[SystemSafety] Candidates for the firing squad

[Olwen Morgan]

I have considerable sympathy with these views and often find myself 
wanting to use C++ simply to get abstract data types. (Note, I do not 
mean I'd use OO programming, which IMO is something that, at least with 
widely used "OO" languages, leading software engineering up a blind alley.)

After having written a C++ coding standard for BT back in the 1990s, I 
decided that C++ was such a cr at p language that I'd never use it and 
would leave any project that did. There is no doubt that, restricted to 
a suitable subset, and with proper tools, C++ can be used safely. 
Unfortunately, C++ allows lumpenprogrammers to produce worse messes much 
more easily than C does. Even though I know how to use it safely - and 
that's probably with even greater restrictions than David Crocker might 
propose - there comes a point where one is so surrounded by people who 
won't use it safely that it's just too frustrating to be working with them,



O




On 21/09/18 09:26, David Crocker wrote:
> This is one of many reasons why I much prefer writing in a subset of C++
> to C. In C++'11 and later you can specify the underlying type used to
> store your enumeration. It's obviously sensible to to have a coding rule
> that requires you to use that facility when declaring an enumeration.
>
> David Crocker, Escher Technologies Ltd.
> http://www.eschertech.com
> Tel. +44 (0)20 8144 3265 or +44 (0)7977 211486
>
> On 21/09/2018 07:12, David Haworth wrote:
>> Typical example: a range check on an enumerated type. Because
>> the signedness is implementation-defined, if you test for < 0
>> you'll get a warning (expression is always true) sometimes.
>> But you can't leave it out ...
> _______________________________________________
> The System Safety Mailing List
> systemsafety at TechFak.Uni-Bielefeld.DE
>

-- 
Olwen Morgan CITP, MBCS olwen.morgan at btinternet.com +44 (0) 7854 899667 
Carmarthenshire, Wales, UK