home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.lang.ada,comp.lang.c,comp.lang.c++
- Path: in1.uu.net!world!bobduff
- From: bobduff@world.std.com (Robert A Duff)
- Subject: Re: C/C++ knocks the crap out of Ada
- Message-ID: <DoBF01.ACF@world.std.com>
- Organization: The World Public Access UNIX, Brookline, MA
- References: <JSA.96Feb16135027@organon.com> <313D4D00.875@ix.netcom.com> <313D94FB.1599@lfwc.lockheed.com> <4hl0g6INNdki@keats.ugrad.cs.ubc.ca>
- Date: Fri, 15 Mar 1996 15:28:49 GMT
-
- In article <4hl0g6INNdki@keats.ugrad.cs.ubc.ca>,
- Kazimir Kylheku <c2a192@ugrad.cs.ubc.ca> wrote:
- >But those checkers _are_ available, and do work. There are a lot of very good
- >such freeware tools, by the way.
-
- Such add-on tools for C have some problems, though:
-
- 1. There are a lot of things that simply can't be expressed in C, or
- can't easily be expressed, except in comments, so it's impossible to
- write a tool that checks these things. Ada isn't perfect here, either
- -- Eiffel has a richer assertion facility.
-
- 2. Such checks are sometimes extremely expensive. For example, to do
- array bounds checking in C, you have to store about 3 extra words of
- dope with every pointer. Ada's array bounds checking is much cheaper
- than that.
-
- >Those checks are valuable. Ada is backed by the right idea.
- >
- >On the other hand, those checks are not omnipotent.
-
- Indeed. Down with fanatics!
-
- - Bob
-