"Information may want to be free, but a little bit of knowledge can be a dangerous thing."
Beyond the somewhat religious software debate of Closed Proprietary Code versus Free & Open Source Software for reason of justice, equity and safety other paradigms are also desperately needed.
Where software is used as a data processing and / or as an generalized information tool, "Free Software" {as per "The Free Software Definition - GNU Project - Free Software Foundation (FSF)" http://www.gnu.org/philosophy/free-sw.html } is arguably the best approach. But unfortunately there is a major policy concern in allowing anybody to "have the freedom to make modifications".
Moving away from software for a moment, governments have among their obligations the public safety, protection and advancement of the population health and security. Accordingly they enact legislation providing laws, regulations, procedures, guidelines & standards, administrators, authorities and instrumentalities.
The 'software libre' (= software that is not only free of licensing fees but whose development is not controlled by a single company) movement commonly makes a distinction as '``Free software'' is a matter of liberty, not price. To understand the concept, you should think of ``free'' as in ``free speech,'' not as in ``free beer.'''. The Beer reference is fortuitous in that historically Beer was the first food product to be regulated for purity and quality as a protection of public health.
Traditionally (prior to the OO paradigm) source code was to varying degrees a somewhat structured contiguous blob of serial decisions and procedural interconnections. Alterations, additions and deletion necessitated that the whole programme be once again tested and debugged. Changes thus become integral to the future evolution of the whole programme. This is the context that the GNU GPL free-software licenses where conceived and snugly fit as the "copyleft" legal construct.
The advent of Object Oriented (OO) programming paradigm strived to treat blocks of software as self-contained independent widgets that could be interchanged and slotted into various programmes analogous to standardized machine part (nuts, bolts, screws, gears, chains, cogs, transistors, pipes etc) in the manufacture of cars or washing machines.
Expert Systems, Genetic Algorithms, Neural-Nets and other Knowledge processing and Artificial Intelligence techniques then leveraged the OO paradigm into professional field where the 'software libre' in allowing anybody to "have the freedom to make modifications" becomes positively dangerous.
Returning now to the policy imperative for wise legislation that advances the public good, not anybody is allowed to hang out their shingle and play doctor, nor is practice of many other professions in the modern world open to any unqualified person that wants to give it a go. Professional training is a long process because the acquisition of any sizable body of knowledge takes time for the mind to absorb then apply so as to full appreciate the implications of said same knowledge.
Software used by Public Authorities would often necessitate the encapsulation such knowledge within the programme. Be they engineering formula for breaking strains in bridge design or chemical concentration in a pharmaceutical, to pick just two examples at random. Thus there exist a whole lot of software where the freedom to make modifications needs to be a least heavily qualified if not proscribed altogether. This raise a string of questions;-
How can this be reconciled with copyleft legal obligations?
How could such software objects / components be authenticated / certified locally and internationally?
How should liability issue underwritten, or accommodated / abrogated?
How could a license be structured to preclude changes by unqualified persons, while still allowing extension and enhancement that do not impinge or negate the core (non-software) disciplines' knowledge integrity?
How could such licenses be framed to prevent capture by proprietary concerns or professional cliques?
How should dangerous knowledge (like the process for a drugs manufacture) be encapsulated with-in such programmes so as to be transparent to qualified practitioners but closed from general scrutiny. Put another way, how could black-boxes & grey-boxes be safely and harmoniously integrated into to the Open-Source philosophy??
When
I first posted this question I got a lot of feedback along the
lines, that under the GPL people can modify the software and not
publish the source, provided they only use it in-house. While this
perfectly acceptable for organizations and governments in
high-tech nations this is not much help to smaller agrarian,
traditional or 3rd world nations. What large hi-tech governments
may do with Open-source code is one-thing al-la the initial
feedback. But how can we foster a body of 'Free Knowledge Objects'
that small say Pacific Island nation could confidently use without
fully auditing or re-inventing the wheel over and over again. Yet
still safely and harmoniously meet the aspirations of the
Free-software movement??
W. Shawn Gray June
2003.
