Re: DO-178

[Enrico Colombini <erix@...>]

On 21/06/2012 20.57, Gé Weijers wrote:
> For insight into why MISRA C is well-intended but ultimately harmful:
>
> http://www.leshatton.org/2005/11/language-subsetting-in-an-industrial-context-a-comparison-of-misra-c-1998-and-misra-c-2004/

Interesting reading, thanks. Shows how committees work :-)
I see two possible weak points in that paper, though:

- The analysis applies to retrofitting existing code, as opposed to code 
written to comply with MISRA C rules from the beginning.

- I suspect those 'noise rankings' could be inaccurate even in a 
retrofitting scenario, because the author implies a constant correlation 
between a fix and its probability of introducing a new error, regardless 
of the rule.
For example, making old code conform to a rule such as "single exit 
point from a function" would have a much higher probability of 
introducing new problems, compared to "always use braces in conditional 
statements".

These are just my impressions; I have no experience of DO-178 or MISRA 
C, even if I routinely apply a number of those 'noisy' rules to my own 
coding.

--
  Enrico