::: Area #157 (comp.sys.acorn.programmer)
Message: #48791 (Read 4 times, has 0 replies, 9900 bytes)
Date   : Wed Jun  7 18:22:33 2000
From   : ahunter@chrysoprase.logicalshift.org.uk
Subject: Re: Software copy protection - how?
Message-ID: <slrn8jqkcp.1s8.ahunter@chrysoprase.localdomain>

(Ooh, I like delayed posting. I can edit my post before it goes out to the
world)

First off, I accept that amendment of a copyright work is illegal. My
point here is that software is not simple stuff in any way.

In article <8hiese$jta$1@nnrp1.deja.com>, gbuchanan@my-deja.com wrote:
>In article <slrn8jo34m.1t8.ahunter@chrysoprase.localdomain>,
>  ahunter@chrysoprase.logicalshift.org.uk (Andrew Hunter) wrote:
>> Hm, does this make something like FX!32 illegal then? For those who
>don't
>> know about this, FX!32 was a system used by Digital to make
>Intel-based
>> applications run on Alpha workstations. FX!32 could run applications
>faster
>> than they did on the Intel processor, because FX!32 could translate
>> Intel instructions into Alpha ones in a similar manner to a JIT. I
>don't
>> recall any court cases involving this software, but I could be wrong.
>> Doesn't transmeta's 'code-morphing' technology work in a similar
>manner,
>> too? (Does it make a difference when this sort of thing is done in
>software
>> or hardware - it's semantically similar to microcode translation in a
>CPU)
>...no, the above would be emulators which don't as such alter the
>copyright work.  In fact, case law clearly upholds the validity of
>emulators.  When the first Spectrum emulators arrived, a number of
>software houses tried to sue the emulator writers.  The case went to
>court in the UK and the emulator manufacturers were cleared of any
>infringement as their emulators didn't alter the original software.

The thing about FX!32 is that it *does* modify the original code, replacing
Intel instructions with their Alpha equivalent. On the first pass, it's
an emulator. On the second pass, it's executing native code. To use (sigh)
a metaphor, it is doing the equivalent of translating a book into a
foreign language without (necessarily) the permission of the copyright
holder. The metaphor breaks down, as the act of translation replaces the
original, whereas a normal translation would result in two copies of the
book... Oh, and the translation is absolutely literal, even though the
languages have different grammar and lexicons...

I think it saved its alterations, too, but I have no specific experience.
VMWare optimises code by replacing code that needs priviledged processor
modes with direct system calls (so it avoids the overhead of trapping
the calls).

Hm, plus there's the use of the word 'permanent' in some of your posts.
Very few digital storage mediums are really permanent. Computer
Organisation classes and texts put magnetic discs in the same category as
RAM. (The way their content is presented to you is purely a user interface
issue, and this can and is changed fairly easily. Disc/memory editors,
virtual memory, debuggers, disc statistic tools, etc, etc)

>> I don't think you've explained why it is a copyright violation to
>alter
>> software stored on my own hard disc, and I can't think of any instance
>> of a licence that specifically forbids this, either.
>...that is rudimentary.  The work given to you is copyright.
>All copyright law specifically forbids what is termed "adaptation",
>which in essence is any change to the original form.  Acknowledging
>the original is not a sufficient defence.
>
>Reverse engineering for interoperability, security analysis and
>bug detection is aok, and specifically provided for in UK law, and
>in other EU states.
>
>> I still don't see why alteration of the application as installed on my
>> hard drive could ever be an infringement of anything? Is it illegal
>> to deface a book I bought?
>...actually, yes.

Really? We studied and annotated books while studying English Literature.
Some of these still had current copyrights. Given that we were an average
bunch of Sixth-formers at the time, I should imagine that some of these
annotations counted as defacements... Eep! Still, they were school copies,
and they don't have anyone's name on them (unless you count references to
teachers that seemed amusing at the time)

What about a copyright book distributed in electronic media? What if the
'defacements' are in a VCS, and can be 'removed' by viewing a previous
version? (I don't expect there's an answer to either of these questions,
as the only examples where this may be relevant are probably technical
documentation, such as that supplied by the ITU on CD-ROM). Heh, what if I
use 'strings' on a word file to see the bits that its author thought he
deleted, but Word never got around to removing?

>> Do you mean you can't distribute the patch or the patched code?
>...both.  The reason being that the patch serves no useful purpose
>on its own.  Thus, under standing copyright law, it is what is
>termed a dependent adaptation, and this is an infringment of
>copyright.  For example, if I took a TV script and wrote a small
>change to it, I could not distribute ether the doctored script in
>full, nor could I distribute the "amendment".
>
>> Does the law really make a distiction between
>> information that a computer can and can't understand?)
>...not quite.  If I tell _you_ how to perform something, then
>that description is an independent copyright work.  So, for
>example, performance notes on the TV script are aok as they
>have an independent use.  It not only refines the work, but
>would deliver an interpretation of it.
>
>I realise this may seem an unlikely, if not perverse distinction,
>but it is clearly the law.

Hrm, most patches are 'refinements' anyway. As a programmer, I can read
the actions described by a patch file the same way a mathematician can
read a system of equations. And this isn't just a theoretical argument -
I have read patch files to see what they do, and then (realising that
the patch program is too stupid to do the patch correctly with an
updated piece of software, say) applied the patch by hand with an
editor. Both machine-code patches (to get some old piece of software
working where the original patch was for a different version) and source-code
patches (as part of the process of merging branches of a version-controlled
system). In both cases the patch descriptions were designed for machine
interpretation. Thing is, if a machine can interpret it, so can I!

I'm not arguing it's not the law, but I wonder if the interpretation is
valid. After all, for all the metaphors, a computer is *not* a book,
or a car, or anything similar, on a fundamental level. (After all, when
you get a book, you don't copy it, in pencil, into a pad to make it legible.
If you did, you wouldn't start from the first free page in that pad,
then skip on when you get to a full page, marking the used pages in
a table at the beginning of the pad, and associating them with that book.
If you read the pad from beginning to end, it would look like a weird
jumble of passages from various differnent books, interspersed with
bits of your own work, information about which books are on the pad,
etc, etc. Or is all that irrelevant? Is it illegal for an operating system
to corrupt a filesystem and reveal the chaos within? Assuming the law to
be as you say it is, I would say not, but if you did that yourself, it is.
Hrm, better not tell anybody about what happened to my copy of Windows '98
when I editted the partition table by hand and accidentally overwrote a
bit of the FAT too [changed heads instead of sectors in the editor, but
didn't notice until I rebooted]. I think that counted as 'defaced' :-)

Software is a weird concept anyway. It's not like a book, as a book is
just information (umm, that's unclear. With a book, you see the authors
exact words. With software, you don't see all that's there unless you
really try). It's not like a machine, because it can't do anything
without a computer. It *is* a series of instructions, intended for use
by a computer - when a computer follows the instructions, you get to see
whatever the software is intended to do - *assuming* you have the right
companion software, otherwise you see something else... Oh, and the software
doesn't do anything - the computer and the user interface just fool you
into thinking it's the software that's performing the actions, when
it's the computer, a seperate entity, that's performing the functions
described to it by the software.

I guess the problem is that the legal system seems to focus on the
presentation layer, even though there are at least two layers below
that (software and hardware. OSI would have us believe there are, in
reality, 6 layers below. TCP/IP gives us 4 or 5 layers). Napster is
getting sued because the presentation layer of its software makes it
appear that copyright information is coming from their servers (to pick a
topical example at random). Computers are general purpose machines - as
far as I know, the only device we've ever created that falls into that
category (I suspect that a strict definition would mean that we,
ourselves, also fall into the same category ;-).

I think the problem is that the law just isn't equipped to deal with the
special problems of software, or digital storage in general. Plus there's
increasing evidence that the lawmakers have even less of a clue[1]...

I think I'll shut up now...

Andrew (IANAL, if you didn't guess already).

[1] Mentioning no home secretaries...

--
http://www.logicalshift.org.uk
Andrew Hunter <andrew@logicalshift.org.uk>... b0rken (but effective) spam
filter in use. Think of this as an excuse to fix your mail client.
Disclaimer: The preceding message may be total nonsense

--- EchoMaker 0.50
 # Origin: GXSN
 * Origin: The Arcade BBS Usenet News Gateway (2:254/27.2)