Autodesk Inc & Anor v Dyason

[1991] HCATrans 97

A	--.i, AUST!l.LI.IA,,ii'
-~» .... ~--~

IN THE HIGH COURT OF AUSTRALIA

Office of the Registry

Melbourne No M29 of 1990

B e t w e e n -

AUTODESK INC. and AUTODESK

AUSTRALIA PTY LTD

Appellants

and

MARTIN PATRICK DYASON,
CHRISTINE DYASON and PETER

VINCENT KELLY

Respondents

MASON CJ
BRENNAN J
DEANE J
DAWSON J

GAUDRON J

Autodesk(2) 79 18/4/91

TRANSCRIPT OF PROCEEDINGS

AT CANBERRA ON THURSDAY, 18 APRIL 1991, AT 9.48 AM

(Continued from 17/4/91)

Copyright in the High Court of Australia

MASON CJ: Yes, Mr Burnside.

MR BURNSIDE: 

If the Court pleases. We start with the

proposition that in any copyright infringement case
the first task of the court is to identify the
subject-matter which is protected by the Act, that

is to say, to identify the copyright work.

In the course of their submissions yesterday,

the appellants said that the program in which they
claim copyright is in the lock and the program is
in WIDGET.C and they assert, without elaboration,

that the two are the same.  They seek to make good

that assertion by saying that the two things

perform the same function.  By this we understand

them to mean that both the relevant part of

WIDGET.C and the AutoCAD lock will produce the same

127 bit sequence when appropriately stimulated, and

once the 127 bit sequence has been completed once

it will begin again.  Now, that is no more than

saying that the two devices produce the same
output.

Can we make a comment in passing about what

was referred to repeatedly as the wraparound

feature. The proper use of language has a great

deal to do with the proper disposition of the questions in this case because the underlying

reality, which we are all attempting to discuss, is

complex and is involved in a technology that is

unfamiliar to most people.

It is discussed in the language of metaphor which is more or less happily conveyed in words. Any misapplication of the words to describe the

metaphor will involve a serious misunderstanding of

the underlying reality. The wraparound feature is

a perfect example of that. It is put forward as if

it is some substantial and consequential thing

which assists the claim for protection.

We were reminded of an expression used by an

anthropologist who wanted to explain the fact that

whilst human beings generally walk upright, they

can also lie down and that was referred to as the

horizontality mode capability, but to dress it up

in elaborate and abstract language is to confuse

the underlying reality rather than reveal it and to

seek to dress the chimera and thereby give it some

sense of reality.

The wraparound feature is nothing more than

this: that the string repeats itself. So if one

device produces the same 127 bit string repeatedly,

you can say that that has a wraparound feature, but

to say that, we would say, simply elevates the

notion beyond its reality and obscures what is

Autodesk(2) 80 18/4/91

really happening. Of course, if one device

produces the same 127 bits repeatedly, then in
order to produce the same output as that device any

other device will also have to produce the same 127

bits repeatedly and it is misleading and irrelevant

and distracting to say that each has the wraparound

feature.

Now, in order to identify the work, we would

say it is necessary first to locate it. In order

to locate it you have to distinguish clearly

between the input to the device which contains the program; the program or alleged program itself and

the output, which it produces by its operation.	In
this case, it is common ground that the input to
the lock can be referred to as transitions.	We
think transitions is the correct description
because it is nothing more than a change in
voltage.	The lock will not respond to anything

except a change in voltage. It does not matter for

present purposes whether it is a change from zero

to positive, or from positive to zero. What

matters is that there is a change in voltage. That

is all that it will recognize.

In the course of submissions yesterday that

phenomenon was referred to frequently as digital

information. This is another example, we would

say, of dressing up a plain reality in elaborate

clothes in order to dignify it. It is a misuse of

words to call voltage transitions simpliciter

digital information, and especially in the context

of computers.

The reason for that is this: the evidence was

that you could stimulate the lock to do what it

does by attaching to it a battery. Do not put it

on the computer at all. Just attach the

appropriate pins of it to a battery, and then

intermittently make contact with the battery so

that the circuit of battery to lock is completed.

Each time you touch the wire on the battery you

will get a voltage transition because if the wire

is voltage transition.

is not touching the battery no electricity flows. flows. That

the lock responds when it is connected to the
computer, but to call the application of the
battery the input of digital information, we would

say, is overstating the matter by a large measure.

There is another reason for saying that. The

evidence was that if a series of voltage pulses

were sent to the lock and if sliced in time so that

they could be treated as binary digits you could

Autodesk(2) 81 18/4/91

read them, for example, as being 101 if you had

that pattern of bits going to it. That would

involve a voltage transition.

If the series of bits sent to the lock was 10001 that would clearly be a different number,

looked at in binary terms, but it would still be

recognized by lock simply as one voltage

transition.

BRENNAN J:  Two, would it not?

MR BURNSIDE: One, Your Honour. That is assuming that the

voltage transition is from positive to negative.

101 involves only one downward transition. When

the next zero comes along there will be another

one.

BRENNAN J:	What about the upward transition?
MR BURNSIDE:	The evidence was, Your Honour, that it only responds to a transition in one direction. In
fact, it really does not matter. Let us suppose it responded to voltage transitions in both
directions.  101 would be two transitions.
One followed by an infinite number of zeros and
another one would still be two voltage transitions.
Looked at as numbers they are different. 101 in
binary is what we would call 5 in decimal, I think.
10001 in binary is what we would call 33. Now, 5
and 33 are clearly data, but to the lock 5 and 33
are the same.  And to call a lock which receives
5 or 33 and regards them as identical, to say that
that lock is receiving digital information, we
would say, is a misuse of language. It responds
simply to voltage transitions.  Now, the output of
the lock - - -
DEANE J:	Mr Burnside, could I ask you a rather stupid

question and that is what is the difference between the one and zero and what I would call positive and

negative?

MR BURNSIDE: It is not a stupid question, Your Honour, at

all. One zero one and positive and negative are

both metaphors, but they are conventional

metaphors. They are ways of describing, in

convenient form, the underlying electrical

phenomena. The underlying electrical phenomenon is

that the voltage level alters from time to time.

The essence of binary and digital equipment is that

it will only recognize either of two electrical

levels and they are typically set at zero and five

or something of that sort. They will not recognize anything in between. Because of the application of

Boolean algebra at a very low level in computers,

they can manipulate series of binary states in a

Autodesk(2) 82 18/4/91

lot of very complicated ways to produce useful

results. In order to manage what is going on for

the purpose of understanding it, you have to

describe the two states in some notation; one and

zero are typically used to describe the presence or

the absence of an electrical current.

DEANE J:  Now, when you said that you could produce this

result by using a battery, just to get it within the confines of my understanding, does that mean

that you would have to have it connected with both

poles of the battery?

MR BURNSIDE: Yes.

DEANE J:  Or would it suffice if it was connected just to

the positive or to the negative?

MR BURNSIDE:  No, in order to get a flow of current from a

battery it is necessary for both poles to be

connected to something which will internally
complete the circuit. In the example I was putting

forward - and let me say Mr Kelly was asked

questions about this and he himself said you could

do it this way - I was postulating that, say, the

negative pole of the battery is attached to the

appropriate input pin of the lock and the positive

pole of the battery is intermittently - randomly,

if you like - put in contact with the other

relevant pin of the lock. By making that

connection you cause a voltage transition to which

the lock will respond.

Now, the output, then, is the 127 bit string

which repeats itself·endlessly as long as

transitions continue to come into the lock.	The
output is plainly not a part of the lock.	It is
the consequence of the lock operating.	The

operation of the lock is the cause; the output is

the effect.	Now, once one looks then in the lock
to see what the program is, or the alleged program,

we would say that one has to be extremely careful

with language because the function or purpose of

the lock is to produce the desired output.	But
what you need to identify is the set of

instructions which will cause the lock to perform

that particular function. And that picks up

exactly the language of the definition of "computer

program".

It is inadequate in this part of the exercise

merely to identify the function which is performed

by looking at the output because to do that is to

identify no more than the function which the

instructions are designed to achieve. But the

exercise of identifying the subject-matter of

Autodesk(2) 83 18/4/91

copyright is the exercise of identifying the

instructions.

Now, I started by saying that the appellants

said, "Well, there is WIDGET.C and there is no

dispute that that is a program. It does the same

thing as the lock. It must follow, therefore, that

there is a program in the lock and it is the same

program." But that simply, in our submission,

misses the point and it is an invalid attempt at
the exercise of identifying the copyright work.

It is regrettably necessary to take the Court to the evidence about what the lock actually does. Now, I will not take the Court to actual transcript

pages but I will refer to any page references that

are needed. I think that my discussion of what the

lock does inside is non-contentious.

Inside the AutoCAD lock there exists essentially two little devices. One is called a

shift register and the other is called an

exclusive-or gate, generally abreviated to

XOR gate. The evidence was that shift registers

and exclusive-or gates are standard over-the- counter components. Any electronic shop will

supply them. What do they do? A shift register is

a device which has one significant and useful

electronic property. This is what a shift register

is. If it does not do what it is supposed to do it

is not a shift register.

What a shift register does is this: it can

hold in what might be regarded as eight cells - in
each of those cells it can have either a one or a
zero. You can start off by putting some ones and

zeros in it in any pattern you like but then when

the shift register is stimulated what happens is

that if you could look into it you would see all of

the ones and zeros shift leftwards one place. The

one on the left-hand end disappears. The slot on

the right-hand end is then empty. It can be

replaced with something else. But if it were

possible to watch what happens in a shift register,

one would see with each incoming voltage transition

that stimulates it the whole array of eight digits
shifting leftwards one place.

If it does not do that it is not a shift register and it does not need to be programmed in

order to achieve that. It is made and sold with

that electronic characteristic.

So, an exclusive-or gate is, in some ways, a

little more subtle. An exclusive-or gate is again a standard electronic component. It has got three wires going into it - actually, two going in and

Autodesk(2) 84 18/4/91

one coming out. The two going in can be called the

input wires, and the one coming out can be called the output wire. What the exclusive-or gate does is to compare the contents of the two input wires.

If one of them, but not both of them, is a one,

then it will produce a one at the output - and I

am, of course, using a one metaphorically. If both

of them is a zero - both of the inputs is a zero -

then it will produce a zero at the output. If both

of them are ones, then it will again produce a zero

at the output, and it is called an exclusive-or

gate because it will only produce a one at the

output if one, but not both, of the inputs is a

one.

Now, the way these devices are linked together

in order to enable the AutoCAD lock to do what it
does, the exclusive-or gate takes its two inputs

from the sixth and seventh cells of the shift

register; and because the shift register keeps

shifting its contents when stimulated the contents

at the two input wires of the exclusive-or gate

will change from time to time. The output of the

exclusive-or gate is fed to fill in the empty gap

which is left at the right-hand end of the shift

register when its contents move, and so one has a

sort of recirculating set of digits, but they do

not exactly recirculate because what comes out of

the exclusive-or gate and back into the beginning

of the shift register will be determined by the

special rules that govern the operation of the

exclusive-or gate.

Now, it would be tedious to go through an

illustration of the two things operating in

conjunction, but the fact is, and the evidence is,

that if you keep stimulating the shift register so

that its cells keep shifting their contents

leftwards, and you keep topping up the right-hand

end of the shift register from the output of the

exclusive-or gate, then if you observe any one of

the cells in the shift register continuously - in	this case it happens to be the sixth cell, but it
does not really matter which one you look at - for a long enough time, you will notice that the
sequence of ones and zeros going through that cell begins to repeat itself after the 127th change.

Now, the evidence was that this combination of

shift register and exclusive-or gate - not this

precise combination, but the general idea of

combining an exclusive-or gate with a shift

register in this general way - is known in

electronic literature as a pseudo random number

generator, and it is a perfectly common circuit,

because it has this property that it will produce

an apparently random string of digits~ only

Autodesk(2) 85 18/4/91

apparently random because depending on how long the

shift register is, so you will get a repetition of

the sequence after a certain number of bits has

gone through, and as a matter of interest, for

those members of the Court whom I have not yet
totally confused, the number in the sequence before
the thing begins to repeat itself is directly

related to the number of cells in the shift

register.	The number, in fact is two, raised to
the power of the number of cells that are used.	I
said the shift register has eight cells - it does
have eight cells, but only seven of them are used
relevantly in this device.

So, the number of possible patterns of bits

you will get before it begins to repeat is 2 to

the 7, which means it will repeat on the 128th bit.

If you increase the number of cells you will

increase the number of bits in the sequence before

it repeats.

So, that is the way AutoCAD lock works.

WIDGET.C was also the subject of evidence.

WIDGET.C is a true computer program, we have no

objection to that proposition at all. WIDGET.C, in

its source code form, was said in the evidence to

comprise between 20 and 30,000 bytes or characters

of program. A part of it, which is responsible for

figuring out what bit should come back from the

lock, is set out at page 818 of the appeal books

and it is worthwhile taking the Court to that page

just for a moment.

BRENNAN J:  We do not need to understand how it is that the

transition stimulates the shift register.	I am not
asking - - -

MR BURNSIDE:  I do not understand either, Your Honour. It

is simply, if a voltage is applied or rather a

voltage change is applied to an appropriate pin of the shift register then it stimulates it to do its

thing. It is just a built in function of the
device.	Now, I do not know which pin it is and I

think it does not matter but the fact is that,

provided it is stimulated by a change of current at

an appropriate pin, it will do what shift registers

do and that will inevitably invoke the exclusive-or

gate to do its thing because it is permanently

hooked up.

DEANE J:  Am I right that the content of the 127 series will

be determined by the arrangement of the zeros and

ones in the cell?

MR BURNSIDE:  Not quite. The precise content - it is useful

for this purpose to look at the sequence of 127

bits as if it is a circle, a clock face with 127

Autodesk(2) 86 18/4/91

ones or zeros arrayed round it. The actual content

of those 127 bits, their order between themselves,
will depend on which of the two pins of the shift

register are connected to the input of the

exclusive-or gate. If you decide that you will

have -

DEANE J:  Can I just interrupt you. Does that mean that the

zeros and ones in the shift register are, as it

were, there in a settled order when you get a shift

register. You do not, as it were, load your shift

register yourself?

MR BURNSIDE:  No, Your Honour, that is not quite right

actually. The shift register is, notionally, all

zeros when you get it because there is no current

and the cells have no content. By attaching one or

other of the pins on the shift register to the

earth line you can start them off as having a one

or a zero in them, in its initial state. As soon

as it is stimulated, of course, that will change.

But that was what was called in the evidence "the

starting key". The starting key is the arrangement

of seven ones or zeros in the shift register before

anything happens to it.

DEANE J: Well, that means, what, they are two to the

seventh possible combinations of ones and zeros in

the shift register?

MR BURNSIDE:	Not in it, Your Honour, no.
DEANE J:	Why not?

MR BURNSIDE: There are only seven cells in the shift

register relevantly as it is implemented.

DEANE J: But you can have one of two in each of seven

cells. Why does that not mean it is a combination

possibility of two to the seventh?

MR BURNSIDE: Yes, Your Honour, that is probably right, yes.

DEANE J: Just trying to fit your explanation and

Dr Emmerson's argument together, do we not need to

focus to some extent on the question whether there

was a combination of ones and zeros in the AutoCAD

lock, which has been copied in charging the Auto-

lock, because if there were I could see an argument

that that could be equated to a set of

instructions.

MR BURNSIDE: Well, Your Honour, there are two difficulties

with that. The first is that what we are talking

about not is not the 127 bit sequence as a

sequence, but merely the initial seven bit group

with which the whole thing starts. As I say, that

Autodesk(2) 87 18/4/91

was referred to in the evidence as the key, and the

evidence was that it did not matter which key you

started with, you would still get the same 127 bit

sequence, but it would start at a different point

round the clock face.

DEANE J:	I see.
MR BURNSIDE:	The only way you can change the content of the

whole 127 bit sequence is by choosing different

pins of the shift register as the inputs for the exclusive-or gate. So if instead of pairing six

and seven into the exclusive-or gate, you chose

three and seven, then you would get a different 127

bit sequence. And there is another interesting

thing that arises out of Your Honours question, and

it is this: the evidence was that when WIDGET.C

sends its challenge, it does not know at the outset

what is the initializing key in the lock and the

locks could potentially have a large number of

initializing keys - in practice they only had a

small number, that is to say any one of them would

have one of a small number of keys which were used

in making them. The evidence was that, on sending

its challenge, WIDGET.C does not know which key to

assume, in other words it does not know whereabouts

around the circle the 127 bit sequence, for this

lock, will begin. So, if it gets a wrong answer on

the first try it assumes a different key,

recalculates what the answer would be assuming that

key, and checks the answer again, and if that is

wrong it tries yet another key and it is only when

it has exhausted all available keys and still got

wrong answers that it decides that it will not

continue to operate. So the starting key is, for

all purposes, irrelevant.

BRENNAN J: All that would be needed is that you would

charge at least one of those cells with a one?

MR BURNSIDE: Yes, because if you start them all with zeros,

will never get anything but zeros, but as long as	the exclusive-or gate will put out a zero and you
there is a single one in the seven bits, then
eventually it will get itself into the exclusive-or gate and that will produce a one and so the process
can continue.

Now, I was taking the Court to page 818 of the

appeal book which shows how the relevant bit of

WIDGET.C works. Can I say, in relation to

page 818, that that is just a fragment of WIDGET.C
but it is, undoubtedly, a computer program albeit a

tiny one. It is written in a high level computer

language which is called C - it is simply the

letter C which is the name of a high level

language. If keyed into a computer which contains

Autodesk(2) 88 18/4/91

the capability of understanding the C language that

program is capable of doing what WIDGET.C does.

Of course, before it causes the computer to do

it it will be translated not just into source code

in the machine but will be translated into object

code and we have no difficulty with the proposition

that the object code generated from the source code

would be most likely an adaptation;	it would be a
version of the same set of instructions.	And,

although one would never really have to do this, we

have no doubt about the proposition that the
instructions to the machine, in object code, are

cognate with the instructions which are represented

here in writing in a computer language.

But to say that without more would be to miss

the point because this, as a set of instructions,

we would submit, is not the same as the set of

instructions, if one could identify them, in the

AutoCAD lock.

GAUDRON J: Well, I was just going to ask you that. Do you

say the AutoCAD lock is anything more than an

electromechanical device?

MR BURNSIDE:  No. a moment arising out of something Your Honour asked I want to come to that in more detail in

yesterday and, in fact, several members of the

Court asked about the program which is fed into the

AutoCAD lock. We have a number of things to say

about that, but perhaps I can defer that for a

couple of minutes.

So that is WIDGET.C. What it actually does is

described in some length by Professor Dillon, an expert called by the appellants at pages 291 and

following of the appeal book. I will not take the

Court to that description. It is sufficient to say that it is as complex as you would expect when looking at page 818.

The Auto-key device is the third device that

needs to be understood. The Auto-key device takes

an approach to the problem which is entirely

different to that which WIDGET.C takes, and to that

which is taken by the AutoCAD lock.

What Mr Kelly had done, it is important to

remember, is this: he first of all detected the

fact that every time a transition went to the lock

something would happen from the lock. It was

necessary to have a transition before the lock

would do anything. That was his observation when

running the lock with AutoCAD. Thereafter he did not use the AutoCAD program at all. The evidence

was, and let me say this: the Full Court slightly

Autodesk(2) 89 18/4/91

misstated this. It was not a relevant misstatement

for the purpose of their reasons, but we think it

is important for this Court to understand what

Mr Kelly actually did. What he did then was to

write a program for his computer which would

produce a square wave.

Now, a square wave is nothing but a series of uniform ones and zeros, a common electronic

phenomenon. If looked at in the language of

computers it is 101010 endlessly. That, of course,

gave him the opportunity of looking at the output
of the lock when compared with a regular input.
Instead of the fluctuating random set of
transitions sent by the AutoCAD program he had a

standard series of transitions which were entirely

uniform. So he stimulated the lock that way. He

said, of course, he could have done it by simply

applying a battery to it, but there is no point in

a cumbersome mechanical arrangement like that. It

is just as easy to generate a square wave, send

that to the lock and have a look what the lock does

then.

By sending a square wave to it and observing

on an oscilloscope, which is able to read two
things at once, he read the output from the lock.

So, in visual form on a screen in front of him he has a square wave, which looks like the battlements

of a castle, and below it, occasional increases and

drops in the output of the lock. That is capable

then of being described metaphorically as a set of

ones and zeros in the square wave, and a set of

ones and zeros in the output. But it is only

metaphorical.

By that process, and after some weeks of considering the matter, he noticed that the wave

form that came out of the lock, or if you like, the

series of ones and zeros that came out of the lock

was not endlessly random, but repeated itself, once
he had sent in 127 square waves.	He was then able

to say, "What the lock does is to send out

these 127 bits, each in its turn, and once it gets

through to them it starts at the beginning again."

So he though, "How does one go about doing that?"

and he took what is the simplest, although

certainly not the most elegant way, of solving the

problem.	He got an erasable programmable read-only
memory - an EPROM.	The EPROM has the

characteristic that it can hold digital numbers;

whether they are data or instructions to no

account. Each cell of an EPROM can hold a number. He needed one that would hold 127 numbers, and the

experts were content to accept that to describe an

EPROM as a set of pigeon-holes was a.reasonable

metaphor for what it is.

Autodesk(2) 90 18/4/91

So what he did was to fill the first 127

pigeon-holes with ones or zeros, according to the

sequence that he had identified. He then connected

the thing up so that every time a transition came

to the EPROM it would send out the next bit.	So
each in its turn was sent off as required.	It is a

very simple way of producing what looks like a

random string of bits, provided it is finite in

length, which this one, of course, is. So, there

is no process that is even remotely analogous to

what happens inside the AutoCAD lock. It is simply the 127 bits, all lined up and each one sent off in

its turn.

BRENNAN J: In response to a transition?

MR BURNSIDE: In response to a transition. I have

simplified the matter a little bit; there is a

counter in Mr Kelly's lock. The counter detects a

transition and when it detects a transition it

increases the ROM address counter. In other words

it will initially start at ROM address zero. When

it is stimulated that will be increased to ROM

address one, and when it is stimulated again it

will increase to two, and again it will increase to

three, and on the 124th stimulation it will be at

address 124, and whatever is the digit, a one or a

zero, at that address, will go to the computer, and

it also had a method of detecting when it had gone

through to the 127th address, and when it did it

automatically reset back to the first address.

DEANE J:  Mr Burnside, why would not somebody of Mr Kelly's

experience look at this 127 series and immediately

identify it as the output of an eight cell shift

register on the basis of what you have told us

about it?

MR BURNSIDE: Yes. Well, Your Honour, the reason for that

was that all of the experts agreed that if the task

is to produce a given 127 bit sequence, the number

of algorithms available to achieve that task is

probably infinite, certainly vast, and the number

of ways of implementing any one of those algorithms

is substantial. So, the difficulty, I think,

Your Honour, is that we have focused on three

solutions to the same problem but to do that is to

ignore that there is a galaxy of other solutions.

DEANE J: Well, I am missing something. I thought you said

that an eight cell shift register was something you

could, in effect, buy over the counter in the

hardware store.

MR BURNSIDE: It is.

Autodesk(2) 91 18/4/91

DEANE J:  And depending upon, or ignoring where the series

starts, it will produce the same 127 series.

MR BURNSIDE: Provided it is connected to an exclusive-or

gate and provided the correct pins are connected to the exclusive-or gate input and provided the output of the exclusive-or gate is connected to the

correct cell of the shift register. Absent any of

those, you will certainly get a pseudo-random

sequence, the length of which will be dictated by

the number of cells used, but the precise content
of which will depend on the exact pin connections.

Now, as I say, the combination of a shift register and an exclusive-or gate was said, in the

evidence, to be a standard electronic circuit for

producing apparently random sequences of bits.

But, whilst it is true that one could say, "Ah,

yes, well, I know that that's one way of doing it
and I can think of all sorts of other ways of doing

it", the easiest way of doing it is to say, "I know

the output I want.	I don't need anything
randomizing.	I know the output I want so I will
just store it in a memory device".	As I say, it is

less elegant but it i~ certainly easy and it is

certainly different to other solutions which are

available.

BRENNAN J: 

Would you get a different 127 bit sequence in

the AutoCAD lock if you connected the input to the
exclusive-or gate from cells seven and eight?

MR BURNSIDE: That assumes that cell 8 is going to be used

and that would increase the length of the sequence.

The shift register has eight cells but only seven

of them are actually used in the implementation;

the eighth cell is ignored.

BRENNAN J: Well then, say four and five, if you like.

MR BURNSIDE: All right, if you connected cells 4 and 5 you

would get a different 127 bit sequence.

BRENNAN J: Although those charges in four and five

gradually find their way to six and seven?

MR BURNSIDE:  Yes, yes. Now, Your Honour, I think that is

probably right. If you have two adjacent cells I

think it must be that they will always produce the

same sequence but you could, for example, have

seven and five or seven and four; you could have

different separations between them. But, yes, I

think Your Honour is right; logically that would

have to be so.

Autodesk(2) 92 18/4/91

DEANE J:  I thought you said that they will always produce

the same sequence but with a different starting

point.

MR BURNSIDE:  we have two available differences to consider.

The first is which pins of the shift register are

connected up to the exclusive-or gate. If they are adjacent pins then we will get sequence A - the one

which is found here. Whereabouts around the circle

of sequence A the thing will begin depends on the

key which is hard-wired into the shift register by

the person who makes the lock. If you choose two

pins that are not adjacent then you will get

sequence B or sequence C and up to however many

possibilities are available. And I think,

Your Honour, on reflection, it cannot be two to the

seventh because it would seem that if they are

adjacent then no matter what pins are connected one

will get the same sequence; if they are separated

by one cell, no matter which pins, then the same

follows. So it may be that there are only three or

four possibilities.

But it is important to bear in mind from this

understanding of the device that what happens

happens as a consequence of the underlying

electronics of the devices involved and, in this

context, I would like to try and expose a few

misapprehensions which, quite understandably, arose

yesterday.

Several members of the Court asked questions

along the lines - asking about the program which

was fed into the AutoCAD lock, or the instructions

which are fed into the lock, or go to the lock. In

our submission, to use that sort of language is to misunderstand why the AutoCAD lock works. Nothing is fed into the lock by way of instructions. Given

the shift register and the exclusive-or gate in the

combination which they have, it will produce the

127 bit sequence when stimulated. Absent either of
those devices it will do nothing.	You do not need

to add anything else once you have got the two

standard devices hooked up together. I was tempted

breathed into him. It is a nice metaphor for creation but modern biological learning would say

to think of the Michalangelo's image of The

it is not accurate except at the most abstract

metaphorical level.

So here one can say, if you like, that the

electronic characteristics of the device are a sort

of program which is fed into them by a chip

manufacturing company in America. But to say that,

really, is to misunderstand the objective of the

exercise and we would say that it is simply wrong,

Autodesk(2) 93 18/4/91

as a matter of language, to say that anything is

fed into the AutoCAD lock; once you have got the

device assembled nothing is fed into it.

GAUDRON J: That may be right but is it not possible to say

that the selection of the cells and the selection

of the starting point really is a set of

instructions to the entire device as to how to

respond?

MR BURNSIDE:  Your Honour, we would have real difficulty

with accepting that proposition because to regard
the connection of a piece of wire to a pin, and
that alone as a set of instructions, let alone an

expression of a set of instructions, seems to us to

do violence to language. We would say that it

simply cannot be described in those terms except in

the most abstract metaphorical sense, but the

Copyright Act is not concerned with abstract

metaphorical senses, it is concerned with real
things and what you have to look for is an
expression of a set of instructions. By the same

reasoning, you might, if you wanted, describe the

operation of a cue hitting a billiard ball as being

an instruction to the other ball to move on. But

that, really, we would say, is again a misuse of

language. The fact that one thing follows as a

physical consequence of another does not mean that

the first is an instruction.

There was another misapprehension we thought

that emerged yesterday and I think understandably.

I think, with respect, Your Honour Justice Brennan

asked whether you can print out the instructions in

the lock. Now, that was a good question, because we say that it exposes the problem that faces the

appellants. Justice Deane asked yesterday whether

you could take the lock apart and identify the set

of instructions. You could certainly take the lock

apart and see what wires were connected to what

other wires, but we do not think that that

corresponds to identifying a set of instructions.

You could identify a circuit, but that is a

different matter again. The reason that you cannot

do either of those things is that there is not any

set of instructions in the lock available to be

looked at.	The best you can do with the lock is
describe what it does.	A few minutes ago I

attempted to describe to the Court what the AutoCAD

lock does, when it is stimulated, but to describe

it in its operation is quite different from

identifying a set of instructions

which causes it to perform that operation.

DAWSON J:  But with an ordinary program it is simply a

matter in the end of electrical impulses, is it

Autodesk(2) 94 18/4/91

not, in the way in which you mechanically connect

the apparatus up?

MR BURNSIDE:	Yes, ultimately, but - - -
DAWSON J:	And by talking about instructions you are really

talking metaphorically again, are you not?

MR BURNSIDE:  No, Your Honour, no. WIDGET.C, page 818,

undoubtedly a set of instructions, not readily

understood by lawyers - - -

DAWSON J: That is merely a representation of what the

electrical impulses will do in a verbal form.

MR BURNSIDE: With respect, Your Honour, we would not agree

with that. It is an expression in a language code

or notation of a set of instructions in ordinary

understanding which will cause a computer to

perform a particular function. Now, take the

simplest possible notion of a computer program, add

A and B. We have no difficulty in saying that the

word "add" is an instruction and that the

expression "add A and B" is an instruction to carry

out a particular operation.

DAWSON J: True, but the computer does not hear that.

MR BURNSIDE:  No, Your Honour, but the words "add A and B'',

provided they are in the conventions of one of the

known computer languages, would be an expression of

a set of instructions. Next down the chain of

transformation will be for those written

instructions to be put into a computer. They will

then reside in a magnetic form on a disk and when
they are about to be carried out they will reside

in memory.

DAWSON J: Well, why can you not say the same way, in

relation to the shift register, you are instructing

it to shift one place to the left every time an

impulse is - you are instructing it to shift one

place to the left and the means by which you give

that instruction is through an electrical impulse.

MR BURNSIDE:	But Your Honour, to say that is to confuse the input with the instruction. Of course, you can if
you like, as a matter of language, say that when I
apply voltage transition to the input, I am
instructing the whole device to do its little
routine, but that is the input.  We are not
concerned about the input, because if that were so,
then the subject-matter of copyright is the random
series of challenge strings, and never once has it
been suggested that the random set of challenge
strings that stimulates the lock is any part of the
copyright work.

Autodesk(2) 95 18/4/91

DEANE J:  On that approach, were you correct when you said

that "add A and B" is an instruction, or is the

instruction what you program the computer, so that

the result is that if somebody gives the add A and

B impulses, you produce A plus B?

MR BURNSIDE:  Yes. Your Honour, we think that the

definition of "computer program" requires one to identify a number of layers of transformation of

the subject work. Take any line of WIDGET.C. We

would agree that even in writing any line of

WIDGET.C would be an instruction. It is an

expression of an instruction intended after certain

processes to cause a computer to perform a

particular function.

The next stage of transformation for any

program properly so-called is for it to be put into the computer. That presupposes two things: first,

it presupposes that the instructions as expressed

in writing follow the conventions of one or other

of the available computer languages. Second, it

presupposes that the computer has within it one of
the programs which will enable the computer to

recognize and apply the conventions of that

language.

There exist two sorts of programs to do

exactly that.	They are called "Compilers" and
"Interpreters".	The difference between them is

simple. If you have a computer program written in

C, for example, and you feed it through a

C compiler, what comes out the other end will be a

different form of the same instructions in object

code. That is ultimately the set of instructions

that the computer will act on. The alternative path is to have what is called an "interpreter"

which means that you tell the computer to run this

little program written in C. It will take each

C instruction in turn, put it through the

interpreter and say, "Yes, that means such and such

in object code. I will do that". It is a sort of

compiler on the run, if you like.

DAWSON J:  So the difference is with instructions you

presuppose comprehension and then acting on that

comprehension, as opposed to an automatic reaction

which you would say the AutoCAD is. I mean, if you

prod someone and they react involuntarily that is

hardly an instruction.

MR BURNSIDE:  I agree with the thrust of Your Honour's

comment. Using notions like comprehension

necessarily introduces difficulties in this arena.

DAWSON J:  We are talking in metaphors, even using the word

"instructions".

Autodesk(2)	96	18/4/91
MR BURNSIDE:	Yes, but Your Honour, the important thing to bear in mind is that one can trace.	From the

starting point the written statement of the

instructions, one can trace a logical and

predictable progression through different forms.

It is a progression through forms which the

definition contemplates. To key the instructions

into the computer, we would say, is to convert them

into a different material form.	We do not have a
problem with that.	To compile them from source

code in the computer to object code in the

computer, we would say, is probably to make an

adaptation of them, although for the present

purpose we would not baulk at the suggestion that

that is converting them into another language code

or notation.

But whichever part of the exercise one

concentrates on you can still identify - as long as

you understand what the object code is doing, you

can map one to one from the instructions on the bit

of paper to the instructions which will be executed

by the computer. One can say that although they

are not in the same form and do not have the same

appearance, they are nevertheless the same because

there is a one-to-one correspondence generally

between the source code in writing and the object

code in magnetic form.

BRENNAN J: Is the program this; that given an input and a

desired output and a capacity for processing

information, the program is that which will

transform the input into the output?

MR BURNSIDE: Certainly the program is the thing that

transforms the input into the desired output, yes.

But the fact that you have something which takes what could be called input, and produces what could

be called output, does not lead necessarily to the

conclusion that what happens in the middle involves

a program.

BRENNAN J: That is why I said, "Given a capacity to process

information".

MR BURNSIDE:	Yes.
BRENNAN J:	In other words if there is no processing of information involved, then you do not have a
computer at all, but if you do have that capacity
then the program is what turns that capacity to
account in transforming the input to the output?

MR BURNSIDE: Yes, we would agree with that, Your Honour.

GAUDRON J:  You agree that capacity simpliciter is

sufficient - digital processing capa~ity is

Autodesk(2) 97 18/4/91

sufficient, or does it have to be engaged, does it

actually have to process?

MR BURNSIDE:  No. You have to have a device with the

capacity and a program which harnesses that
capacity to respond to certain input and produce
certain output, and we think that a great deal

hangs on that proposition in this case because we

say that there is no program in the AutoCAD lock

for this reason, that it does not have both a set

of instructions and a device with the relevant

digital information processing capabilities. We

say that the language of the definition makes it

plain that you must have two things. You must have

a set of instructions and you must have a device

which has the capability of processing digital

information.

Test that in the case of this lock. You have

a hard-wired arrangement of two standard

components. Without adding anything more they will

do what the lock does. It is possible to say that

the shift register and the exclusive-or gate have the capability of processing digital information,

although we would resist that, but if you take away

the digital information processing capability, the

electronic characteristics which enable them to act

as they do act, where then is the program?	One
cannot point to anything at all, even the
connection of pins - - -

GAUDRON J:  Not even that?

MR BURNSIDE:  Not even the connection of pins. Because if

the shift register turns out to be a dud, then

nothing will happen. The device not only has no

digital information processing capability, it has

nothing else that you can point to as being a

program intended to harness that capability.

Now, we do not have any difficulty with the notion that the instructions, on the one hand, and

the device, on the other hand, can be physically

united. Your Honour the Chief Justice, in the

Apple case, identified exactly that possibility in saying that the object code held in the ROMs was certainly physically embedded in them but it was

nevertheless separate from them. We agree with

that, we think that is quite right, but to say that

is merely saying that the storage device and the

thing stored are different, that is certainly

right. But in this case, the digital information

processing capability is exactly the same thing as

the instructions. One cannot identify any

instructions beyond the inherent characteristics of

the devices themselves. Take away those

Autodesk(2) 98 18/4/91

characteristics and they cease to have any

capability of processing information.

Now, we say that the definition contemplates

that the instructions should be intended to cause a

device to perform a particular function and the

device is the vehicle on to which those

instructions are carried into effect. But you

cannot have the cause and effect bound up together
as a single entity, and that is what you have in

this lock. So, for those reasons, it is our

submission that the lock does not contain a set of

instructions or an expression of a set of

instructions. It is merely a device that has

digital information processing capabilities of a

character specific to the nature of the devices it

contains. It can do nothing else. It has nothing

else added to it to make it do what it does.

BRENNAN J:  Does the string of impulses that it is generated

by the operation of a WIDGET.C program and which

are the input to the lock, a string of impulses

which have any significance for the operation of

the lock?

MR BURNSIDE: No. In WIDGET.C, so far as the evidence

allows one to understand it, it is just a series of

truly random numbers and the form and content of

those random numbers was never revealed. As seen

by the lock, they are not even random numbers, they

are just transitions. So that, for example,

decimal 5 and decimal 33 will be treated and

understood identically by the lock. We would say

that by no stretch of reasoning could the challenge
strings form any part of the work which is sought

to be protected.

BRENNAN J:  So that there is no quality attached to the challenge strings which was copied by Mr Kelly?

MR BURNSIDE:  No, no. Mr Kelly's device only starts at the

point where challenge strings are coming in. His

device -

BRENNAN J:  Is it material to the operation of the lock what

the challenge string is?

MR BURNSIDE: Absolutely, as witnessed the fact that it

could be produced just as well by randomly applying

a lead to a battery when it is hooked up to the

lock. That will make the lock work. That is one

reason, I might say, why we have real difficulty

seeing the lock as a device with information

processing capabilities because if it produces its

127 bit string merely by having a battery applied

to it randomly, we would have trouble understanding

what goes in as information and, therefore, what

Autodesk(2) 99 18/4/91

happens to it is the processing of information and

we would say that the Full Court was right in
saying that the thing does nothing more than

respond automatically to incoming transitions.

BRENNAN J: If one were to regard the relevant device as a

combination of the computer and the lock and the

relevant program as incorporating WIDGET.C and

whatever else goes with it, was there any part of

that program considered as a composite which was

copied by Mr Kelly?

MR BURNSIDE:	No, Your Honour, we would have difficulty with the idea of the two things being regarded as a
composite.
BRENNAN J:	I appreciate that.
MR BURNSIDE:	The thing that has to be identified is the expression of a set of instructions. If it is once
found that the lock does not contain an expression
of a set of instructions or a reproduction of an expression of a set of instructions then it will
never contain them merely because you add it to something else that undoubtedly contains a set of
instructions.  The cart does not become horse-like
merely by being hooked up to a horse; it just
remains a cart.  The lock will always remain just
what it is, if it does not contain a program, when viewed alone; it does not acquire a program or any
of the characteristics of a program by being
plugged in and put in contact with WIDGET.C.
GAUDRON J:	I still do not understand why, at one level of

generality at least, you cannot say that the lock

was programmed to produce a particular series of

127 responses endlessly;	why you cannot say it was
instructed to do that.

MR BURNSIDE: Well, at one level of generality you can say

it was programmed to do it in the sense that you

can say that animals with instincts are programmed
to follow their instincts.	The difficulty is

whether that formulation takes you anywhere.

GAUDRON J:  The difficulty is in the notion of a set of

instructions, as such, when you are talking to

something which we assume reacts in a programmed

manner. We are not talking about a set of

instructions to somebody who analyses them and

says, as for example, a set of instructions to

barristers. We are talking about a set of

instructions where stimulus produces fixed,

predetermined response.

MR BURNSIDE:  Yes.

Autodesk(2) 100 18/4/91

GAUDRON J:  And so, in one sense, all you have got to do is

see the stimulus as a set of instructions. Now

why, therefore, is not the instruction to the -

whatever the lock looks like - plus the XOR gate

and its shift register, why is not that just a

program as defined?

MR BURNSIDE: Well, extending the analogy, you could say by

flicking on a light switch you cause the light to

come on, therefore the light switch is programmed,

it contains instructions to the light. At one

level of generality you can get away with that but

we would say that it is a level of generality that

misuses the language as the Act uses it.

GAUDRON J:  But the difference is not that. I mean, the

analogy is not that. I am not sure what it is.

And the reason why the analogy is not that is

because of the shift.

MR BURNSIDE:  One still has a stimulus at one end and a

predetermined, automatic, electronically determined

result at the other end. The difficulty, we think,

Your Honour, is this, that there is an a priori

assumption that because a device can respond to an

electronic stimulus by producing an electronic

output that there must therefore be something in

the nature of instructions in the middle.	We
question that assumption.

GAUDRON J: Yes.

MR BURNSIDE:  But even let it be supposed that there is, at

some level of generality, a set of instructions

inherent in the shift register and the exclusive-or

gate - - -

GAUDRON J: Inherent in the selection of the starting point

and the cells.

MR BURNSIDE:  When you say, "the starting point", by that do

you mean the key that is embedded?

GAUDRON J: Yes.

MR BURNSIDE: Well, the way in which that is put suggests

that a choice is the same as an instruction and

that is problematical.

GAUDRON J: Yes, but is not the fact this? This shift

register with its exclusive-or gate, at one level

of generality, operates by producing endlessly a

sequence of 127 - as at one level of generality -

and that is your analogy with turning the light on:

there is no choice, it is on or off; but this one

can produce any of a number of particular but

different sequences, depending on the point at

Autodesk(2) 101 18/4/91

which you start and the cells you select and your

instruction is as to which one to take.

MR BURNSIDE: Well, yes. The question is that there is an

unstated anticipation of the future in that

formulation of it. You say, "In this device you

can choose to connect this pin or that". Until you

connect the pins you do not have the device. That

is not intended to sound flippant, it is a serious

problem.

GAUDRON J: Yes.

MR BURNSIDE:	The device only comes into existence once you have made the necessary number of contacts,
regardless of which particular ones you choose.

GAUDRON J: Yes. Well that may be so, but it can still give

different outputs. Let us assume - I am not too

sure what the number of variations might be. Let

us assume it is something relatively manageable

like five different sequences, there is still

something that selects one sequence rather than

another.

MR BURNSIDE: That is a very abstract metaphor.

DAWSON J: Well, it is the human mind that selects the

sequences, is it not?

MR BURNSIDE: Well, a human mind has made a choice to create

a device that will produce a particular sequence,

that, we would accept, yes. It implements that

choice by building the device, but it does not

instruct the device to produce a particular

sequence by building the device. The device does

not exist until you have made your choices and then

decided to implement it, and so we would have real

difficulty with the suggestion that by the act of

choosing which pins to connect you are giving

instructions to the device which, ex hypothesi,

does not even exist until you have made the choice.

And perhaps it is just a different way of putting

our earlier proposition that the instructions and

the vehicle into which they are carried into effect

are identical.

GAUDRON J:  Does that mean any more than - accepting that a

set of instructions in the definition is, in this

context, a metaphor - - -

MR BURNSIDE:  Yes, it does go beyond that.

MR BURNSIDE: Yes, it does go beyond that.

GAUDRON J: Well, I have not quite finished. Assuming that

is a metaphor, does it really go beyond saying it

Autodesk(2) 102 18/4/91

depends at which level of generality you apply the

metaphor?

MR BURNSIDE: In this case no, because in this case you

have, inevitably, the problem that whatever you

identify as the set of instructions has no

capability of existing independently of the device

on which it is to operate. It cannot at any level

of metaphor or generality be regarded as having a

separate existence.

GAUDRON J: Well, I do not know. You say that, but somebody

might have said - there might have been a bit of

paper somewhere in which somebody said, "Connect

cells 6 and 7 and start at cell 4".

MR BURNSIDE: If that is a computer program within the

meaning of the definition, Your Honour -

GAUDRON J: Yes, well that is the question, is it not?

MR BURNSIDE: It is not possible that that is a computer

program within the meaning of the definition
because that, in our submission, is really in the

realm of ideas. By what means could such a notion

be described as a set of instructions to a device

having information processing capabilities to

perform a particular function.

GAUDRON J: It is an instruction: start moving at cell 4 and

direct the current from whatever happens to be in

cells 6 and 7 to this thing.

MR BURNSIDE: But, with respect, as you formulated it a

moment ago, it is an. instruction to an electronics

engineer to make certain solder connections, it is

not an instruction to a device.

GAUDRON J:	Thank you, yes.
DEANE J:	Mr Burnside, while you are being interrupted, can

I ask you a question right off what you are
currently dealing with?	Why is not the output 127

series itself a computer program or set of

instructions directed to the computer to make it

function?

MR BURNSIDE: First, because all the expert witnesses

accepted that there is a clear distinction between

a set of instructions and output produced by that

set of instructions - - -

DEANE J: But, no, I am not talking about it as output, I am

talking about it as input to the computer.

MR BURNSIDE:  Yes, well they also accepted the distinction

between input and the program and that is a

Autodesk(2) 103 18/4/91

distinction I tried to make good earlier on.	The
difficulty with the general proposition,
Your Honour, is that -

DEANE J: Can I take you back? Is the distinction you are

drawing between the instruction to the computer,

add A to B, which is not a computer program, and

the program that has gone into the computer to

enable it to react to the instruction to add A and

B?

MR BURNSIDE: Yes, but there is also this: that the input to

any computer could not be regarded as instructions

to the computer to perform a particular function.

It may be, at one level of abstraction, regarded as

an instruction to the computer to perform a general

function. So, for example, if I type in part of a

novel into a keyboard, I am in a sense instructing

the computer to put those letters on the screen,

but to start typing in a novel and to say that what

I am doing is programming the computer, or giving

it instructions in the sense used in the Act, would

be a serious misuse of language.

DEANE J:  I follow that, but this is a bit different, is it

not, because here you have the computer programmed

to do things? The programming will not work unless

it is supplemented by the input of the output from

the lock - - -

MR BURNSIDE: With respect, we would not agree with that.

First, it was accepted on the evidence that

Mr Kelly never had access to WIDGET.C never made

any inquiry about what WIDGET.C did -

DEANE J:  I am not suggesting - I am obviously being

obscure. The way the computer is programmed to do

the function requires what is in the computer as

supplemented by what is going in from the lock. That simply in combination charges it so it can

perform the function that it is desired to have

performed.

MR BURNSIDE: Well, perhaps. We are not entirely sure. It

is not clear on the evidence what WIDGET.C did with the input coming from the lock. It appears that it

compared it with another number that it had

generated elsewhere to see if they were the same.

But there is no reason in logic why WIDGET.C could

not operate by merely checking that there was

something coming back from the lock, and then

simply discarding it saying, ttWell, look, there is

something coming back. Let us just assume there is

a lock in placett, because if there is not a lock
there nothing will come back. It could do that
just as well. It would still have its intended

effect, but once one accepts that, then it becomes

Autodesk(2) 104 18/4/91

very difficult to say that the input from the lock

causes the computer, the big computer, to perform a

particular function. It is far too general.

DEANE J: But if the output from the lock could properly be

regarded as a set of instructions or a computer

program, that would be the end of the case, would

it not, as against you, in that the whole purpose

of this Auto-key lock was to reproduce that output,

and if it is properly seen as a set of

instructions, to reproduce that set of

instructions?

MR BURNSIDE:  Yes, at the most fundamental level, what

Your Honour says is right. If the 127 bit

string - - -

DEANE J:  Do not read into what I have said to you that I am

not agreeing with your demolition of what I put to

you.

MR BURNSIDE: If the 127 bit sequence is regarded as a set

of instructions within the meaning of the Act, then

there is no doubt that we have copied it. We

would say just two things: that the case has never

been pleaded or argued that way and there would be

serious procedural difficulties associated with

such a finding -

DEANE J:  Or evidence has never been led directed to that.

MR BURNSIDE: Yes, but in addition, it would not only be the

end of our case. It would be the end of computing

civilization as we know it, because to regard the

output of a device as being a computer program

would be revolutionary in a way that is

inconceivable.

DEANE J: But that, I would suggest, does not do justice to

what I was putting to you, in that I was not

treating it as the output of the device, I was

treating it as the required input to enable the

computer to perform its function.

MR BURNSIDE:  Even so, and I did not mean to diminish the

argument Your Honour was putting, even if it is

treated just as the input, that would be equally

revolutionary. I think Your Honour, yesterday, put

forward the possibility of a person wishing to

write some program to work in conjunction with

Wordperfect and, as a matter of reality, lots of

programs are written to work in conjunction with

large programs like Wordperfect. In order to

achieve that it is necessary to know what precise

form of input is acceptable to the program in
conjunction with which the new program will

operate. If the example Your Honour puts involves

Autodesk(2) 105 18/4/91

infringement in this case, then it would involve

infringement in that case as well, because it would

follow that any analysis of the acceptable or

required input to a program would, itself, amount

to a program, and we would say that there is no

warrant in the Act or in the second reading speech

to justify such a radical departure from accepted

notions of computer programs.

DEANE J:	Thank you.
MR BURNSIDE:	The judgment at first instance attracted a

certain amount of interest in relevant parts of the

press and professional literature. We think a

result of the sort Your Honour suggests would

create a furor.

DEANE J: Well, is that by way of encouragement or

discouragement?

MR BURNSIDE: Discouragement, Your Honour, that is in the

sense that we would not encourage Your Honours to

do it. Whether the creation of a furor is regarded

by Your Honour as something to be desired is

another matter. But we think that the words of the

definition simply would not allow it and, in

particular, we would identify the word
"particular", that is to say the set of
instructions have to be intended to cause the

relevant device to perform a particular function

and data, which is accepted and acted on, as input

cannot be regarded as instructions which cause a

particular function.

Now, I revert to the attempt to locate the copyright work, the expression of a set of

instructions which attracts the operation of the

Act. The appellants repeatedly identify the

program as being the set of instructions set out in
paragraph 5 of their outline. Now, we would say

this, what paragraph 5 of the appellants outline

contains is certainly not a set of instructions.	It does not even rise to the level of being an
algorithm.

BRENNAN J: Could you tell me what an algorithm is?

MR BURNSIDE:  Yes. An algorithm is quite well explained at

page 124 of the appeal book, Your Honour, and it

was explained in these terms: an algorithm is a

narrative and visual statement of a set of

procedures by which a particular problem can be

solved. Now, it is plain, both as it is used in

mathematics and computing and as it is defined in

the dictionary and as it was explained in evidence,

an algorithm must be a sufficiently precise steps

to enable a real solution to be derived from a

Autodesk(2) 106 18/4/91

specific set of variables if one goes through the

steps which are explained in the algorithm.

Now, an algorithm, therefore, although one would not seek to solve problems by working through

the algorithm - at least in most circumstances -

the algorithm can be implemented intellectually to

produce a solution.

The next stage of program development after

the algorithm stage is, generally speaking, writing

a set of instructions in an available computer

language which implements the logical flow of the

algorithm because a computer will not accept an

algorithm in its form as an algorithm.

The difficulty with what is set out in

paragraph 5 is that it is not sufficiently precise

even to constitute an algorithm.	What it is is a
description of the problem.

DAWSON J: At page 818, is that an algorithm?

MR BURNSIDE: No, Your Honour, that is a program; that is

an expression of a set of instructions. I can take

the Court to some algorithms that describe the

logic of the devices in question: at pages 924 to

929 of the appeal book.

DAWSON J:	I see.
MR BURNSIDE:	Now, let me make it clear, the algorithm is

the logical solution or the logical method by which

a solution is derived. A flow chart which is a

common enough thing is simply a visual

representation of the algorithm. But the algorithm

really is the logical structure of the solution.

At page 924, what you see is an algorithm which

describes the processes which are implemented in a

few lines of the code of WIDGET.C found at

page 818.	You will see that we have the words at
the top:

SET KEY TO 100.

To the right of that, you have the expression:

key= 100;

"Key= 100;" is an instruction in C language and it

can be located, if not on page 818 then in another

representation of WIDGET.C.

Then:

SET i TO=KEY -

Autodesk(2) 107 18/4/91

is the description of the next logical step.	The
words to the right of it:

for (i=key -

is an expression of that step in a computer

language and is thus another instruction, and so

the process goes right down to the end and you will

see that the expressions on the right, which we

would accept without question, are instructions

within the meaning of the definition. They

implement the logic which is described in the

algorithm on the left of the page.

If I can take the Court to page 927: this

represents both the algorithm, or logical flow, of

a part of WIDGET.C and the instructions by which it

is implemented. The Court will see, looking at

page 818, the words "static char mightbe", "static

char sequence [16]" and so on. Those words are
also found on page 927 and by following the
algorithm set out on page 927 you can follow the

way in which the logical approach to the solution

has been implemented in C code at the top of
page 818.

Now it is plain to see that the logical description of the solution contained on the

left-hand side of the pages that I have taken the Court to is far more precise than the description
of events contained in paragraph 5 of the

appellants' outline and we would therefore say that

paragraph 5 of the outline, so far from being a

computer program, is not even an algorithm but is

rather simply a description of a general approach

to the solution of a problem. The difficulty that

that entails for the appellants is this, that it

was made plain in the second reading speech and in
the explanatory memorandum that the amendments to
the Act were not intended to protect ideas or
algorithms but just the particular expression of the set of instructions by which algorithms were implemented.

Now, if paragraph 5 does not rise to the level

of an algorithm, it clearly cannot be within the

scope of protection intended by the amendments.

Can I take the Court briefly to what was said

in the explanatory memorandum? Paragraph 16, when

referring to the phrase "expression .... of a set of

instructions", it says:

The phrase "expression .... of a set of

instructions" is intended to make clear that

it is not an abstract idea, algorithm or

mathematical principle which is protected but

Autodesk(2) 108 18/4/91

rather a particular expression of that

abstraction. The word "set" indicates that

the instructions are related to one another

rather than being a mere collection.

I should perhaps say, in passing, following on

Your Honour Justice Deane's question, that that

would be another difficulty with regarding the

sequence as a program because they are not

relevantly related, one to another.

Whilst I am talking about it, too, I think it

was Your Honour who, yesterday, suggested what you

referred to as the simple example of the program

which simply produces the alphabet in sequence. It

is interesting to observe that compared to the

AutoCAD lock, that is a relatively complex example

because by ordinary conventions, a single letter of

the alphabet would be expressed in eight bits and,

therefore, to express the alphabet in bits would

need 208 bits, not 127. So, not a great deal

follows from it, but it is nice to put these things

in perspective.

The second reading speech has also got

something to say about the matter of algorithms. Perhaps I can just refer the Court to pages 2421 to 2422 where very similar language is used. At

the very end of 2421 the Attorney-General said:

The term 'computer program' is also defined in some detail. Without going into

all aspects, let me just mention that the

definition is intended to make it clear that

abstract ideas or procedures for solving

problems are not protected but only particular

expressions of them. the program may be

written in any language, code or notation and

may include related information. It may be

expressed in a high level language

intelligible to humans, such as FORTRAN,

COBOL, ALGOL or BASIC. Equally, it may be

expressed in a lower level or machine language

such as is obtained after compilation or

assembly and as may be directly suited to the

operation of a computer.

So, in our submission, when tested against a

proper understanding of the definition of "computer
program", what the appellants have identified in

paragraph 5 is not a computer program. What it is,

is a broad description of the function which a

device, the lock, will perform when it is supplied

with a set of instructions, but the function

performed is not the set of instructions, and a

description of the function is not a computer

program.	The instructions, in our submission, are
the cause and the function is the effect.	To

Autodesk(2) 109 18/4/91

identify the function of a program, we would

submit, is to do no more than to identify the idea

which the program is intended to implement.

We base a substantial part of our case on the proposition that copyright law does not, and never has, protected ideas but, rather, protects the form

in which the ideas are expressed and it is plain,
in our submission, the 1984 amendments did not

intend to dislodge that fundamental tenet of
copyright law.

The difficulty here, as in all cases of this sort, is to distinguish between idea and expression

but, in our submission, for the reasons already

advanced, one can equate the function which a

program is intended to achieve with the idea of the

program, and the code or other method of

implementation as being the expression of that

idea.

In seeking to distinguish between idea and

expression where the subject-matter is something as

arcane as computers, we would say that there are

three significant milestones.

At the starting point, one identifies the function which it is desired to achieve. In this

case it would be the sequence of 127 bits which it

is desired to create. Next, you describe the

fulfilment of that function in some logical

sequence, typically in the form of an algorithm, and the evidence in this case was that there was

probably an infinite number of algorithms which

would produce the 127 bit sequence. There may be

some hyperbole in that, but every expert agreed

that the number of possible algorithms to produce

this 127 bit sequence was incalculably large.

Once you have described in the form of algorithm the way in which you will achieve your

purpose, you must then write a set of instructions

which will cause a computer to perform the

particular function you desire. It is only when

you get to the third stage that we would say there
comes into existence a work which attracts the

protection of the Copyright Act.

So it is a set of instructions which gives

effect to an algorithm which is a computer program.

Anything before that point is in the province of ideas. It should be possible to say that on the

facts of this case the solution is therefore very

easy because on any view, the method employed by

the AutoCAD lock, if described in the language of

algorithms, is quite different from the method

employed either by WIDGET.Corby the· AutoCAD lock.

Autodesk(2) 110 18/4/91

And if the algorithms which describe what they do are different, then it follows inexorably that the

set of instructions, if ever identified, which

implements that algorithm will also be different.

We, of course, say that even if you look at

them and describe them in the language of

algor~thm, the next step was never taken. There

never was the expression of a set of instructions,

but merely a direct implementation in hardware

which does not necessary involve the creation of a

computer program. But that is a separate question

from the present part of our submission.

BRENNAN J: Can I interrupt you for a moment there? Could

you program a computer to recognize a transition
when received from some external impulse, or by way

of some external impulse?

MR BURNSIDE: Yes, and in fact that is commonly done really

in all programs - almost all programs that are used

by people from day-to-day rather than more obscure

things like car ignition systems. There is a

commonly used routine which continues to look at

the keyboard to check whether any key has been

depressed. The keyboard is constantly scanned to

see whether anything has happened there and if

something has happened then what has happened will

be taken in as input and then reacted to. Now,

whilst that would not necessarily be treating - in

fact it probably would not be treating what happens

as a transition but rather treating it as a binary

number. It is equally possible to scan the

keyboard and only identify transitions regardless

of what number they also represent.

BRENNAN J:	Or you could have a program which would search binary numbers to detect transitions?
MR BURNSIDE:	Yes, it would be simple.

BRENNAN J:  And then you could have a program which would, on recognizing a transition, move through 127

different numbers and you could put that program

into an EPROM.

MR BURNSIDE:	Yes.
BRENNAN J:	Would you then have copyright protection for the

EPROM?

MR BURNSIDE:  Your Honour's example would also need a

microprocessor which would fetch and carry out each

of the instructions we have been talking about,

each in their turn, then, yes, the answer is that

the expression of that set of instructions would be

a computer program and would attract copyright

Autodesk(2) 111 18/4/91

protection, but it would not prevent another person

from setting about to create another program or

another device that would create the same output.

BRENNAN J:	Yes.
MR BURNSIDE:	And could I illustrate that point by reference

to an example that emerged several times yesterday?

Log tables: a common example or illustration of the difficulties when science and copyright come together. Clearly, by publishing a set of log

tables you do not prevent anyone else from

generating log tables or using logarithms. What

they cannot do is to copy the printed work. Let it

be supposed that what the AutoCAD lock did was

highly complex; let it be supposed that it

contained a microprocessor capable of doing very
complicated things; let it be supposed that what

it did was to send, each in their turn, each

successive element of the standard Napier log

tables by calculating the logarithms of each

successive number from zero through to ten in

increments of 0.1, it would be perfectly easy to

implement such a program.

So let it be supposed then that that is what

the AutoCAD lock set out to do and inside WIDGET.C

there would be a corresponding routine which

calculated each successive logarithm of each

successive number, compared them to see if they

were the same and then, with or without irrelevant

side issues, it would then allow the processing to

continue. Suppose Kelly intercepted this and,

having a great memory for his schoolboy arithmetic,

recognized the succeeding expressions in a log

table.	Does that then prevent him from writing a
program which will generate a log table?	We would

say that it could not, it simply could not, and

there is no difference, in principle, between

generating a log table as the desired password and

generating an otherwise meaningless 127 bit string,

except that the 127 bit string is vastly less

significant in its size and complexity.

Now, we think that if the Court agrees with

writing a program to generate log tables, if that proposition is right then this appeal must fail.

the proposition that the use of log tables in

Now, I was inviting the Court to understand the distinction between an algorithm on the one

hand and the code which implements the logic of the

algorithm on the other. At some point after
algorithm one finds the transition from idea to

expression of idea. The American courts have had

to deal with the idea expression dichotomy in the

Autodesk(2) 112 18/4/91

context of computer programs, in the case of

Whelan v Jaslow, and we would take the Court to

that decision.	I take the Court to page 1235 of
the report, in the left column.	I am referring to
the report of the case in 797 F 2d 1222.	The

heading where the relevant passage starts is ''A

rule for distinguishing idea from expression in

computer programs". At page 1235, the court said:

It is frequently difficult to distinguish

the idea from the expression thereof. No less

an authority than Learned Hand, after a career

that included writing some of the leading
copyright opinions, concluded that the

distinction will "inevitably be ad hoc."

The court then went to look at Baker v Selden

and then on page 236, at the top of the page, said

this:

In deciding this point, the Court -

in Baker v Selden -

distinguished what was protectible from what

was not protectible as follows:

(W)here the art (i.e., the method of

accounting) it teaches cannot be used without
employing the methods and diagrams used to

illustrate the book, or such as are similar to

them, such methods and diagrams are to be

considered as necessary incidents to the art,

and given to the public.

Applying this test, the Court held that

the blank forms were necessary incidents to

Selden's method of accounting, and hence were

not entitled to any copyright protection.

The Court's test in Baker v Selden

suggests a way to distinguish idea from

expression. Just as Baker v Selden focused on

the end sought to be achieved by Selden's

book, the line between idea and expression may

be drawn with reference to the end sought to

be achieved by the work in question.	In other
words, the purpose or function of a

utilitarian work would be the work's idea, and

everything that is not necessary to that

purpose or function would be part of the

expression of the idea.

And there is reference to one of the American Apple cases -

Autodesk(2) 113 18/4/91

Where there are various means of achieving the

desired purpose, then the particular means

chosen is not necessary to the purpose; hence,

there is expression, not idea.

Now, translating that to the circumstances of

this case, the purpose, plainly, is to produce a

device which will generate the 127 bit series. If the AutoCAD lock did not do that then it would not

work as the lock. Likewise, if the Auto-key lock

did not do that it would not work. The evidence

was and a comparison of the two locks demonstrates

that there are many ways of fulfilling that purpose

or performing that function. The function, it is

submitted, is the idea of the work and is not

protected. The way the function is achieved is the

subject of protection.

It then is necessary to say, well, is there an

expression of that function, or of the set of

instructions that implements that function, in the

AutoCAD lock? For reasons already advanced, we say

there is not, but, even if there exists something

in the AutoCAD lock that can be described as an

expression of a set of instructions that caused the

device to perform the particular function,

production of the sequence, it is plain, on looking

at the two locks side by side, that Mr Kelly's lock

does not copy that method of expression. He has

performed the function in a different way and, in

our submission, that is an end of the appellants'

case. This problem was dealt with in the judgment
in the court below, formulated as the question,
"What is the relevance of function in determining

objective similarity?". We say, first, that the

objective similarity test is still one of the two

tests of reproduction or infringement by

reproduction and, with respect, we adopt what

Chief Justice Gibbs said in the Apple case.

Well, how does one determine substantial similarity in the case where a computer program is	the subject-matter. First one has to identify the
work which attracts the protection of the Act. In
our submission, the function of the work is no part
of the protected work. Function is the idea and is not protected. Accordingly, when seeking to
determine whether there are substantial
similarities between the protected work and the
alleged infringing work, it is illegitimate to look
at function, because to do that is to look at
something which is outside the scope of protection and something outside the scope of protection, we
would say, cannot be of assistance in determining infringement.

Autodesk(2) 114 18/4/91

It was plain, in all of the judgments below,

that the court accepted that, apart from function,
the two locks had no other similarity and as a

consequence of that it is our submission that there

is no similarity at all, function being an

irrelevant consideration once one accurately

identifies function as being outside the work which

is protected.

The court below equally held that there was

nothing but function in common between WIDGET.C or

WIDGET.C and the AutoCAD lock combined, and

Mr Kelly's lock, With respect, we agree with that,

but we would say this: that not even the function

of WIDGET.C is the same as the function of the

Auto-key, and that is for this reason: the function

of the Auto-key and the AutoCAD lock is to generate

the 127 bit string. WIDGET.C does not have to

generate the 127 bit string at all. All it does -

its function - is to look at the incoming string of

bits and way, "Well, is this the one that I

expect?", and in order to do that, what it does is to check which number ought to be the next number, but it does not generate them at all, and it

certainly does not produce the 127 bit string as

output because WIDGET.C, relevantly for this

inquiry, does not produce output, except when it

says to whatever other part of the program is

involved, "You had better stop". That part of the

program is not in evidence and is not involved in

the case against us.

So, if one accurately identifies the function of the AutoCAD lock and the Auto-key lock as being

to produce as output the 127 bit string, it is

apparent that that function is not the same
function as WIDGET.C's function but, in any event,

it is plain that the way in which WIDGET.C does

what it does is quite different from the way in

which the Auto-key works and accordingly there is

no similarity, for any relevant purpose, between

the two. Now, once the Court comes to the conclusion
that function is not part of the work which is

protected, then a real question does arise, whether

it can be said in this case that Mr Kelly derived

the Auto-key by use of the protected work, and we

would say that the answer to that question is,

"No", because Mr Kelly did not relevantly use the

protected work; what he did was to use the
function. All he identified was the function; he

set the thing going to see what it did. In other

words, to see what output it produced, and he did

that without even using WIDGET.C because, as I

tried to explain earlier, he wrote a program that

generated a square wave that would simplify the

Autodesk(2) 115 18/4/91

task of detecting the pattern in the lock's output.

So, all he did was to harness the lock's function

in order to make his device, and to harness the

lock's function is not to use any part of the

protected work.

Now, in seeking to dislodge the finding of

each of the judges in the Full court, our learned

friends relied on a number of cases concerning
plays and the translation of novels into pictures

and the like. We would say, first, that reference

to cases in which novels have been translated into

musical" and we think that any reference to the transference of ideas from one dramatic form to another as being in the outer reaches of

plays is unlikely to be helpful in this context.

metaphorical abstraction so far as computer

programs are concerned.

The fact is that those cases do not help, in

any event, because there one is comparing a

protected work with an alleged infringing work.

The details of plot and character are within the

protection of copyright and so to say that a play

contains all the details and incidence of character

and so on as a novel, when looking for substantial

similarity, is legitimately to compare the

protected work with the infringing work. But to

look here at the function of the two, to elevate
the repeating nature of the output to the status of
a wraparound feature is simply to look at things
that are no part of the protected work.

But in addition, we would say, that those cases are unhelpful in this context because, in any

event, the legislature has resorted to deeming

provisions in order to make it clear that things

which are quite different in form will nevertheless

be regarded as versions of one another. So

section 21(3) gets over difficulties that might

arise in transformation from two dimensions to

three, or vice versa, by deeming one to be a

version of the other. And the definition of

"adaptation", so far as it relates to

transformation between dramatic and non-dramatic

forms and the like, gets over the problem that

might otherwise arise in determining whether or not

one thing is a version of another in ordinary

language.

Now, that leads readily to the related

question: if one cannot find any similarity between

the Auto-key and any program identified by the

appellants, that leads to the conclusion that there

is no reproduction.	What about the question
whether there is an adaptation?	We would say, that

Autodesk(2) 116 18/4/91

it is impossible to assert the two things which

have no similarity, can be a version one of the

other. It is wrong as a matter of language and it

would be a nonsense as a matter of law. If in this

case the Court decided that function is not

relevant to the process of comparison when seeking

to identify reproduction, but then said, "Well let

us have a look at the function to see if one is an

adaptation of the other", that would simply allow a

monopoly on ideas by an indirect path where the

direct path was closed. Once one gets to the

position that all the two things have in common is

their function, then it becomes impossible to say

either, that one is a reproduction of the other, or

that it is an adaptation of the other.

We cannot conceive any possible circumstance

in which language could properly be used in saying that the Auto-key lock is a version of the AutoCAD lock, any more than one could say that a mouse is a

version of an elephant. They have nothing in

common except function, and function is not part of

the protected work.

Now the appellants say that if their appeal

does not succeed the amendments in 1984 will be

demonstrated to be unworkable and useless and no

doubt civilization, as we know it, will be

seriously impaired. They also say that it will be

necessary, in every case, to examine minutely the

workings of the mechanisms involved. We say that

neither of those consequences follows. The reason

for that is that in almost every case it will be

possible to identify a program which is the subject

of the claim for protection.

This case is right at the outer reaches of any

possible application of the Copyright Act for the reason that the device which is said to have been

copied cannot readily be called a computer program.

Put to one side whether, on one analysis of things,

it can be said that the AutoCAD lock contains a

program, it is clearly a program in a most unusual

form. It is not a program in the form that the

Attorney-General appeared to have in mind in the

second reading speech where he referred to the

problem in the amendment sought to address as being

that software programs are hard to write but very

simple to copy, and clearly what the

Attorney-General had in mind, for the generality of

cases, was that programs would be found contained

on disks. In that form, there is no doubt, they

are easy to copy.

If there is a program in the AutoCAD lock, it

could not be said that it is a program that is easy

to copy, except by dismantling it, breaking it

Autodesk(2) 117 18/4/91

apart, identifying the electronic components and

putting together a similar arrangement of

electronic components. But we think, and we submit

that that is not what the Attorney-General had in

mind, and it is not what the amendments were

directed to.

So in most cases that arise under the Act the

Court will be confronted with a computer program in

the ordinary understanding of that word, and it

will be possible to identify a listing in source

code, perhaps an algorithmic description of the

problem, almost certainly a copy of the program in

object code, and one would also have an alleged

infringing work, most likely in object code form.

In those circumstances, the application of the

Act will present no difficulties. There may be a

question involved whether the sets of instructions

are substantially the same if there is not a one to

one identity between the protected work and the

infringing work, but that is a special case. In

ordinary cases, therefore, we say that the failure
of this appeal will not have the least bearing on

the efficacy of the 1984 amendments.

Unless there are any matters that I have left

unclear, those are the submissions I wanted to make

on the principal ground of the appeal, the one that took up most of yesterday, and I would then turn to

the matter that occupied the last few minutes of

yesterday, but with some hesitation.

If was difficult, given the brevity of our

learned friend's submissions, to understand quite

how the matter of authorization was being put. As

we understood it, the case put was this: in order

to run AutoCAD the program, the big program, you

have it on a hard disc in your computer and when

you invoke the program, part of it gets lifted into

random access memory and its instructions will be

carried out. It is suggested that that is an

activity which requires the licence of AutoCAD.

That is a serious question which raises some quite

difficult issues.

But our learned friend conceded that there was

another question, and one which we say is logically

anterior, and that is whether any infringing

behaviour of that sort had been proved. He

conceded that no infringing behaviour of that sort

had been proved in evidence, but merely invited

that an inference be drawn that there would be

infringing behaviour; that is to say, that there

would be running of the AutoCAD program by people

not licensed to run it. That was an inference

which was not drawn by the learned trial judge and

Autodesk(2) 118 18/4/91

which was not drawn by the Full Bench of the

Federal Court, and we would say that this Court should not be burdened with the task of trying to go through the evidence in order to see whether

there is an available inference of infringing
behaviour of the sort suggested in order to get to

the much more difficult and general problem of

whether, in any event, running a program is an act

comprised in the copyright.

If the Court wishes to embark on that

difficult and interesting question we are happy to

assist, but we do see that there is a logically

anterior question which presents serious problems for the appellants in that they are flying in the face of the facts found in the courts below.

BRENNAN J:  So flying in the face of the facts found or that

there were no facts found relevant to this in the

first place?

MR BURNSIDE: Well, the finding that there was no proved

infringement. The problem arose in this form -

this, I should say, goes well beyond the

submissions made yesterday and, really, goes back

to what is found in the judgments in the court

below. It was said, in the court below, that in
order to run the program it was necessary to be

licensed to do so. That is the general question of

law.

The evidence which was accepted at first

instance and in the Full Court was that all of the

people to whom Auto-keys were sold were licensed to

run AutoCAD, but the question was, "What was the

term of that licence?". Was there to be implied

into that licence a restriction that you could only

run AutoCAD, the program, as long as the AutoCAD

lock was in place rather than any other lock?

Now, that is purely a question of fact and it

the appellants before they get to the question of	is one which is necessary to resolve in favour of
principle, namely, whether in fact a licence is
necessary at all when you run a program.

Now, we would invite this Court to say that

the findings of the court below should not be

disturbed, that is to say that there is not any

restriction in the licence which AutoCAD users get

which prevents them from running AutoCAD with some

other lock than the AutoCAD lock. It is purely a

question of fact. It is partly a question of

whether or not terms should be implied into the

written licence agreement which AutoCAD uses.

Autodesk(2) 119 18/4/91

DEANE J:  But is that really putting the effect of the

evidence accurately. I thought that the effect of

the evidence was that the Auto-key lock would

enable the unauthorized use of copies of AutoCAD?

MR BURNSIDE: Clearly, the use of Auto-key would enable a

person to use an unauthorized copy of AutoCAD, that

is true.

DEANE J: But is that not the advantage of the thing, that

you only have to buy one and you can then make and

use copies by simply getting the lock?

MR BURNSIDE:	Your Honour, that is the way the case for the appellants was put at the trial but the learned
trial judge rejected that on the facts
because - - -
DEANE J:	Can you direct us to the relevant findings of

fact?

MR BURNSIDE: Yes. Whilst that is being got, Your Honour,

could I explain this: the evidence was that a lot

of people who use AutoCAD, people who have

purchased a proper copy of it and use it in their

work, operate from more than one location.

Typically they have an office and they work at home

as well, like barristers and judges, and the

difficulty that the AutoCAD lock creates for them

is that every time they want to pursue their work

at home they have to undo the printer from the back
of the lock, undo the lock from the back of the

computer, take it home, reinstall it and then start

working, and witnesses for both sides agreed that

that was a nuisance.

DEANE J: What, the AutoCAD people will not sell an extra

lock to people who want to do that?

MR BURNSIDE:	No, they will not, but what they do, both in the way the product is sold by their dealers and in
the material that they enclose with the program, is
that they say you can make as many copies of the
program as you like and you can copy it into as
many computers as you like.  What you are not
allowed to do is to run two copies of it at the
same time. That was the restriction which is
expressly imposed both at the point of sale and, if
it become legally effective, in the documentation
that is sealed in with the program.

What the Auto-key lock enables a person to do

then is to have a lock on his computer at the

office and a lock on his computer at home and he

can then use one or the other without having to

carry the lock about. There is real benefit in

that because there is not only the inconvenience

Autodesk(2) 120 18/4/91

associated with removing and reinstalling the lock

repeatedly, but the evidence also was that there

was a risk of damage to the connections of the lock

by repeated plugging in and unplugging and, of

course, there is the risk of losing the lock. And,

the evidence was that AutoCAD would not replace a

lost lock in any circumstances; you had to pay for

a new program - $5,000. So the lock, acquired by

that fact, has substantial intrinsic value as well

as a real nuisance component. The legitimate use

of the Auto-key, on the evidence, was to prevent

exactly that irritation and that risk of loss.

Now, a further challenge was mounted below -

which we do not understand to have been pressed

here - which was that well, even allowing that

there are legitimate uses for the Auto-key lock,

you are enabling purchasers to use it

illegitimately as well and there are two ways in

which that might happen. The first is that they

will be very naughty and they will operate both

their computers at the same time. But, of course,

that is not authorization within the meaning of the

word as defined in Moorhouse's case because, whilst

it makes it possible, it happens at a time when

there is no possibility of supervision or

prevention.

The other suggestion was, well, perhaps these

locks will be bought by people who have acquired

copies from a licensed user but who themselves do

not have a licence. And there was no evidence, and

the court below accepted that there was simply no
evidence of a purchase of the Auto-key lock in

those circumstances. In fact the evidence was to the contrary, that each time an Auto-key lock was delivered to a purchaser, they would seek to test the lock - not surprisingly - and the way they

tested the lock was to remove their AutoCAD lock,

install the Auto-key lock and see whether the

program still ran. But, of course, the fact that

they had an AutoCAD lock demonstrates, indirectly,

that they are an authorized user of the program

because otherwise they could not have an AutoCAD

lock at all.

So, by reason of that evidence, the court declined to find that there was any authorization

of infringement by the ways just suggested. And we

would invite this Court simply to say that it will

not seek to disturb that finding.

BRENNAN J:  Can I just interrupt you there for a moment?

Was it common ground that if a person other than

the purchaser had run AutoCAD with the Auto-key,

then that use, by a person not being the purchaser

Autodesk(2) 121 18/4/91

of AutoCAD, would have amounted to an infringement

of the copyright?

MR BURNSIDE:  No. That is the question of very substantial

interest and difficulty which would arise if the

facts had been found otherwise but, we say, plainly

the question does not arise in this case because
the facts are not there to give rise to it and it

is for that reason that although, no doubt,

Your Honours would be delighted to have a second

technical issue to consider, we would urge the

Court not to embark on consideration of that

problem in the absence of a sufficient factual

foundation.

There was a variation on that theme as put

below, and again it is one which has not been

articulated before this Court, and it was,

well -

DEANE J: Well, are you not best not articulating it?

MR BURNSIDE: For the last 10 minutes I have been fishing

for a suggestion like that, Your Honour. I must

say, we have real difficulty in meeting submissions

that have not actually been put to this Court,

although we have some recollection of what they

are, or were in an earlier hearing, but - - -

MASON CJ:  Mr Burnside, you might refer us to the pages in

the judgment of the trial judge where these

findings of fact are made upon which you rely, as

an answer to this submission.

MR BURNSIDE: 

Yes, pages 992 to 993.

BRENNAN J: 

And could your junior also identify the part of

the pleadings in which this aspect of the case was
being made against you?

MASON CJ: 

Where is this dealt with in the judgments of the Full Court?

MR BURNSIDE: 

The matter was put before the Full Court in a

somewhat different way, and that was the gloss that

I was about to go on to. In the Full Court there

really was not a challenge to the learned trial

judge's finding that there had been authorization

of infringement by supplying Auto-keys to people

who were not licensed users of AutoCAD. So the

finding at 992 to 993 really was not challenged,

although we understood that that was the aspect of

the matter referred to in submissions yesterday
afternoon.

The way the matter was argued in the Federal Court was rather more subtle and, trying

Autodesk(2) 122 18/4/91

not to do any injustice to it, it was this: first,

that running AutoCAD requires a licence; second,

AutoCAD was acquired by users who were licensed;

third, the licence is to be understood as

restricting the use of AutoCAD to a use when the

AutoCAD lock is in place, albeit that no such term appears anywhere in the documentation, or anywhere

in the circumstances of sale; fourth, that the

Auto-key lock enables a licensed copy of AutoCAD to

be run without the AutoCAD lock in place. That

circumstance involves a reproduction of the program

in RAM whilst it runs in breach of the implied

term.

Now, the Full Court said that there was no

such implied term and accordingly, the question did

not arise whether, running a program by a licensed

user, involves a reproduction in material form.

And that is the question of general interest that

might one day have to be resolved, but which, we

say, does not arise here for the reason that the

factual foundation is not made out.

Now, there is no question of general interest

involved in saying whether or not the contract

contains an implied term of the sort contended for.

It is trite law, what is the test for the

implication of terms, and thus far the appellants

have simply failed to get any court to find that

such a term was to be implied, and we would invite

this Court to say that it is not interested in

hearing argument on that aspect of the matter because the facts are against the appellants.

MASON CJ: Well, Mr Burnside,· we are minded to think at the

moment that in reply Mr Emmerson can have the

opportunity of explaining, with a little more

clarity perhaps, what his attack upon the judgments

of the court below is, and we will reserve to you,

if we think it appropriate, an opportunity to

respond to what Mr Emmerson says.

MR BURNSIDE: Yes. In those circumstances then, unless I

have left any matters unclear, those are our

submissions.

MASON CJ: Well, I mean there may be matters unclear, but I

am not sure that that entitles you to address us

any further.

MR BURNSIDE: If there are any matters that I can clarify, I

would be grateful to have the opportunity.

MASON CJ: Only you would know the answers there, of course,

to that, Mr Burnside. We will hear from

Mr Emmerson at this stage.

Autodesk(2) 123 18/4/91

MR BURNSIDE: If Your Honour pleases.

MR EMMERSON: If the Court pleases, the parties are, of

course, at issue on the two matters that I set out

in paragraph 2 of our outline of argument. It is

our submission, in general terms, that my learned

friend in his submissions misapplies the
distinction between an idea and the working out of

the idea in a copyright work and also misapplies

the distinction between a work and the mechanism

used to implement the work.

However, in order to reach that point, my

learned friend put to the Court a number of
propositions of fact, some of which we would
challenge. These dealt with propositions as to how

the various programs and the various locks operate.

I should make it clear that on the issue of

infringement, we say that both limbs of the test

for infringement are satisfied to enable this Court

to conclude that the respondents have reproduced in

a material form the relevant program in the

computer. They have also reproduced in material

form the relevant program in our lock.

I should make it clear that we say that if

either of these two are made out, then it is our

submission that we should succeed in this appeal.

It is not necessary for us to succeed in both.

My learned friend, however, directed very

detailed submissions to the way in which he says

that the appellants' lock works. Now, our broad

response to that is that it confuses the

distinction between a work and the mechanism used

to implement or reproduce the work.	He
concentrated on the mechanism.	He, as it were,

took it apart and said of each bit, "Well, clearly,

that bit is not a computer program". This is an

incorrect way to approach the matter, in our

submission, where you are dealing with something

which is, as our lock is, a hard-wired piece of mechanism because, of course, ex hypothesi, the program is wired into the device.

However, that does not conclude the matter.

The program was wired into the device in the Apple

case.	The programs there were not programs which
could be put in and put out at will.	They were

part of the ordinary functioning of the computer in

that case. The only difference is that in that

case the Court was concerned with computers as

ordinarily understood, whereas here, we are

concerned with a statute which refers to a device

having digital information processing capability.

Autodesk(2) 124 18/4/91

BRENNAN J: Apple was concerned with the manufacture of a

computer.

MR EMMERSON: It was concerned, with respect, strictly

speaking, with the importation. It was the

computer itself which was imported and the argument

for Apple in that case was that because the

computer itself had within it as a permanent part

of itself memory, and that memory was preprogrammed

with the program in question, so this was said to

be infringement by importation and sale of - - -

BRENNAN J:  The strict question was whether or not the

manufacture of the silicon chips that were in the

computer, that were imported, would, if they had

been manufactured in Australia, have amounted to an

infringement of copyright.

MR EMMERSON: That is correct, yes, and they were something

which were permanently in that state.

BRENNAN J:  It was not concerned with the operation of the

computer but with its manufacture.

MR EMMERSON: 

Yes. Equally, in the present case, the

question is whether the manufacture of the
respondents' lock and the sale of that lock in

Australia infringes.  Now, that lock, just as the

Wombat computer in the Apple case, has a memory in

which there are, we say, instructions. And so the

effective question, in each case, raises whether

the manufacture and sale of the device with the

program in it amounts to an infringement.

DEANE J: 

What if the locking device had been to program the

computer so that it would only work if the

electricity current was at 13 volts for one minute,

followed by 17 minutes for a second minute, back to

13?

MR EMMERSON:  Yes.

DEANE J:  Would you say that anyone who manufactured a

transformer which produced that electrical current

was in breach of the copyright in your computer

program?

MR EMMERSON:  The particular example that Your Honour gave

to me does not involve a device which processes

digital information. It simply involved a device

which received alternating current - - -

DEANE J: - - - Which reacts to - - -

MR EMMERSON:  - - - and itself changed the voltage.

Autodesk(2) 125 18/4/91

DEANE J: It might even have to have a little computer in it

to do it on alternate minutes if you did not wind

it up.

MR EMMERSON:  On that the question may well turn. But, to

take Your Honour's example as initially formulated,
the mere fact that without, so to speak, anything

coming in from the outside to the inside - it

switches voltage every few minutes - would not make

it a device with digital information processing

capabilities within the Act and so the answer would

be, "No, that would not infringe".

Now, having said that, one could imagine cases in which this result was achieved by a computer

program within the meaning of the definition in the

1984 amendments and, if this occurred, then, again

depending on whether the other test was satisfied,

that could well be an infringement, we would say.

Now, in dealing with the lock produced by the appellants which, as I say, is one way in which we put the case but it is not the only way in which we

put the case, my learned friend made a number of

statements which we would wish to correct or

comment on or challenge. One of the statements he

made was that one could duplicate the effect of

signals going to the lock by taking a battery and a

switch and turning the battery on and off.

Now, we say that that may well be the case but, of course, that does not answer the question

because by turning a.battery on and off one is, in

fact, sending digital information. This is exactly

the way digital information in the form of Morse

code was sent by classic telegraph operators. The

possible use of the battery, just like the possible

substitution of other devices to do the same thing,

cannot, we say, be decisive in this case. All that

it establishes is that the respondents were able,

if they had so wished, with a battery, to duplicate

signals which could have been sent out by a

computer.

Now, so far we would not challenge that, but

that does not mean that the device does not have

digital information put into it, it necessarily

does, and the whole notion of a battery where you

are switching the thing on and off at random

entails the proposition that digital information is

being put in. The question, however, is how the

lock processes that digital information.

Now, in describing the nature of our lock, my

learned friend proceeded by saying, "Well, it

really only contained a shift register and an

exclusive-or gate and each of these are things

Autodesk(2) 126 18/4/91

which you can buy in the shop". Now, we accept

that you can buy shift registers in a shop and you

can buy exclusive-or gates in a shop just as you

can buy computer memories in a shop. However, in

the case of our device, the way that it is

programmed is the specific way in which it is wired up. Now, this has a number of components to it and

it is the combination of that which makes up the

particular way in which the program is implemented

in this device.

In the first place there is a starting point

which involves a choice of a specific seven-digit

number. Now, as to that, my learned friend was

inclined to discount the importance of this and he

did so on two different grounds. One is, he said,

"Well, all this determined was where in the

repeating sequence of 127 bits you started." Well
now, granted that that is correct taken in

isolation. It involves first the proposition that

you have the device so set up that it does generate

that sequence in repetitive form, and secondly it

involves the proposition that it does not matter

where you start. Now, for the functioning of the

program, it does matter how you start, and the way

we have put our case throughout is to say that what we are concerned with is the particular combination

that I set out in paragraph 5 of our outline of

argument, but there is nothing in that which says

it does not matter where you start. If you do not

start in the right spot then you will not get the

same numbers being produced in response to the

input data by the lock as you would by the program

itself, and so the system will not work.

BRENNAN J: Is that so?

MR EMMERSON: Well, yes, and this is the second point I want

to challenge my learned friend on, because he made

one qualification which, I think, does not appear

in any of the judgments, but he said, "Well, if it

starts at the wrong point, the program will try

again". Now there was some evidence that there

were either two or three different possible

starting points, but we would challenge my learned

friend's assertion that it was clear that whatever

the starting point was the device would still

function. We would say that was not a finding of

fact made by any of the courts, and it was not a

finding of fact which was properly open.

BRENNAN J:  Does that mean that you could not take an

AutoCAD lock off one computer that was running

AutoCAD and put it on to another?

Autodesk(2) 127 18/4/91

MR EMMERSON:  You could, for this reason: that one part of

the wiring which sets the program in the AutoCAD

lock is that the starting number is permanently

wired in position, so that there are wires attached

between the relevant lines and the relevant keys,

so that when you start the device, for a particular

lock you always start in the same spot - you do not

just start at random.

BRENNAN J:	How is it that Auto-key is able to work?
MR EMMERSON:	The way that the respondents' one works is

that that sequence of numbers is part of the

sequence of numbers which is actually stored in

their memory, and so they have taken the same

sequence of numbers and they have stored it in the memory of their device and it is put out in order.

BRENNAN J: But it matters where it starts.

MR EMMERSON:  But it does matter where it starts, and their

device starts at cell No 1 and goes on to cell No

127, then wrapsaround and goes back to cell No 1

again.

BRENNAN J:	And does yours always start on cell No 1?
MR EMMERSON:	Ours always starts at the same spot, yes, but

you see -

BRENNAN J: Well, I asked you if did it start at cell No 1?

MR EMMERSON: But, you see, cell No 1 is a reference to the

particular way in which the program is stored in

the respondents' device. In our device, the

program is stored by being hard-wired and that has

various aspects which I am seeking to develop, but

one of them is that you permanently wire a set of

wires to the shift register, so that when you turn

the device on, it starts off automatically with a

sequence of seven digits.	Now, those
seven - - -

BRENNAN J:  The same sequence?

MR EMMERSON:  Yes, Your Honour, always the same. Now, that

sequence of seven digits is the first sequence that

gets put out be the device because what happens is

that the shift register shifts through, digit by

digit, moving to one side and in place of one that

has been moved out, it puts in the result that

comes in from the exclusive-or gate. Now that

means that the first seven digits permanently wired

into our shift register, are the first ones that

come off and then further ones come out.

Autodesk(2) 128 18/4/91

Now, the way the respondents' device stores

the same information is simply in a memory, but

that is just another sort of memory unit and in the

first seven of the cells, the memory unit, it has

the first seven digits that are put out and this is

necessarily so because it is essential to the
respondents' underlying purpose in this, is that

they should produce a device which produces exactly

the same result as our device and so they cannot

change from one number to another; they have to

take our precise choice in this regard.

So WE say that part of the program which is

wired into our lock is one which is wired in by the

starting point which gives you the initial numbers

in the seven cells in the shift register. Then

there are further choices involved. There is the

choice of using only seven cells of the shift

register rather than eight or six or five, some

smaller number. There is the choice of which cells

you use for the input for the exclusive-or register and which cells you use for the output and, indeed,
whether you use cells at all rather than, for

example, putting incoming data immediately into the

exclusive-or register.

Now, the result is that you have a lot of

choices and as a result of the specific choice of
the way the system is wired up so, we say, the

program is wired into the device. It is therefore,

in our submission, quite wrong to say that, as it

were, all you have got is an exclusive-or gate in a

shift register. If that was all you had that, in

itself, does not give you a preprogrammed device;

it is the particular way you wire up the various

components to each other, the various components to

the fixed input and output lines that gives you

your programming.

In our submission, what my learned friend

really is seeking to do here is by slicing the

thing up into two small bits and ignoring some of the major bits to say that there is, accordingly,

no program here. We say there is a program here.

GAUDRON J: But you still have the difficulty, if you say

that, of identifying what is the set of

instructions and it really does depend, at the

level of generality, that you identify those

instructions as to whether there is anything like

reproduction or adaptation.

MR EMMERSON: Well, at the level of generality that I used

for much of my address in-chief, I took the set of

instructions as being sufficiently summarized in

paragraph 5 of our outline of argument.	Now, I
would stand by that.	My learned friend said,

Autodesk(2) 129 18/4/91

"Well, it is not a program." In our submission, it

is a sufficient description of the program for the

purpose of explaining what we say are the set of

instructions. It is this: the particular program

chosen has two aspects. It receives incoming

digital information and detects whether each digit

particular direction then the program steps on to the next stage.

is the same or different from the preceding digit.

Now, that, we say, is the first part of the

instructions. Next: at each stage the program puts

out a digit which may be nought or one according to

a specific predetermined sequence. The sequence is

chosen so that it is essentially impossible for an

outsider to duplicate it by guesswork, and so on.

GAUDRON J:	The problem really is the words "put out" because that suggests that that might really be its
function and, at a more specific level, its
instruction might simply be "start here".

MR EMMERSON:  I want, in the course of this reply, to try to deal with the function point. We say that you

cannot ignore function in considering whether there

is sufficient objective similarity, and when one is

talking about specific function at the level that

we say is required by the working of the 1984

amendments, we would simply say that you are, to

that extent, protecting function.

DEANE J: But if said S(b) to a computer that could speak

back to you it would say, "It is no use saying that

to me, if you want me to do that you have got to

program me to do that".

MR EMMERSON:  No, with respect, what it would say is, "Put

it into a language code or notation with which I am

familiar".

DEANE J: That is the same thing, work out a computer

program.

MR EMMERSON:	But if you look at "computer program", it is:
an expression, in any language, code or
notation, of a set of instructions (whether with or without related information) intended,
either directly or after either or both of the
following:
(a) conversion to another language, code or notation -

et cetera -

Autodesk(2) 130 18/4/91

to cause a device ..... to perform a particular

function.

DEANE J: So, you would work out the computer program and

then you would have copyright in that computer

program and if somebody used that computer program,

in breach of copyright, you would have a good

action but if somebody used a completely different

computer program to get instruction 5(b) you would

not have an action.

MR EMMERSON: Well, with respect, Your Honour, we would

challenge that. I do not want to be taken to be

saying that S(a) and 5(b), which were of course

prepared as a description of the program, are

themselves in the sort of language one would use
for computer programming, but I am saying that
those are the effect of the instructions which we

say are relevant in this case.

Putting that, however, to one side, we say

that that does state the essence of the

instructions which are found in WIDGET.C and found

in the lock made by the appellants and by the lock

made by the respondents and that the difference

between them is not that these instructions are not

being carried out, because they are, but the

specific way in which they are implemented in

hardware.

The real question is whether one can use the old idea expression "dichotomy" to say, in effect,

that you simply cannot have any protection for
instructions of this form. It is only for one of

the infinite numbers of different ways that you

could apply it to specific hardware. And it is

there that we say, well, if that is the law then

the effect of the 1984 amendments is worthless

because one can always make the thing at the level

of the actual operation of the machine look

different.

But that, we would say, is not the purpose or

effect of the definitions that were introduced into

the Act in 1984, and that it is quite wrong to

assume that anything which is specific as what we

have set out in S(a) or (b) is merely an idea. You

could say that the idea might be to set up a lock

which receives information and sends it out. You

could say that an idea might involve a checking of

the precise operations carried out within the

computer and within the lock against each other.

That is an idea, but if you work it out into a

particular way that you do it and a specific

sequence, then we would say that it would make

nonsense of copyright law to say that that is no

more than an idea. You might equally well say when

Autodesk(2) 131 18/4/91

you have refined a literary work down to the stage

when you have a sonnet that that is only an idea

and that the real problem is how you put on paper

the various words in the sonnet.

BRENNAN J:  But the idea that you have for the wiring up of

this would be incorporated in engineering drawings

in which copyright would subsist, and there is no

doubt that these are constructed according to those

drawings, no doubt the lock is constructed in

accordance with appropriate drawings.

MR EMMERSON:	No doubt it is.
BRENNAN J:	And once constructed, can you change its

function by any instructions that you give it?

MR EMMERSON:  You could change its function by changing the

manner in which the various components are wired

together, yes.

BRENNAN J: Quite, but my question to you is that if it is constructed in accordance with the drawings, once constructed, can you change its function by giving

it any instructions?

MR EMMERSON:  One has to answer that at two different

levels. As regards the program that we are talking

about, if you wanted to change the program you
would have to change the wiring. Just as if you
wanted to change the way that the Apple computer
programs in the Apple case were there, you would

actually have to change the wiring, assuming you

could do that, because in each case you have a set

of instructions which are permanently fixed into

the machinery of the program device.

At the second level, but not as I apprehend

it, the level which Your Honour had in mind, one

could no doubt think of the incoming stream of data
as being instructions. They are certainly things

which lead to the device performing a particular

function, but that is always variable. That is not

something which is permanently wired into the

device.

MASON CJ:  Mr Emmerson, we will adjourn now until 2.15.

AT 12.46 PM LUNCHEON ADJOURNMENT

Autodesk(2)	132	18/4/91
UPON RESUMING AT 2.17 PM:
MASON CJ: Yes, Mr Emmerson.

MR EMMERSON:  If the Court pleases, returning to what we say

in reply on the matter of facts, so far as the

appellants' lock is concerned, the program is

hard-wired and the manner in which that particular

program, rather than any one of the large number of

other possible programs, is inserted is, first,

there is a hard-wiring in of the first seven

numbers to the output, and these are put into a

storage unit called a shift register. Secondly,

the input into the exclusive-or device involves a

choice of where the input comes from; we choose two

particular cells, the others would give different

results. Third, the output of the exclusive-or

device can be put to any of the cells or anywhere

else which would be useful, and again that will

lead to different results. So, by the choice of
the manner in which the various components are

wired together, one hard-wires the program into the

computer, and that is what we say occurred in this

case.

Now, there are two matters which my learned friend raised which we would challenge. One is the

suggestion that it does not much matter where you start in the sequence of 127 digits. We say that

that is not established by findings of fact, nor is

it established on the evidence. It is important to

know, if you want to copy this device, which are

the numbers with which you start and the only

flexibility is that the program is adapted so that

it can be used with any one of- it was either two

or three locks - I was not able to turn up the

precise reference in the evidence over the luncheon

adjournment and it is not mentioned in any of the

judgments.

So far as what must be received back; my

learned friend suggested that one might be able to

receive back from the lock a variety of different

signals and the device would still continue to run

AutoCAD, we would say that that is incorrect. It

is clear that the precise and exact number must

come back and I would refer the Court, in this

factual context, to some of the findings of the

trial judge.

In dealing with the various numbers that can

be said, the learned trial judge said, at page 956

in the paragraph beginning at line 17:

Autodesk(2) 133 18/4/91

The number of sequences that can be

created by the use of the 127 digit system is

enormous being 2 raised to the power of 127.

The actual response to AutoCAD depends upon a

transition, but the response must be correct

every time and the response is checked by

AutoCAD by reference to the number chosen to

constitute the challenge. A number is

transmitted back to AutoCAD. That number is

compared by AutoCAD with the number that

should have come back and if correct the

challenge has been met.

So, it is a comparison of a precise number which

comes back from the lock. That is compared with

the number generated within the program itself in

the computer and those two must be precisely the

same. The need to have a precise check is also set

out by the learned trial judge in the passage at

pages 953 to 954 under the heading "How the AutoCAD

Hardware Lock Operates". What His Honour says is:

Written into the AutoCAD program is a set

of instructions described as Widget C.

Widget C is contained in the source code and

the object code and thus is contained in

AutoCAD when stored in the hard disk within a

computer. Widget C comprises instructions to

the computer requiring it to send out what has
been described as a challenge to the AutoCAD
lock. On receipt of the challenge, the lock

processes the challenge and transmits its

answer to the computer. The computer receives

the answer, checks it against its own

instructions, and if the answer is correct,

continues to execute AutoCAD. If the answer
is not correct, the computer ceases to run.

Then further down the page in the paragraph beginning at line 17 on page 954, His Honour sets

out:

The basic form of WIDGET.C is a 127 state

machine table which generates a series of l's

and O's.

Those are the digits that have been already

referred to many times in argument. His Honour

then deals with a vast number of possibilities that

you could get and with the wraparound feature and

he deals with the sending out of a challenge and

the result that comes back.

Now, in our submission, none of this supports

any suggestion that one can have back a whole

variety of responses from the lock and the program

will still continue to run. It is clear that His

Autodesk(2) 134 18/4/91

Honour's findings were specific and were to the

contrary of that. You must get a precise number

back and that is, of course, what we are claiming

here. We are not claiming a general principle in

which any number will work; we are claiming however

that we have protection for a specific device in

this case.

As to the similarity of the respondents'

program, His Honour says this at page 961:

With the help of computer programs developed

by himself, Mr Kelly prepared a source program

consisting of a series of numbers based upon

the repeating pattern of the AutoCAD lock
operation. It was a tabular representation of

the pattern.

So His Honour treats quite correctly, in our

submission, the essential part of the respondents'

program as being, in fact, what he had copied

directly from the AutoCAD lock. His Honour

continues:

It was the program to be used for the

preparation of a look up table to be coded

into an EPROM (erasable programmable read only

memory).

And then he says the reason for the use of the

memory and so on.	He deals with the 127 digits and
the wraparound feature.	At the bottom of page 962
going over to 963:

The EPROM contained in the Auto-Key lock contains a set of digits in hexadecimal notation in the first 128 consecutive memory locations but the result is the same as if the

digits being l's or O's as set out under the D

heading in the program prepared by Mr Kelly

were used. The sequence is identical to that

contained in the original state machine table

contained in the AutoCAD program and used in

the AutoCad lock shift register.

In our submission, that is a correct statement

of the facts. And then, as to the working of the

respondents lock, His Honour quotes a passage from

Dr Forward, again on page 963 beginning:

"The Auto-Key lock works in a similar way to

the look-up table that is used in the AutoCAD

programme which runs from the main memory of

the computer.	The look-up table is programmed
into an EPROM -

Autodesk(2) 135 18/4/91

and goes on to deal with the manner in which it is

stored.

Now, in our submission, the position is that

both in WIDGET.C and in the respondents' program

you have the critical matters encoded into a

look-up table. It is the same in each case. The

difference, so far as the appellants' lock itself

is concerned, is that while some of those numbers

are hard-wired into the lock the remaining numbers

are generated.

Now, that brings me to the matter of

hard-wiring. My learned friend tended to suggest

in his submissions that the 1984 amendments were

not intended to cover the case in which hard-wiring

of computer programs occurred. Now, as to what I

mean by hard-wiring, I would refer the Court to

what was said by Dr Forward on that point at

page 87 of the appeal book beginning at line 3.

There is also a transcript page number of 154 at

the bottom of the page but it is page 87 in the

appeal book, and Dr Forward is asked:

Are programmes ever hardware?---Yes,

programmes can be implemented in hardware.

After the programme has been written, it

depends on how it is treated after that what

the eventual implementation of the programme

is. Complicated programmes normally would not

be implemented as hardware, but fairly simple

programmes are sometimes implemented as

hardware because the hardware can operate at a

much faster rate than a software programme

being interpreted by a computer can operate.

Now, the question then is whether that type of

hardware programming is intended to be excluded

from the definition of "computer program" that is

to be found in the Act. In our submission, it is

not to be excluded. The definition reads:

"computer program" means an expression, in any

language, code or notation, of a set of

instructions (whether with or without related

information) intended, either directly or

after either or both of the following:

(a) conversion to another language, code or

notation;

(b) reproduction in a different material form;

to cause a device having digital information

processing capabilities to perform a

particular function.

Autodesk(2) 136 18/4/91

Now, the definition is silent about what is

the actual chain of causation in this case.

Clearly, as a matter of general logic, instructions

as such do not cause a device to perform a

particular function. One has got somehow to alter

the state of the device so that that function will

be performed.

In our submission, the chain of causation can be, as with a use of a home computer, that the

operator might type the program in on the keyboard.

That is one possible chain of causation. But

equally, a possible chain of causation is the one

that you have where programs are hard-wired into a

device which has information processing

capabilities. Where that occurs, then there is

still the relevant causation because one is looking

again at the causal chain of events without saying

that one must find something perceptible in the

same way as the original computer program was

perceptible in the device which performs the

function.

That brings me to another aspect of

hard-wiring. My learned friend submitted that the

set of instructions could not be physically part of

the device. We would resist that. There is

nothing in the definition of "computer program"

which says that the set of instructions cannot be

physically in the device. While one might say that conceptually one can distinguish between the device and the instructions, that is a conceptual and not

a physical distinction.

It is somewhat analogous to the case that one

has in designs law in which you talk about a design

being applied to a physical object. For the

purpose of analysis you draw a conceptual

distinction between the design and the physical

object, but it does not follow that you can look at

a jug made according to a particular design and

say, "Well, part of it is the design and part of it	is the jug". In the same way, it is our submission
that you cannot necessarily look to an object which
has been programmed in a particular way and
separate out physically as distinct from conceptually part of that object and say, "That
part is the object and the other part is the
program".

The next point that we would make deals with

this: my learned friend, in a number of places, put

submissions which appear to assume that the 1984

amendments included various matters which, in our

submission, they do not include. For example,

first; the 1984 amendments do not require that the

instruction should be in writing or other visible

Autodesk(2) 137 18/4/91

form; quite the contrary, as I attempted to show.

Secondly, the 1984 amendments do not require that the device having information processing capabilities must be or include a microprocessor.

That is not part of the definition and, in our submission, is not required. Third, the 1984

amendments to not require that the device should

have a compiler or something similar to a compiler.

that is something which is possible, but it is not

required by the amendments. Fourth, in dealing
with the various steps that can take place between

the devising of a computer program and the device

performing a particular function, there is nothing

in the definition of "computer program" which says

That you must be able to find a one-to-one

correspondence between some instruction, written
out somewhere, and something in the device. The

definition does not import that notion of a

one-to-one correspondence.

So, in our submission, much of my learned

friend's submissions on these aspects of the case

depended on reading into the definition of

"computer" the notion of "computer" as it may

ordinarily be understood in common speech, and the

notion of "program" as it may ordinarily be

understood in common speech.	We say that that is
simply not correct.

I want next to deal with the matter of the

algorithm. In our submission, in the way that my
learned friend's submissions were put, the
introduction of the idea of an algorithm simply

confuses the issue and we would note this: an

algorithm is not something which is mentioned in

the Act. There is no requirement anywhere in the

Act that a computer program must either be an

algorithm or must have been derived from an

algorithm.	The Act is entirely silent on that.
The explanatory memorandum deals with algorithm only for the purpose of drawing a

contrast - I have just lost the immediate

reference, but it draws a contrast between abstract

ideas and algorithms on the one hand and says,

well, as such they are not protected. Now, in our

submission, that casts no light at all on the

problem before this Court in the present case.	My
learned junior has pointed out that the relevant
paragraph is paragraph 16 in the explanatory
memorandum:

The phrase "expression .... of a set of

instructions" is intended to make clear that

it is not an abstract idea, algorithm or

mathematical principle which is protected but

rather a particular expression of that

Autodesk(2) 138 18/4/91

abstraction. The word "set'' indicates that

the instructions are related to one another

rather than being a mere collection.

We accept that it is not an abstract idea, an

algorithm and mathematical principle as we would

understand it being examples of abstract ideas

which are protected, but that simply does not

answer the question because my learned friend then

went on to say that the program described in

paragraph 5 of our outline of argument could not be

a description of a computer program because it was

not even a description of an algorithm.	Now, in
our submission, it does not have to be a

description of an algorithm; it has to be, as we

say it is, a description of a computer program and
it is precisely because we were not concerned, in

paragraph 5, with writing down any algorithm or abstract idea but we are concerned with writing down what the relevant instructions are that we did

not include an algorithm in paragraph 5.

BRENNAN J:  Does one find, anywhere in the appeal book, a

set of instructions expressed being the set of

instructions for which protection is claimed?

MR EMMERSON:  Yes, the passage to which my learned friend

drew attention from WIDGET.Chas just that.	The
particular page that my learned friend drew

attention to, which I think was at page 818 of the

appeal book, sets out a set of instructions and we

refer to those as setting out the relevant

instructions when expressed in the C language.

BRENNAN J: All of them or only part of them?

MR EMMERSON: This page sets out all the relevant

instructions.

BRENNAN J:  And then some?

MR EMMERSON:	I do not think it sets out anything else, no.
BRENNAN J:	So these are the relevant instructions for which you claim protection?
MR EMMERSON:	I think the short answer is, yes, Your Honour.
BRENNAN J:	And therefore we must look to see what it is

that infringes that set of instructions.

MR EMMERSON:  We must look and see what it is that infringes

our copyright instructions, yes. This is the form

in which the instructions appear in source code - I

was going to say, in the computer, but in the

program which, when compiled, goes into the

computer.

Autodesk(2) 139 18/4/91

BRENNAN J:  Were these instructions, or any part of them,

applied to the lock?

MR EMMERSON:	We would say, yes.
BRENNAN J:	Were they applied otherwise than by hard-wiring?

MR EMMERSON:  No. The way in which these instructions work

on this page is, of course, directed to a computer

and so soft-wiring but the way in which they were

applied to the lock was by hard-wiring.

BRENNAN J: Well then, the hard-wiring that was effected,

was not the consequence of these instructions?

MR EMMERSON:  Yes, it was, because, as His Honour found,

these instructions in the computer and the hard-

wiring of the lock come together. They must be

identical, and that is the whole purpose of the

thing.

BRENNAN J:	What must be identical?
MR EMMERSON:	What must be identical is what a device

programmed with these instructions does when an

input string of digits comes in.

BRENNAN J: That is, it must produce the reciprocal

function.

MR EMMERSON:  Yes. and they must be identical in each case. It must produce specific results, yes,

BRENNAN J:	And you do not have any difficulty with the notion of the distiriction between function and
program?

MR EMMERSON: There is bound up in the notion of "computer

program" a notion of functionality. It is true, of
course, that the word "function" does not mean the

same thing as the word "program", but functionality

is, of course, part of the program, yes, and once

it gets to the stage, we say, of being as specific

as this is, then that is what is protected.

That brings me to the point of adaptation. I

had put in my submissions in-chief that it was of
assistance in understanding what the word

"adaptation" means in the context of computer

programs to look at the meaning that had been given

to the word "adaptation" in other parts of

copyright law and I pointed out that, in those

examples, one might well find that something was

protected as an adaptation, notwithstanding that it
did not have the very thing which had initially

given rise to the subsistence of copyright and the

examples I gave included an adaptation by conveying

Autodesk(2) 140 18/4/91

the story of a literary work in pictures only. A

literary work, initially, of course, traditionally,

acquired copyright by reason of the fact only that

it involved letters, it was in print or writing.

Yet the ambit of protection, when one starts to

look at an adaptation, goes beyond that. Similarly

with the adaptation of something in dramatic form

to non-dramatic form and vice versa.

There was a third example that I gave which my

learned friend criticized and I want to come to,

and that is the example drawn from section 21(3) of

the Act because my learned friend seemed to suggest
that this, in a relevant sense, was merely a resort

to a deeming provision and not something which cast

light on the meaning of the word "adaptation".	We
would challenge that because what section 21(3)
says is this:

For the purposes of this Act, an artistic work

shall be deemed to have been reproduced:

(a) in the case of a work in a two-dimensional

form - if a version of the work is produced in

a three-dimensional form; or

(b) in the case of a work in a

three-dimensional form - if a version of the

work is produced in a two-dimensional form;

and the version of the work so produced shall

be deemed to be a reproduction of the work.

The Court will note that that subsection does not

deem something to be a version if it were not

otherwise a version. What it does is deem

something to be a reproduction, but it gets there

by way of saying that a version is deemed to be a

reproduction. Now, it is the word "version" on

which we rely and it is the word "version" which is

also a word which appears in the part of the

definition of "adaptation" which deals with

"computer programs".

So, we would stand by our submissions on

section 21(3). We would say that where something

acquires artistic copyright by reason only of it

being a drawing, then something which is in three
dimensions and which ex-hypothesi cannot be a
drawing is still a version of that drawing for the

purpose of the Copyright Act.

This, in our submission, makes it clear that

where one is talking about a version of a computer

program not being a reproduction of the work, in

our submission that makes it clear that Parliament

is taking one beyond the rather narrow literal one-

for-one comparison of instruction for instruction

Autodesk(2) 141 18/4/91

that my learned friend contended was necessary in

order to make a finding of infringement in this

case. We say that the use of the word "version'' in

the definition of ''adaptation" makes it clear that

in computer copyright law, as in the other branches

of the law that I have mentioned, you look to see
what is valuable in the work and that is what is

protected; you do not look merely to see what was the thing which led to copyright subsisting in the

first place.

The next short point that I wish to make is

this:	my learned friend said that our submissions
in this case would lead to a monopoly on ideas.	In
our submission, that is clearly not so. We are
claiming here not a monopoly on any idea in the abstract, nor are we claiming any monopoly on the notion of having a lock, or even a lock which
carries out a particular type of arithmetic or logical operation.

What we do say, however, is that where one

has, as in the respondents' case, a lock which is

programmed to carry out precisely the logical

operations which we say are the essential part of
our program, then that is clearly a reproduction of
the copyright work, and if that is wrong it is in
any event a reproduction of an adaptation of the

copyright work.

I would next wish to take the Court to a few

references in cases and textbooks which we say are

relevant here. First, it is important to

recognize, in our submission, that the causal link

required for a finding of infringement may be

direct or indirect, and that is, of course,

relevant if one is looking at infringement of the

program as it exists in WIDGET.C by the making of

the respondents' lock. This is dealt with in a

passage in Copinger which, I think, was not on our

original list of references but, if I may, it is a

short passage.	I will give a reference to it and
read it out to the Court.	It is paragraph 461 in

Copinger, Twelfth Edition, 1980 and it is headed

"Indirect Copying".	What the learned authors say
is this:

Copyright may be infringed by copying

something which is itself a copy of the
plaintiff's work. Indeed, in most cases of

alleged infringement, the copying is done

indirectly, the plagiarist never having seen

the original manuscript, engineering drawing,
dress designer's sketch and so on, only the
published work, piece of furniture, dress, or

whatever, derived from the original work.

Autodesk(2) 142 18/4/91

Pausing there, we say that that is, of course, what

has happened in this case and it is not to the
point for my learned friend to say that the

respondent, Kelly, never had access directly to

WIDGET.C. There is nevertheless the sort of

copying contemplated by the Act under the heading

"Indirect Copying".	Now, the learned authors
continue by saying:

However if, notwithstanding there is a "chain"

from the defendant's work indirectly to the

plaintiff's work, all that is copied is the

idea of the plaintiff's work, then there is no

infringement.

And much argument has been directed to this Court

on that, and they continue:

But if the original work has been reproduced,

it is no answer to say that it has been copied

from a work which was itself, whether licensed

or unlicensed, a copy of the original. This
is so, even if the intervening work is of a
nature not capable of being an infringement of

copyright or enjoying copyright, or even if

there is no intervening work, the link being

an aural description.

Pausing there, it is therefore irrelevant to the

question of whether the respondents have infringed copyright in the program as it appears in WIDGET.C,

to ask whether or not my clients' lock would, if

made and sold by the respondents, have been an

infringement. It is·on that part of the case,

merely an intervening step and it does not matter

whether our lock is capable of being an

infringement of copyright, or of enjoying copyright

because all that is relevant is that it forms part

of the chain of causation.	The learned authors
continue, saying:

The latter point underlines the risk that,

even if an independent designer is employed in

an attempt to avoid infringement, it may be

necessary to give him so much information that

the work he produces is still an

infringement -

and there is a reference to the case of Solar

Thomson Engineering v Barton, (1977) RPC 537.

If I may, although it is not on our list, I

would wish to give the Court a reference to one

relatively recent case in which this type of

copying occurred where the intervening step was

certainly not itself something which was capable of

Autodesk(2) 143 18/4/91

protection; that case is Winstone v Plix

Products Ltd, (1985) 5 Intellectual Property

Reports at 156. It is a decision of the Court of

Appeal of New Zealand and it concerned a design of

a kiwi fruit pocket pack. The infringement

occurred not by sight of the pack, but by coming
across a description of it and the Court of Appeal

held that this was nevertheless an infringement. I

quote that merely as a recent example which upholds

the general principle set out by Copinger and,

indeed, points out that what it is doing is
following that principle, does so at page 158.

The next point that we would make is that the cases support the view that the absolutely hard and

fast distinction between idea and expression of

that idea cannot be maintained.

MASON CJ:  Mr Emmerson, I would not want to stop you from

raising this but, of course, this material should

have been elaborated and developed in-chief because

you put this submission in-chief and you should

have supported it with authority at that stage.

MR EMMERSON:  I would accept that, Your Honour. I had, at

that stage, assumed that authority would not be

necessary; that the matter was clear. I do not

think that anything that I am about to put is doing

more than showing how these established matters

meet some of the points that my learned friend made

in his submissions.

MASON CJ: Well, as I said, I would not stop you from doing

it but I do not retract what I said, that the

authorities should have been dealt with in-chief.

MR EMMERSON:  I accept that, Your Honour, and I can only say

that there are a vast number of authorities to

which one could have referred and, in fact, I would

now wish to refer to a small number of those. One

case which illustrates that is the case of Football

League v Littlewoods Pools, (1959) 1 Ch 637. That

was a case in which the alleged infringement

occurred in a list to be used in football pools and

the headnote at page 638, paragraph (2) of that

headnote says this:

That, although there can be no copyright

in information per se, including that of the
predetermination of future events of one's own

creation, and even though there was

insufficient skill, labour or effort in

reproducing the chronological list from the
clubs' list to support a claim to copyright in

the former, of necessity the information in

the programme of fixtures had to be reduced to

writing, and the skill, labour, time, judgment

Autodesk(2) 144 18/4/91

and ingenuity of the League, their servants or

agents, which, on the facts, had been brought

to bear in compiling the clubs' fixtures

lists, could not be separated from the work of

compiling the chronological list, and that,

accordingly, the League were entitled to
copyright in the chronological list.

Now, in our submission, that is relevant to the way in which my learned friend put his

submissions because we say that you cannot simply

carve up the various steps that we went through and

say the part and the only part that can be

protected is the final adaptation of the program to

a specific physical device.

If one goes to page 655 at the bottom of that

page and over to page 656, one finds that

His Lordship says:

Every case must depend on its own facts.

It is perfectly true that Sutcliffe was not

employed to produce a work of art per-senor

even a work primarily as a book of reference,

such as a directory or a railway guide, but he

was employed to produce the best possible

programme of fixtures. Of necessity, however,

that programme had to be reduced to writing.

The League's duty is to arrange the best

possible programme of games to please the

football public in general and the clubs'

finances in particular, and they can only do

that by producing a list or lists of those

games. If, as a result of prolonged

cogitations, Sutcliffe reaches the conclusion that it will be best if, for example, Arsenal

plays Manchester City at Highbury on September

20, 1958 (as the programme provided), he is

doing so no doubt primarily because that is

best from the point of view of league

football, but if, as the result of the whole

produces in a particular form the season's of his prolonged and skilled cogitations, he list -

he then sets that out -

in my judgment he or the League (who have, by

direct assignment, any copyright which might

otherwise vest in him) are entitled to claim

that the chronological list is produced as a

result of the entire skill, labour, time,

judgment and ingenuity of the League, their

servants and agents. In my judgment, on the

facts of this case, it is not open to the

defendants to try and dissect and break down

the efforts of Sutcliffe in the way suggested.

Autodesk(2) 145 18/4/91

Accordingly, in my judgment, the plaintiffs

are entitled to copyright in the chronological

list.

DAWSON J: Well does that say anything more than that it is

just difficult to draw a line between an idea and

its expression, a sharp line?

MR EMMERSON:  No, it does not really. It says that you can

support copyright in the subject-matter which
eventually attracts copyright by looking, amongst
other things, at the effort that went into the

aspect of the work which did not directly attract

copyright. So, in this case, the position is that

you acquire copyright by reason of Sutcliffe

committing certain things to writing. Now that act

of committing those things to writing is no doubt

something which is relatively easy to do, but when

it is then contended, well that is merely a trivial

exercise, what the court says is, you do not just

look at the committal of something to writing and

say that that is a trivial exercise, you look at

the whole work that went behind it.

Similarly, in the case before the Court, you

do not just look at the exercise of adaptation of a

program to a particular device having information
processing capabilities and say, "Well, it is only

that which one is looking at as the effort that

underlies the making of the work", but rather one

looks at the totality, and so you separate out the

question of whether this copyright subject-matter

and the question of whether sufficient effort went

into it to be protected, and you look at the

totality of the effort in that context.

Another example of the same sort of thing was

in Mirror Newspapers v Queensland Newspapers,
(1982) 59 FLR 71. That was a similar type of case.

There, what you had was a list of numbers forming a

bingo game and that was held to be capable of being

protected and the fact that one should not draw a

hard and fast line is set out by His Honour at the

bottom of page 76 where His Honour says:

Now it is not legitimate to draw a line

between the work which went into the idea and

the work which went into the expression of the
idea and to confine the protection of

copyright to the latter.

And he refers to the well-known case of Ladbroke

v William Hill, (1964) 1 WLR 273 and he also refers

to Football League v Littlewoods to which I have

just taken the Court.

Autodeck(2) 146 18/4/91

Another reference in Copinger which we say is

of assistance in considering the position of an
intervening work between the copyright work and the

alleged infringement, is to be found at

paragraph 545 in Copinger where the learned authors

say this:

Difficult cases have arisen where the

work alleged to have been infringed consisted

of an arrangement of non-copyright tunes -

that does not arise here, of course, but they go on

to say -

or where the infringing work was produced by

using an intervening work. In all such cases

the question is one of fact whether the

plaintiff's skill and labour has been

appropriated -

and that, in our submission, is of assistance in

looking at the present case. In the present case
the appellants' skill and labour have, we would

say, been appropriated by the respondents in the

manner in which they have made a very close copy of

what we say are the relevant bits of our program.

The next point to be made is this:  my learned

friend seemed to suggest in some cases that somehow

one could dismiss the program in suit on the ground

that it was a simple program. We would challenge

that. It is well established, in our submission,

that a work may be subject to copyright,

notwithstanding that it is a simple one, and again

a reference to Copinger is appropriate at this

point. Dealing with artistic works at

paragraph 177 the learned authors say this:

Simplicity is not enough to prevent copyright subsisting and drawings for rivets

and the like have been held protectable. Also

the artistic part of a trade mark and a

signature and Union emblem. Also drawings for

standard parts such as engines and gear boxes

and engineering drawings.

He mentions a whole lot of other matters in which

mere simplicity has not been enough to cause a

court to hold that copyright did not subsist.

Could I add to that passage a reference to

Ogden v Kis, (1982) 45 ALR 129, in which the

copyright subject-matter was the drawing of a key

profile. I do not propose to take the Court to the

next three cases I am mentioning in detail: the

Express Newspapers case, (1985) 5 IPR 193, where

the copyright subject-matter was a five letter code

Autodesk(2) 147 18/4/91

used in a newspaper game; and to the Kero Step

case, (1977) RPC 255, where the drawing in question

was a very, very simple logo used as a trade mark.

In all those cases, and one could multiply

them, the mere fact of simplicity has not been held

to lead to a conclusion that one cannot have

copyright.

Next, I want to say a little bit about a case

cited by my learned friend, Whelan v Jaslow,

(1986) 797 F 2d 1222. It is, of course, is

something that arises under United States law

which, in the area of copyright, differs fairly

substantially from our law. However, my learned

friend drew the Court's attention to the passage at

page 1236, in the left-hand column, and we would

say that this illustrates, admittedly under a

different legal system, some of the points which we

have been seeking to make. Beginning reading at

the top of the column, what the court said was:

In deciding this point, the Court

distinguished what was protectible from what

was not protectible as follows:

(W)here the art (i.e., the method of

accounting) it teaches cannot be used without

employment the methods and diagrams used to

illustrate the book, or such as are similar to

them, such methods and diagrams are to be

considered as necessary incidents to the art,

and given to the public.

Well, now, that particular problem, of course,

does not arise here even if it were to be part of

our law because in no sense could one say that the

computer programs here in suit had been given to

the public. On the contrary, the appellants took a

great deal of trouble to ensure that members of the

public did not find out what they were and, as we	know, it took Mr Kelly some weeks before he found
out what they were. So there is no question of
dedication. Then, in the next paragraph, the court says:

The Court's test in Baker v Selden

suggests a just way to distinguish idea from

expression. Just as Baker v Selden focused on

the end sought to be achieved by Selden's

book, the line between idea and expression may

be drawn with reference to the end sought to

be achieved by the work in question.	In other
words, the purpose or function of a

utilitarian work would be the work's idea, and

everything that is not necessary to that

Autodesk(2) 148 18/4/91

purpose or function would be part of the

expression of the idea.

Now, we would have no difficulty with that

passage but we would read it in the present context

as follows. We would say that the underlying

idea - that of using a lock which carries out some

processing on digital information which it receives

and sends back the result to a computer - is to be

described as the idea. But, of course, that idea -

as the evidence makes clear - could be worked out

in any one of an unlimited number of different

ways. We have worked out that idea in one

particular way with one particular aspect of the

incoming data stream which is used to trigger the

activity of the lock and with a particular set of

states and sequence of numbers that is sent back.

Now that is not necessary for working out the

idea of a lock or even for working out the idea of
a lock of the same general type as in the present

case. So that working out is not on this test, in

our submission, part of the essential idea

underlying our lock. It is to be noted that on the

succeeding page 1237, left-hand column, beginning

in the middle of the paragraph, about half-way down

the page, the Court says this:

As we stated above, see supra at 1231, among

the more significant costs in computer

programming are those attributable to

developing the structure and logic of the

program. The rule proposed here, which allows

copyright protection beyond the literal
computer code, would provide a proper

incentive for programmers by protecting their

most valuable efforts, while not giving them a

stranglehold over the development of new

computer devices that accomplish the same end.

Now, applying that test, it is our submission that

entirely consistent with the views that we have	that leads to an important distinction which is	been putting in the present case. There is nothing
in the submissions that we would put which would
stop other people protecting their programs by devising locks and those locks could be digital information processing devices, as in the present case. They could carry out arithmetic and or
logical operations on data sent out from the
computer and could transmit them back, and there is
nothing in the submissions that we have put which
would stop somebody protecting his program in that
particular way.

What we seek to do is stop the respondents

from selling a lock which can be used1 and used

Autodesk(2) 149 18/4/91

only, with our AutoCAD program - that is the only

use of it, as the trial judge found, and was fully accepted by the Full Court, and the only use of it

is to overcome a precaution which the appellants

have taken to prevent piracy.

Now, in our submission, what the appellants are doing is analogous, not offering keys generally

as by a locksmith, but by saying, I have found out

a key to a particular safe where there are things

of value and I am now selling commercially keys to

that safe. Now, that distinction is entirely in

accordance with what Whelan Associates are saying,

because they say that computer protection goes

beyond the literal code and the aim should be to

provide a proper incentive for programmers by

protecting their most valuable effects, which is,

of course, precisely what we are seeking to do
here, and then the converse is, while not giving

them a stranglehold over the development of new

computer devices that accomplish the same end, we

do not seek a stranglehold of new computer devices

which would accomplish the end of being locks.

What we do, however, is seek to prevent other

people from selling a lock which can be used to
provide, in essence, the key to our own AutoCAD

program.

If the Court pleases, the remaining matter

that is to be dealt with is the matter of

authorization. As I put this matter in my

aspects of the authorization issue on which the

submissions yesterday at page 77 of the transcript,

court below had found against us and the first one

that I mentioned, which I said was the important

one, was the one with which I dealt more

substantially and it is that one, which we say, is

a matter of such importance that it is proper that

this Court should consider it.	I then went on to
say that:

The other aspect -

on which the courts had found against us was -

that the court held that it was not actually

established that the use of the respondents'

locks would involve infringing conduct of the

sort that I have just described, and it is

true that there was no direct evidence that

that had happened.

I then put the submission:

that the inference is blazingly clear. The

only use for these particular devices is an

Autodesk(2) 150 18/4/91

infringing use, that is to say, unauthorized

running of the AutoCAD.

Now, we stand by those submissions. The reason why

the point is important is, of course, the first

point. My learned friend in his submissions,

understandably enough, deals with the second of

those two points and as to this, we simply say that
the first point is sufficiently important and that

the Court should be disposed to consider it, as

soon as a suitable case arises, and that this is a

suitable case because, in our submission, while we

accept that the courts below refused to draw the
inference that there had been infringing activity

as a result of the sale of the respondents' locks,

we say that that is the correct inference to be

drawn.

My learned friend tended to suggest that we

had put this matter in an inconsistent fashion

before the learned trial judge and before the Full

Court. We would not accept that.

BRENNAN J:  Do you not have to deal with, first of all,

whether the question arises on the findings that

have been made, however it has been put?

MR EMMERSON:  I am directing my submissions at this moment

to the question of whether the question arises and

I am saying that before the Full Court we, in fact,

or alternatively, no person is authorized to run

put two arguments that the question arose, and
reading from my notes on this occasion, we said

first, no purchaser of AutoCAD is authorized to run

any unauthorized copy of AutoCAD. So, in fact, we

argued both those points before the Full Court.

Now, it is because the Full Court found

against us on the construction of the licence

document that,in my submissions yesterday, I did

AutoCAD was entitled to run AutoCAD save with an	not seek to put the proposition that no owner of
AutoCAD lock for the reason that we would accept,
as my learned friend puts it, that that does
involve construction of documents. It is a point
which has been determined against us and a point which we would accept for present purposes is not
of sufficient importance to trouble this Court.

But there is, nevertheless, the other way that

we get to it and we just simply say that it is

obvious from the surrounding circumstances that the

activities of the respondents are such as to

encourage the use of pirated copies of AutoCAD.	My
learned friend suggested, but I may have misheard
him, that there were actual findings of fact that

Autodesk(2) 151 18/4/91

his clients sold only to authorized AutoCAD users.

That was not the po~ition. What the trial judge

found was simply that we had not satisfied him that

he should draw the inference that there was

unauthorized use.

What he says, at page 992 - and what we have

to accept is his finding against us on this - is:

There is no evidence before the Court to

suggest that users of AutoCAD have purchased
the Auto-Key lock to enable them to make extra
copies of AutoCAD to be disposed of and to be
used together with the Auto-Key lock, at a

profit to third persons. There is no evidence

that any person, not being a purchaser of the

AutoCAD program is using AutoCAD. The

advertisements for the Auto-Key lock are

directed to AutoCAD users. The names of some

persons who have purchased the Auto-Key lock

are before the Court, but there is no evidence

that they have used the Auto-Key lock or the

methods by which it has been used. There is

no evidence to suggest that any AutoCAD user
has used the program on more than one computer

at any one time.

So what His Honour found and what we, of course, cannot challenge here, is that there was no

positive evidence by which we were able to

establish actual illicit use. What we say is that

His Honour should have found, and the Full Court

should have found, that there was an almost

irresistible inference from the surrounding

circumstances and the nature of the respondents'
locks that this would be the use for them.

The Court will recall that His Honour found that the respondents went to very great pains to conceal their identity, that they kept no detailed

records of purchases, first names only were used,
and so on.	Indeed, we would say that it was not

really challenged that the circumstances were such

that it would be very easy for the respondents'

locks to be used for unlicensed running of AutoCAD

programs or the running of unlicensed copies of

AutoCAD programs.

BRENNAN J:  Be it so, what is the cause of action and where

was it pleaded?

MR EMMERSON:	The cause of action here is authorizing infringement of copyright.
BRENNAN J:	By whom?

Autodesk(2) 152 18/4/91

MR EMMERSON: Authorizing by the respondents of infringement

of copyright by the respondents' customers.

BRENNAN J:  But there is no finding that any customer has

infringed.

MR EMMERSON:  No, I accept that, and the way that we must

put it is to say that this Court should draw the

inference nevertheless from the circumstances that

some customers would infringe.

BRENNAN J:	Not would - have.
MR EMMERSON:	Have, yes.
BRENNAN J:	And the finding is against it.
MR EMMERSON:	No, there is no finding that there has been no

infringement by customers. What His Honour found

was that there was no evidence, we would say, of

any particular infringement.

DAWSON J: Well, is the onus on you to establish the fact?

MR EMMERSON:	Yes, it is, Your Honour.
DAWSON J:	You have not established it.

MR EMMERSON: Well, I have put the matter as high as I can.

We simply say that this is an inference that the

Court should draw. I should perhaps add that it is

of concern that the courts below seem to have

found, albeit it tentatively, against us on the

point that I described as being the really

important point and it is for that reason that we

have sought to air that point in this Court. But

if the Court is not disposed to hear argument on

that point, I cannot put the other aspect more

highly than I have.

MASON CJ: Yes, thank you, Mr Emmerson. Mr Emmerson, as

presently advised, the Court is of opinion that the

grant of leave should be confined so as to exclude

the authorization issue.

MR EMMERSON: If the Court pleases. There are no other

matters which I would seek to raise in reply, if

the Court pleases.

MASON CJ:  In the light of that, the Court will specifically

confine the grant of leave so as to exclude the

authorization issue and that will dispense with any

necessity for you to reply on that issue,

Mr Burnside.

MR BURNSIDE: Yes, Your Honour, it does.

Autodesk(2) 153 18/4/91

MR EMMERSON: If the Court pleases, those are the

submissions for the appellant.

MASON CJ:  The Court will consider its decision in this

matter.

AT 3.42 PM THE MATTER WAS ADJOURNED SINE DIE

Autodesk(2) 154 18/4/91