law that applies. Secondly all sorts of other
> >considerations arise for PGP: the patent's late filing, the US "munitions"
> >law, the Swiss IDEA patent, etc.
>
> Since we are discussing copyright here, the UK/US argument seems immaterial.
> Copyright is subject to an international convention to which both countries
> are signatory. Are you eliding copyright and patent? If so, why?
The boundary between patents and copyrights (and their overlap) is
different in the US and in the UK. I was just pointing out I'm not
prepared to draw any conclusions from US interpretations.
Kevin Martin
jfh...@avondale.demon.co.uk (John F Hall) wrote:
[157 lines with about 10% new content snipped]
Time to take it to e-mail, gentlemen?
Peter G. Strangman
> The problem with copyright is that it lasts indefinitely, unlike
> patents which have the grace to expire after 18 years.
Dunno 'bout the States but in Blighty copyright lasts until
25 years after the death of the author, that is definitely
not indefinetly. 8-)
Iolo Davidson
"HOLT A" writes:
> Pe...@adelheid.demon.co.uk ("Peter G. Strangman") writes:
>
> >Dunno 'bout the States but in Blighty copyright lasts until
> >25 years after the death of the author, that is definitely
> >not indefinetly. 8-)
>
> Not quite correct .... it used to be 50 years, but recent
> European regulations have extended it to (?)75 years.
How recent? The 1991 "Writer's & Artist's Yearbook" has a
detailed section on copyright which states 50 years as the
duration, quoting the 1989 statute changes. There is a 25 year
duration applying to typefaces and the case of "artistic works
which have been industrially exploited", whatever that means.
The duration normally runs from the year of the author's death,
but in some cases from the year of publication.
Life of author plus fifty years is also the term agreed under the
Berne Convention.
USA copyright law (1955 ratification of the Universal Copyright
Convention) set the duration for life of author plus twenty-five
years. In 1978 this was changed to life of author plus fifty
years. Because of the change, it is possible in the USA to apply
for an extension of copyright for works first copyrighted before
1978. Depending on the timing, the extension might be for a
further 47 years, or for a total of 75 years from when copyright
was first secured. I think this provision is the source of the
75 year period you mention.
--
SPECIAL SEATS WHO SCRATCH
RESERVED IN HADES THE LADIES
FOR WHISKERED GUYS Burma Shave
HOLT A
... quoted material elided
>>
>> Not quite correct .... it used to be 50 years, but recent
>> European regulations have extended it to (?)75 years.
>How recent? The 1991 "Writer's & Artist's Yearbook" has a
>detailed section on copyright which states 50 years as the
>duration, quoting the 1989 statute changes. There is a 25 year
>duration applying to typefaces and the case of "artistic works
>which have been industrially exploited", whatever that means.
I think the change occurred within the last 13 months probably
Jan 1st 1994). The EC "harmonised" copyright regulations which in each
country had been a particular compromise between the interests of
the creators and consumers by taking the strongest definition
from each of the member states, thus:
70 (or 75) years [ as in Germany? incresing that in most others]
from creators death [ as in UK, where others had from date of
creation]
weak definition of original [in Germany a work had to be creative
to be eligible]
probably also modifications to "fair dealing"
The publishers lobby won hands down at the expense of the consumer.
Andy
Owen Lewis
io...@mist.demon.co.uk "Iolo Davidson" writes:
>> >Have you ever actually written a program?
>>
>> Yes. Hence, my rebuttal of your proposition that to do so
>> is something magical, mystical, and essentially different from
>> any other field of human endeavour, to the point where the
>> ownership of rights in such work should be treated in an entirely
>> unique manner.
>
>So copyright is the way to go, then.
I weep :-)
Most of the world uses letters patent to protect rights in an inventive idea.
Call this A.
Copyright is used to protect an original implementation of an idea which, in
itself need not be inventive.
Call this B.
You may hold A and B.
You may be able to obtain A but not own B.
It is common that B exists without A. This may be because A has not been
applied for; there is insufficient invention to do so or there is a legal
prohibition on an application for A.
Finally, you may join the happy mass of humanity who neither know nor care
about A or B.
Clear now :-)?
Owen Lewis
Fr...@foxhouse.demon.co.uk "Fred Read" writes:
>In article <791096...@eloka.demon.co.uk>
> o...@eloka.demon.co.uk "Owen Lewis" writes:
>
>> In particular regard to the patenting of software, I tend to the view >> that it would be in implicit recognition of the lack of substantial
>> inventiveness in programming, per se, that, in this country, makes it
>> impossible to patent a program, per se. However, as has been previously
>> pointed out, by others as well as myself, it is quite as possible to patent
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>> an invention whose practical expression is through the medium of a computer
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>> program as it is to patent an invention made of brass.
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>
>But surely just as the *brass* is not fundemental to the patent, neither is
>the computer program.
You have it exactly right! It is the idea which is patented. The
implementation will be copyright. This is true even where the invention is one
that is only made practicable through a computer program.
>A similar device could be made using:
> (1) a different processor
> (2) a different language
> (3) a different algorithm
>
>Would any of these infringe the patent ?
Again, straight to the guts of it. Yes, any other implementation of the same
inventive idea will infringe the patent. The changes that you suggest might
well evade copyright infringement. Hence the particular value of a patent.
Owen Lewis
Pe...@adelheid.demon.co.uk "Peter G. Strangman" writes:
>As a 'small' writer I cannot afford to go through the process of
>obtaining a patent, does this mean that I should not be allowed to
>enjoy the fruits of my labours?
>
>Copyright has served 'writers' very well for a very long time,
>let's keep it that way.
That is the rub isn't it. International patenting - and these days nothing
else will really do the trick - is expensive.
The inventor has to weigh the balance of commercial advantage; but that is
something that those in business must do every day. As you say, copyright is
a cheap and pretty effective means of protecting your *implementation* of an
idea.
Owen Lewis
fla...@v2.ph.gla.ac.uk "Alan J. Flavell" writes:
>In article <790453...@eloka.demon.co.uk>, o...@eloka.demon.co.uk (Owen Lewis)
> writes:
>>A bore I know but what else is there to do? Put the boot on the other foot:
>>You spend a year, time and money, on developing a super idea to have someone
>>say (and there are those who do try it :-)), "Oh,
>
>I though the problem was rather a different one. You use an idea that
>seems obvious to you. Later you find that someone else has patented,
>copyrighted, published or is otherwise claiming it as their own
>original work. It never occurred to you to keep a record of such an
>"obvious" (as it had seemed to you) thing.
I don't think so. You first have some experience of a problem or limitation.
You then have an (imperfect/incomplete) idea for a solution. You specify,
develop, test, retest and trial an implementation of your idea until it works
well enough. That's if you're lucky. Most of the original ideas I've ever had
have never made it past the waste basket or else have only been useful in a
pretty limited sense.
>If you had spend a year and time and money on it, you'd be that much
>more likely to have some evidence. The problem was that you were too
>clever, believed it obvious, and have no evidence of your work.
>
>>>> >> >I have no desire to enforce copyright. My concern is only that someone
>>>> >> >doesn't produce a 5 year old patent, and tell me that I can't use my 25
>>>> >> >year old ideas.
>
>something like that.
Well, you pays your money and takes your choice.
Phillip M. Hallam-Baker
In article <791415...@adelheid.demon.co.uk>, Pe...@adelheid.demon.co.uk ("Peter G. Strangman") writes:
|>In article <3gbrtr$2...@news1.best.com> dim...@best.com "Phil Jones" writes:
|>
|>> The problem with copyright is that it lasts indefinitely, unlike
|>> patents which have the grace to expire after 18 years.
|>
|>Dunno 'bout the States but in Blighty copyright lasts until
|>25 years after the death of the author, that is definitely
|>not indefinetly. 8-)
Unless the author is dead then copyright does last an indefinite time, just one
that is bounded by the expiry date of the author.
IF (its not certain yet) the Germans get away with extending the protection
to 75 years then copyright could last for almost 200 years. If Alice (aged 3)
writes some cutesy bit of scribble, and pegs out at the age of 123 thats
200 years.
If someone perfected cryogenics then there would be a clever way to keep a
practically perpetual copyright. imagine the authour writes a book, and
is immediatey shoved in the freezer and pronounced "dead". 74 years later
the greedy publishers defrost him just long enough to prove that he is still
alive and once this is established refreeze him...
The main danger of such a scheme is that as the years rolled by the authour
would have to become celibate or face a significant risk of committing incest
with one of his descendants.
--
Phillip M. Hallam-Baker
Not Speaking for anyone else.
Bo Noren
We are all wanting for an answer on the question to patent computer
software. However,there is a previous deciousion passed by the Swedish
Government Court (Regeringsratten) which indicates that computer software
should in fact be patentable. However, this decision (the Pillips case)
indicates that sutch a patent must be written in a form that supports
the basic invention relating to how a certain precedure is performed
(and evaluated) by a computer, i.e. hense no protection for a program code.
Considering how this topic was treated in Switzerland, i.e. based on
whether or not a certain percentage of the program code was different to
the original code, I do support a judgement based on the fact whether or
not a certain program sequence can be regarded as priorly known.
I must say, knowing of the possiblity to change program code by using a
different language, that I support any judgement based on the actual way
that a problem is solved, not on the ability to change the program code
from one programming language into another. And, accordingly, one can not
base the question of infringement on differences in program code, since
that this would would make it possible for another part to use a new
method simply by changing the programming language.
Bo Norén
European Patent Attorney
Owen Lewis
>Well, this question is indeed a major question throughout Europe.
>
>We are all wanting for an answer on the question to patent computer
>software. However,there is a previous deciousion passed by the Swedish
>Government Court (Regeringsratten) which indicates that computer software
>should in fact be patentable. However, this decision (the Pillips case)
>indicates that sutch a patent must be written in a form that supports
>the basic invention relating to how a certain precedure is performed
>(and evaluated) by a computer, i.e. hense no protection for a program code.
That's interesting. It reflects the professional advice I was given regarding
a UK patent application supported by an implementation as a computer program.
>
>I must say, knowing of the possiblity to change program code by using a
>different language, that I support any judgement based on the actual way
>that a problem is solved, not on the ability to change the program code
>from one programming language into another. And, accordingly, one can not
>base the question of infringement on differences in program code, since
>that this would would make it possible for another part to use a new
>method simply by changing the programming language.
>Considering how this topic was treated in Switzerland, i.e. based on
>whether or not a certain percentage of the program code was different to
>the original code, I do support a judgement based on the fact whether or
>not a certain program sequence can be regarded as priorly known.
I'm not sure I follow. Many software programs comtain chunks of pretty
standard library code. However, these are merely the means to an end. So long
as the end being served is original and inventive.....?