Assistance needed

Assistance needed

[Chris Sparks (aka Mr. Ada)]

Hi fellow Ada-philes,

I need some ammunition against using C++.  There has been some talk here
at work that some of the
future add-on work may  be using C++.  The software in question is
safety-critical and I wanted
to quote "BAD" examples of C++ usage on projects.

Examples I have seen are:
  Navy plight with Windows NT
  Colorado Airport baggage snafu

URLs will be greatly appreciated.

Thanks,
Chris Sparks
http://www.catalina-inter.net/mrada

Re: Assistance needed

[Robert B. Love]

In <35BCF867.E7EA0EF0@catalina-inter.net> "Chris Sparks (aka Mr. Ada)" 
wrote:
> to quote "BAD" examples of C++ usage on projects.
> 
> Examples I have seen are:
>   Navy plight with Windows NT
>   Colorado Airport baggage snafu

If memory serves, the language used in the Denver baggage system had
nothing to do with the problem.  The system was poorly designed, it
ran open loop and used timing to know when bags were in a certain area,
not a sensor.  As much as I'm pro Ada I can't see blaming the choice
of C++.  Double check this.

Of course you could say anyone silly enough to try this open loop
strategy would be silly enough to use C++ but I wouldn't go that
far.
----------------------------------------------------------------
 Bob Love                                   MIME & NeXT Mail OK
 rlove@neosoft.com                            
----------------------------------------------------------------

Re: Assistance needed

[Travis C. Porco]

In article <6pm0aq$8bu$1@uuneo.neosoft.com>,
Robert B. Love  <rlove@antispam.neosoft.com> wrote:
>In <35BCF867.E7EA0EF0@catalina-inter.net> "Chris Sparks (aka Mr. Ada)" 
>wrote:
>> to quote "BAD" examples of C++ usage on projects.

>> Examples I have seen are:
>>   Navy plight with Windows NT
>>   Colorado Airport baggage snafu

>If memory serves, the language used in the Denver baggage system had
>nothing to do with the problem.  The system was poorly designed, it
>ran open loop and used timing to know when bags were in a certain area,
>not a sensor.  As much as I'm pro Ada I can't see blaming the choice
>of C++.  Double check this.

>Of course you could say anyone silly enough to try this open loop
>strategy would be silly enough to use C++ but I wouldn't go that
>far.

Aside from the Joyner paper, which I'm sure you've seen, I don't know
of much.  There is very little good data on programming languages,
and a lot of hot air (computing is very emotional).  I only do small
time one-person stuff, where programmer time is at a premium, and I
learned a long time ago that (1) programming is extremely demanding,
and (2) anything that lightens the load is a big help.  

So I won't
do imperative programming in anything other than Ada; the attention
to readability and "human factors" in general in the language design
is a Good Thing!  (We need _a priori_ reasons for language choice,
since there is no good data; to me, one significant fact about Ada
is the fact that human factors played a conscious role in the design.)

When Ada83 first came out, a lot of hackers rejected it because of
its complexity, and association with the military (the notion of 
"private" types was particularly offensive to some people).  But a few
years later, after straining at this gnat, they swallowed a camel:
C--.  I made a real effort to learn C++, spending all together several
months, and it's just too difficult to bother with; the juice isn't
worth the squeeze.  That's just one person's self-selected testimonial,
so it's not worth diddly-squat as data, but there you have it.  

-- 
Travis      **standard disclaimers apply**

"The ultimate result of shielding men from the effects of folly
is to fill the world with fools."  --Herbert Spencer

Re: Assistance needed

[Travis C. Porco]

I apologize for this "language war" post; I have lurked here for years,
learning things, but usually abort this sort of post before sending it;
this time I hit send instead of abort.  Oh well, I meant every word of
it.  

In article <6pmanb$6qu$1@agate.berkeley.edu>,
Travis C. Porco <porco@stat.Berkeley.EDU> wrote:
>In article <6pm0aq$8bu$1@uuneo.neosoft.com>,
>Robert B. Love  <rlove@antispam.neosoft.com> wrote:
>>In <35BCF867.E7EA0EF0@catalina-inter.net> "Chris Sparks (aka Mr. Ada)" 
>>wrote:
>>> to quote "BAD" examples of C++ usage on projects.

>>> Examples I have seen are:
>>>   Navy plight with Windows NT
>>>   Colorado Airport baggage snafu

>>If memory serves, the language used in the Denver baggage system had
>>nothing to do with the problem.  The system was poorly designed, it
>>ran open loop and used timing to know when bags were in a certain area,
>>not a sensor.  As much as I'm pro Ada I can't see blaming the choice
>>of C++.  Double check this.
>
>>Of course you could say anyone silly enough to try this open loop
>>strategy would be silly enough to use C++ but I wouldn't go that
>>far.
>
>Aside from the Joyner paper, which I'm sure you've seen, I don't know
>of much.  There is very little good data on programming languages,
>and a lot of hot air (computing is very emotional).  I only do small
>time one-person stuff, where programmer time is at a premium, and I
>learned a long time ago that (1) programming is extremely demanding,
>and (2) anything that lightens the load is a big help.  
>
>So I won't
>do imperative programming in anything other than Ada; the attention
>to readability and "human factors" in general in the language design
>is a Good Thing!  (We need _a priori_ reasons for language choice,
>since there is no good data; to me, one significant fact about Ada
>is the fact that human factors played a conscious role in the design.)
>
>When Ada83 first came out, a lot of hackers rejected it because of
>its complexity, and association with the military (the notion of 
>"private" types was particularly offensive to some people).  But a few
>years later, after straining at this gnat, they swallowed a camel:
>C--.  I made a real effort to learn C++, spending all together several
>months, and it's just too difficult to bother with; the juice isn't
>worth the squeeze.  That's just one person's self-selected testimonial,
>so it's not worth diddly-squat as data, but there you have it.  
>
>-- 
>Travis      **standard disclaimers apply**
>
>"The ultimate result of shielding men from the effects of folly
>is to fill the world with fools."  --Herbert Spencer


-- 
Travis      **standard disclaimers apply**

"The ultimate result of shielding men from the effects of folly
is to fill the world with fools."  --Herbert Spencer