Xref: utzoo comp.lang.scheme:2138 comp.lang.lisp:4611 comp.lang.smalltalk:2740
Path: utzoo!news-server.csri.toronto.edu!cs.utexas.edu!swrinde!zaphod.mps.ohio-state.edu!usc!apple!mikel
From: mikel@Apple.COM (Mikel Evins)
Newsgroups: comp.lang.scheme,comp.lang.lisp,comp.lang.smalltalk
Subject: Re: implementing an interpreter
Message-ID: <50278@apple.Apple.COM>
Date: 15 Mar 91 00:01:43 GMT
References: <1991Mar8.153120.19836@tripos.com> <50211@apple.Apple.COM> <3653@paradigm.com>
Distribution: comp
Organization: Apple Computer Inc., Cupertino, CA
Lines: 16

In article <3653@paradigm.com> gjc@paradigm.com writes:
>In article <50211@apple.Apple.COM>, mikel@Apple.COM (Mikel Evins) writes:
>> 
>> A third approach is the Big Bag of Pages (BiBoP) arrangement, in
>> which objects of various types are each allocated only in certain
>> ranges of addresses. In this way, an object's type can be
>> determined by examining its address. This scheme somewhat complicates
>> an already somewhat complicated gc algorithm if you plan to
>> use, let us say, generation scavenging, but type determination
>> is fairly cheap.
>
>No heaper than the smalltalk technique that the poster mentioned.
>(Because of the extra shift. Although you may need a shift anyway).

Sorry, I didn't mean to imply that BiBoP was cheaper, only that it
was an, as yet unmentioned, alternative.