[SystemSafety] Limitation of liability (was Qualifying SW as "proven in use" [Measuring Software])

Pekka Pihlajasaari pekka at data.co.za
Wed Jun 26 12:50:06 CEST 2013


On 26/06/2013 01:09, Steve Tockey wrote:
> In fact, I would even advocate a removal of the liability waiver on
> software licenses. Let programmers who write code that fails be liable
for
> the damage that their defects caused. Then, and probably only then,
will
> we see proper professionalism applied to software development. It's
> clearly (to me, at least) not an issue of software not being able to
be
> built in a solid, reliable way, it's simply that the practitioners are
too
> used to complete immunity from their sloppiness.

Software developed in safety critical industries is presumably supplied
with licenses very different from the shrink-wrap disclaimers we are
familiar with on commodity software. It are also much more expensive and
provides less choice.

Is it possible that both the user nor developer would not benefit from
mandating more strict liability? The probable reduction in choice as
players exit the industry and greatly increased costs, at least
partially resulting from insurance premiums, would clearly influence
their choice.

Perhaps the invisible hand of the market has established that shifting
the risk to the licensee is efficient. Certainly the model was prevalent
even before dominant software companies could have created a status quo
around this.

Regards
Pekka Pihlajasaari
--
pekka at data.co.za 	Data Abstraction (Pty) Ltd	+27 11 484 9664



More information about the systemsafety mailing list