[SystemSafety] Safety-Critical Systems eJournal

[Derek M Jones]

John,

> As those of you already registered as a Reader on the web-site will know,
> the launch issue of the Safety-Critical Systems eJournal is now available
> for viewing at: https://scsc.uk/journal/index.php/scsj/issue/current

Congratulations on a successful launch.

Dewi Daniels and Nick Tudor's article hits a few nails on the head.

Yes, the characteristics of software failures are different from hardware ones,
and a lot of the statistics that appears in software reliability papers is
out to lunch.

One very important point that was not really covered was the lack of data
needed to validate reliability models.  Having a plausible sounding theory
is all well and good, but how do we know it has any connection to relaity
without the data to validate against?

Billions get spent sending spacecraft to planets to take snapshots of them,
but nothing gets spent gathering software reliability data.

When people complain about the reliability of software, I point out
that nobody is actually interested in this reliability, and if they were
then they would spend some money on it.

-- 
Derek M. Jones           Evidence-based software engineering
blog:https://shape-of-code.com