Path: csiph.com!usenet.pasdenom.info!weretis.net!feeder4.news.weretis.net!feeder2.ecngs.de!ecngs!feeder.ecngs.de!Xl.tags.giganews.com!border1.nntp.ams.giganews.com!nntp.giganews.com!local2.nntp.ams.giganews.com!nntp.brightview.co.uk!news.brightview.co.uk.POSTED!not-for-mail
NNTP-Posting-Date: Thu, 21 Jun 2012 17:23:59 -0500
From: Gareth Owen <gwowen@gmail.com>
Newsgroups: comp.lang.c
Subject: Re: Design decisions
References: <jrvvlb$8k1$1@speranza.aioe.org> <js025k$hjm$1@dont-email.me>
Date: Thu, 21 Jun 2012 23:23:59 +0100
Message-ID: <87a9zwb8xs.fsf@gmail.com>
User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/23.1 (gnu/linux)
Cancel-Lock: sha1:dYs6mS4ZHrAMe/ICQNHU9f015Fg=
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Lines: 26
X-Usenet-Provider: http://www.giganews.com
X-Trace: sv3-2FF91hVkjspYjpCXwNicY9InIttfQyBlsjOXQv4hKUdq5j9awEagO9KUWOH9aPujWZqIR3iwPikXhIQ!mszBB3ZlYxluakHalDV7zV40jf6DCLt1kaqDReEcKJS9t6NEFKF8YbbpRsu/1j8+uCzZ
X-Abuse-and-DMCA-Info: Please be sure to forward a copy of ALL headers
X-Abuse-and-DMCA-Info: Otherwise we will be unable to process your complaint properly
X-Postfilter: 1.3.40
X-Original-Bytes: 2029
Xref: csiph.com comp.lang.c:22507

"BartC" <bc@freeuk.com> writes:

> How does C++ deal with it?

Most (all?) C++ STL containers give a bounds-checked accessor that
returns a reference on success, but throws an exception on out-of-bounds
access (I assume that's the main failure condition Jacob has in mind for
GetElement also).  

Exceptions give you a return path which, when combined with well
encapsulated objects, automatically cleans up any resources.  Of course,
writing such exception-safe objects is a non-trivial undertaking.

jacob navia <jacob@spamsink.net> writes:
> Before returning, the "GetElement" API calls the error function and
> will return ONLY IF the error function returns.

What do you have in mind for a well designed, non-returning error
function.  Obviously one that calls abort()/exit() would work but what
else? 

Could the error function provide a default value (e.g. out-of-bounds
access always returns a 0 value, or attempts to expand the vector and
pad with zero)?

Or were you thinking in terms of longjmp()ing to some prior state?