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

From: Smith, Brian E. (ARC-TH) < >
Date: Mon, 9 Mar 2015 16:15:40 +0000


Thanks for attempting to return the discussion to Œhome base.¹

In addition to the manifold academic and technical dimensions debated in this thread, there is a moral one also that falls out of the socio-tecnhical environment in which these software-intensive system operate.

In his piece entitled Computing and Moral Responsibility, at,, Merel Noorman writesŠ

³The design and use of technological artifacts is a moral activity and the choice for one particular design solution over another has real and material consequences.

³Accountability Š is different from liability. Liability is about looking for a person to blame and to compensate for damages suffered after the event. Once that person has been found, others can be let Œoff the hook¹, which may encourage people to look for excuses, such as blaming the computer. Accountability, however, applies to all those involved. It requires a particular kind of organizational context, one in which answerability works to entice people to pay greater attention to system safety, reliability and sound design [my emphasis], in order to establish a culture of accountability. An organization that places less value on accountability and that has little regards for responsibilities in organizing their production [or research] processes is more likely to allow their technological products to become incomprehensible.²


On 3/9/15, 5:25 AM, "Martyn Thomas" <martyn_at_xxxxxx

>You will recall that I started the thread with the title above by asking
>a question. To paraphrase: if some software is operating in a given
>operational environment, and fails repeatedly, and the causes of the
>failure are corrected by error corrections to the software, and the
>process continues for a significant time during which the number of
>failures per hour are observed to have reduced, what property of the
>software has improved? Specifically, why would it be wrong to call this
>property "reliability"?
>Is it meaningless to say that one release of a software system is "more
>reliable" than an earlier release? If you believe it is meaningless to
>talk of software reliability in these contexts, please explain what
>phrase you would prefer to use, and why.
>On 09/03/2015 10:14, Nick Tudor wrote:
>> ...
>> In my view, the reason so many have commented on the list is that the
>> kind of thinking espoused regarding so called "software reliability"
>> costs industry and tax payers money and it is frustrating to have such
>> written in standards which ill-informed users, such as those in
>> government, take as read.
>The System Safety Mailing List

The System Safety Mailing List
systemsafety_at_xxxxxx Received on Mon Mar 09 2015 - 17:15:55 CET

This archive was generated by hypermail 2.3.0 : Tue Jun 04 2019 - 21:17:07 CEST