Programming language C vs c++ (was: [GRASS5] Want GNU libavl ?)

[David D Gray]

Bernhard Reiter wrote:

> On Mon, Jan 21, 2002 at 10:39:07PM +0000, David D Gray wrote:
> 
> [...]
>>But the truth is that few people _if 
>>any_ today in engineering, research, academia use C as the development 
>>language. It is almost universally C++ that is used. 
>>
> 
> I do not agree with the statement.
> And even if I did I call it a mistake to bet on C++.
> We can agree upon that this mistake is made by a large group of
> people, though. There is no reason to repeat it.
> 


Well I'm afraid I must repeat it again, because it is my regular 
experience. Though I do agree it is a mistake to bet on C++, and I think 
its wide adoption is to do with the political ramifications of the 
commercial software world and its central role for many years in Visual 
Studio. It's popular because it's just what's used in 'Software 
Engineering' circles. Of course among many engineering disciplines, the 
coding requirements are very low-level - on a par with kernel hacking, 
and here you will find C ; also in RT systems.


> Without going into the details I do not consider C++ a nice
> programming language. There are also general doubts regarding 
> the object-orientated approach in computer science now. 
> It concerns databases and programming languages.
> 


No disagreement here on the first point,  but i find that 
object-oriented features become a natural solution as you move towards 
more complex data structures. I also find it a real pain to spend so 
much of my development time taken up with trivial memory bugs, that 
really should be able to be handled by the language - but that takes us 
in a different direction from C++.


> For a pointer to skepticsm regarding C++ see one of the 
> better "language discussions":
> 
> 	http://www.advogato.org/article/207.html
> 
> 
> 


Well, these things have been repeated often, and are just brought 
together here in one place. They are perfectly true, but to repeat I 
don't advocate major adoption of C++ code in the GRASS core, we should 
just consider accomodating it now in the case of libraries that we would 
like to use and where it is found not to have major portability or 
stability problems.

Regards

David