[SystemSafety] Software reliability (or whatever you would prefer to call it)

[C. Michael Holloway]

On 3/6/15 6:18 AM, RICQUE Bertrand (SAGEM DEFENSE SECURITE) wrote:
> The aim is to find a way to convince an authority (which is far from demonstrating something to a scholar) ...
In my experience, demonstrating something to a "scholar" may be often 
easier than convincing an "authority". That is, at least some scholars 
seem to be swayed by arguments of dubious cogency and evidence of 
questionable credibility that would not sway many, if any, approval 
authorities.  The temptation to cite this thread (and its predecessors) 
as supporting evidence is strong, but I shall resist it.

To be slightly less snarky, I'll elaborate a little bit in very generic 
terms.

Consider some sub-field of study R. Very smart (and, as with all fields, 
some not-so-smart) academics are involved in R.  Many papers are 
published.  Many of these papers carefully enumerate the conditions C 
necessary for R to be relevant.  For most of the researchers in R, 
mentioning C is important for reasons of integrity, but irrelevant for 
their work, because they are only interested in situations in which C holds.

Suppose that in the real world, C never, or almost never, holds. 
Scholars may likely continue to believe that R is a field worth studying 
and continue to produce results and continue to seek to convince each 
other that one approach is a better than another.  But approval 
authorities will not, and should not, be convinced that R should be used 
in systems they approve, because C does not hold for those systems.


-- 
/*cMh*/

*C. Michael Holloway*, Senior Research Engineer
Safety Critical Avionics Systems Branch, Research Directorate
NASA Langley Research Center / MS 130 Hampton VA 23681-2199 USA

The words in this message are mine alone; neither blame nor credit NASA 
for them.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.techfak.uni-bielefeld.de/mailman/private/systemsafety/attachments/20150306/5f9fb407/attachment.html>