On software quality and engineering

[Simon Fowler]

On Sat, Nov 02, 2002 at 09:55:27AM +1100, Brad Hards wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
> 
> A concept that has been missed in all of this is that all things are designed 
> to meet acceptable risk. The Pinto example is where they mis-defined what 
> level of risk was acceptable.
> 
> Aircraft software is not designed to be bug free. It is designed to contribute 
> less than an acceptable conponent of aircraft crashes. Typical sort of 
> numbers are "1x10-9 per hour for safety critical items". You can't test to 
> that sort of number, so you go with "process" approach.
> 
> However most of the software on an aircraft isn't safety critical (RTCA 
> DO-178B level A). So you design to a lower level of reliability (eg. the 
> intercom is probably level C, and the in flight entertainment systems is 
> probably level E - so it doesn't have any software process requirements).
> 
> It is unrealistic to expect that complex systems will not fail. It only 
> realistic that a system fails at (or below) an acceptable level. Normally the 
> risks are defined in terms of probability of failure (or partial performance) 
> and the consequences of failure (or partial performance). 
> 
I think the difference with software is that it's not like a
physical system - it doesn't wear out over time, it doesn't have an
inevitable failure at some point in the future, even if it's perfect
now. Software is either correct or it's not correct. 

Of course, back in the real world software is /never/ correct, since
it's impossible to get all the bugs out of any non-trivial piece of
software. So although the class of failures you get in physical
systems due to things wearing out don't happen, you can probably
still apply much the same reasoning about failure rates . . . 

> This might make a decent topic for November CLUG. Not sure if I'll be there at 
> this stage, but I'm willing to present on this.
> 
It's certainly an interesting topic, and it's one I haven't seen
mentioned in any of the courses I've done in my degree . . . I'd
love to hear about it from a /real/ engineer ;-)

Simon

-- 
PGP public key Id 0x144A991C, or http://himi.org/stuff/himi.asc
(crappy) Homepage: http://himi.org
doe #237 (see http://www.lemuria.org/DeCSS) 
My DeCSS mirror: ftp://himi.org/pub/mirrors/css/ 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 232 bytes
Desc: not available
Url : http://lists.samba.org/archive/linux/attachments/20021102/f7d9ade2/attachment.bin