NetNews Offline 2/NetNews Offline Volume 2.iso/news/comp/lang/c++-part2/18037

Internet Message Format | 1996-08-05 | 3.9 KB

Path: news.sprintlink.net!datalytics!usenet
From: Rob Stewart <[e-mail]>
Newsgroups: comp.lang.c++,rec.games.programmer,alt.msdos.programmer,comp.programming
Subject: Re: Ada 95 vs C++ (was: Young programmers read me.)
Date: Thu, 18 Apr 1996 12:45:16 -0400
Organization: Datalytics, Inc
Message-ID: <[e-mail]>
References: <4icpp9$[e-mail]> <4imqe4$[e-mail]> <[e-mail]> <4j52hn$[e-mail]> <[e-mail]> <[e-mail]> <[e-mail]> <4kv046$[e-mail]> <4l51al$[e-mail]>
NNTP-Posting-Host: 204.62.224.71
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
X-Mailer: Mozilla 2.0 (WinNT; I)

Peter Joachim Unold wrote:
>
> Thus spake [e-mail] (Todd Coniam):
>
> [cut of someone's praise of c++]
> >--
> >While experience in the commercial world has proven time and time again that
> >programmers are not capable of capturing most bugs in the program when using
> >C/C++ until well into the development.  C/C++ is inherently one of the most
> >error prone languages around.  Dangling pointers and memory leaks are all
> >slip into code much too easy, and are horrible to try and track down.  The
> >folks at Purity Software(C) are making money hand over fist selling products
> >that help with just these problems.  Perhaps you've heard of 'Purify'?
>
> Sure we have. While dangling pointers can be a hell to find, I never understood
> why memory leaks should be that horrible. Especially not in a language like
> C++ where resource allocating/freeing can be dealt with in such a nice way.
> (see Stroustrup's 'the c++ programming language').
> Also providing your own global new/delete operators, which takes extra debug
> arguments is quite easy.
>

There are many programming idioms that preclude stupid mistakes.
The point is that C++ gives you the freedom to do a great deal,
including hang yourself.  That freedom can be used to write
good, efficient programs too, don't forget.

> >If C++ is soo great why do most major developments severly limit the features
> >that they allow used in the project.  Templates and multiple inheritance
>
> Huh? What major C++ projects doesn't use templates? I can understand why
> anyone wouldn't use the most advanced new features, but banning good old
> ARM defined templates is just stupidity. As for MI, it depends entirely
> on your design, wether you use it or not.
>

There are only rare circumstances in which multiple inheritance
is appropriate.  MI can be quite troublesome, so it stands to
reason that most C++ developers would avoid it.

> >(C++'s _great_ features) are usually the first to go.  Wouldn't you rather
> >have the compiler tell you that your code is crap before you try and run it?
>
> Sorry? C++ is a static checked language.
>

In other words, the compiler catches all sorts of things for us.

> >Try Ada 95, all the benefits of easy to understand syntax (you can read
> >tommorow or next year!) with all the great object orientated features you want
> >along with full multi-tasking in the language!
> >Check out the web sites below and see for yourself.  You won't be sorry.
>
> OK thanks. We'll all check it. But you'd be a better advocate of Ada 95 if you
> provided us with some arguments of why Ada 95 is better, instead of stupid
> c++ flames.
>
> I've substituted your Ada 95 URL with an URL to the latest public ansi c++ draft.

Actually, your URL was to the April 1995 DWP.  Let's point him
to the January 1996 DWP instead:
http://www.bby.com.au/%7Egnb/wp/jan96/index.html

> There you can learn some more about the language before you flame the next time.
>
>  best regards
>    peter unold - [e-mail]
>
> c++ draft: http://www.cygnus.com/misc/wp/draft/

--
Robert Stewart        | My opinions are usually my own.
Datalytics, Inc.    | [e-mail]