🚀 go-pugleaf

In one of the reactions to the "Soap Software" teaser I posted a while
ago, Michael J. Farren commented:

> One of the biggest problems I have with the current 'debate' on copy
> protection which is raging throughout the microcomputer groups is,
> simply, that people are vehemently defending positions that they have
> taken with, it seems to me, no actual evidence at all to support those
> positions.

I can't claim finders credit for "discovering" software theft.  
Anecdotal evidence is widely available (e.g. PERSONAL COMPUTING, May 
1987). More verifiably, one can cite three major sources of evidence: 
surveys, legal literature, epidemiology.

Granted, the best "proof" would be a scientific, random sample of 
hard disks in use at large.  We attempted to do just such a study - and 
reported some data. Not surprisingly, we ran into some sampling problems.

The Federation Against Software Theft (in the UK) published a 1986 
estimate for national annual loss of about 150 million British Pounds 
in Great Britain alone.  Their data show about five illegal copies for 
each duely purchased package.  Future Computing (based in Texas) 
completed a large scale survey of users, and estimated an annual 800 
million dollars loss, extrapolating a single purchase for each four 
acts of theft.  They projected a single stolen copy for each legally 
paid-for package.  Write to each of these organizations, as well as 
North American based ADAPSO for details.

There is a deluge of legal literature on the topic.  Hardly likely 
that this literature is all hypothetical.  Also, see the Lotus vs. 
Rixson 1984 case.

The propagation of viruses (viri?) is another indirect (unobtrusive?) 
epidemiological measure of the phenomenon.  Viruses seem to flow so 
rapidly across sites and even continents (talk to any administrator 
of microcomputer clusters in industry or education).  The "carriers" 
are mostly executable files.  And these are mostly originally, ostensibly 
(and by contract) intended "for use on one machine".

Our study is available as part of the Proceedings of the 1987 National 
Conference on Data Processing, Jerusalem, Israel, pp. 106 - 135. 
Other versions have appeared in a local computing journal and as a 
chapter in a legal textbook. An English version is in the works.

However, I think you are missing the point by fixing on the financial 
loss of manufacturers.  The three most important issues about theft have 
to do with users, not manufacturers.  These are not the taking of high 
moral ground, blaming the thieves, rather:

a) Morally and philosophically, software theft is of concern.
   There is a second-order reflection on the perceptions of
   computers, intellectual property, the worth of creativity, the
   subculture of programming, etc. Our respondents agreed that
   software theft is a less serious offense (or social problem) than
   jaywalking, cutting-in ahead of a line, smoking in a public
   place.

b) Efficiency-wise: it seems that (most?) stolen software gets 
   disseminated sans some unpilferable ingredients, such as the
   tutorial, manual, installation files. These (numerous?) users 
   are likely to be using the software in a less efficient,
   knowledgeable manner. We are now trying to study this question
   systematically. Can "quality of use" be measured?

c) Software itself (like art) is supposed to be a process, not a product.
   A lot of MIS and computer science literature is devoted to
   showing this. I wonder whether there is a nice way to say the
   same (process, not product) about art, film, journalism, etc.
   Software maintenance, iterative design, ongoing and context-
   sensitive documentation, structured programming, etc. are all
   concepts related to the need for software to be dynamic,
   developing.
   
   If software is a process why is it priced like a product?  What 
   happens to processes that are priced like stagnant products?
   How does the software market reward excellence, reward early
   adopters, treat updates, encourage rethinking, ensure
   intellectual property rights, guarantee its own ongoing size?

I would like to bring the discussion back to the original proposition, 
of advertising support for software.  I have received numerous reactions 
(in this forum, as well as privately).  I will summarize and post these, 
as well as report some empirical findings from our experiments.  (Data, 
Michael Farren, data!).  But I think the more interesting issue has gone 
mostly unnoticed:

	What are the implications, promises or perils of advertising
	in software? Are the only valid questions whether it will be
	acceptable, or whether theft exists?

Sheizaf Rafaeli