One concern which supports a conservative approach to these matters relates to background knowledge.
The well-intentioned experimenter (and hapless third parties) may be exposed needlessly to unnecessary risk of a greater or lesser extent due to the simple fact that his/her technical background does not encompass lessons learnt earlier by the design and test industry.
Even those who ought to know better sometimes can fall foul of this trap. I recall one fatal test accident within a factory program some years ago where an ill-considered modification was being tried out as a hoped-for solution to an ongoing handling difficulty. The result ? ... two dead and one very seriously (and permanently) injured.
To me, if there is not some sort of sensible and competent peer-review assessment done prior to development, then the risk of mishap goes up unacceptably.
While I do not support the idea of excessive and oppressive paternalism in the process, there ought to be a balanced middle ground which protects most of the various stakeholders' desires and interests.
Or do we simply wish to enshrine the right of the individual occasionally to be stupid, wilfully reckless, negligent, or just unfortunately in the wrong mod at the wrong time ... ?
[ 06 January 2002: Message edited by: john_tullamarine ]</p>