Path: utzoo!utgpu!jarvis.csri.toronto.edu!rutgers!att!ulysses!andante!alice!bs
From: bs@alice.UUCP (Bjarne Stroustrup)
Newsgroups: comp.lang.c++
Subject: Re: C++ design
Summary: aims
Keywords: information
Message-ID: <9475@alice.UUCP>
Date: 13 Jun 89 19:40:41 GMT
References: <9474@alice.UUCP>
Organization: AT&T Bell Laboratories, Murray Hill NJ
Lines: 27



Given the widespread rumers (sic) to the contrary, it might be worth
pointing out that C++ was designed with a careful consideration of
principles such as type safety, consistency, encapsulation, orthogonality,
minimalism, and completeness. These ideals are, in particular, the reason
why C++ could not simple preprocessor adding a few features to an otherwise
untouched C. The need for a coherent and unified type system precludes that.

The stated aim for C++ (see the first paragraph of my book) is to ``make
programming more enjoyable for serious programmers.'' It does so by directly
attacking the problems that can make life miserable for a programmer or
a group of programmers: by providing support for good design, by allowing
ideas to be expressed directly in the language, by providing checking of
programs, by allowing the programmer to provide sufficient structure in
a program to make tools effective, by making good design and programming
techniques affordable, and by making good design and programming techniques
available on essentially all kinds of systems.

This may not have been stated clearly enough in the past. At least I have
seen postings presenting the view that compatibility was C++'s major or
even C++'s only goal. It wasn't and isn't. Compatibility provides a way of
co-existing with more traditional systems, a path for people who cannot
make immediate radical breaks with the past to follow, a way of preserving
skills, a way of re-using tools and libraries, and a way of guaranteeing
completeness of a new language.