NetNews Offline 2/NetNews Offline Volume 2.iso/news/comp/lang/c++-part2/16905

Internet Message Format | 1996-08-05 | 1.7 KB

Path: fido.asd.sgi.com!news
From: [e-mail] (Matt Austern)
Newsgroups: comp.lang.c++
Subject: Re: Why don't you use garbage collection
Date: 12 Apr 1996 16:56:52 GMT
Organization: SGI
Distribution: world
Message-ID: <[e-mail]>
References: <[e-mail]> <1996Apr12.170708.1835@ittpub>
Reply-To: [e-mail]
NNTP-Posting-Host: isolde.mti.sgi.com
In-reply-to: [e-mail]'s message of 12 Apr 96 17:07:07 WET

In article <1996Apr12.170708.1835@ittpub> [e-mail] (Wil Evers) writes:

> In my opinion, this subtle difference is why garbage collection would be
> appropriate for C, but not for C++. C++ destructors are often used to
> release other resources than memory. If C++ programmers would get sloppy
> about deleting their heap objects ("don't worry about deleting it, the
> collector will do it for you"), we might end up with too many open file
> descriptors, eternal database locks, forever blocking semaphores and
> expensive telephone bills.

In C++, I think the most reasonable scenerio is where some objects,
but not all, are garbage collected.  You might, for example, inherit
from some GC_Base mixin class if you want garbage collection; GC_Base
would overload operator new to do the appropriate magic.  And it's
entirely reasonable to imagine that you would have to accept some
restrictions (such as restrictions on destructors) in exchange for
garbage collection.

In C++ we're used to the idea that heap-allocated and stack-allocated
objects have different rules.  It's not so radical to introduce
garbage-collected objects as a third category.
--
Matt Austern
SGI: MTI Compilers Group
[e-mail]