Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!linus!decvax!duke!mcnc!ncsu!uvacs!mac
From: mac@uvacs.UUCP
Newsgroups: net.lang
Subject: CLU mutations
Message-ID: <1084@uvacs.UUCP>
Date: Thu, 1-Dec-83 15:17:46 EST
Article-I.D.: uvacs.1084
Posted: Thu Dec  1 15:17:46 1983
Date-Received: Sun, 4-Dec-83 04:24:21 EST
Lines: 14



Does CLU still have the mutable/immutable value distinction?  My impression
was that certain system-defined types (immutable) are copied on assignment,
but that structured data are assigned with pointers and shared (mutable).
Thus after A:=B operations on B can change A iff they are of a mutable
type.

This feature occurs in languages with pointers.  In Algol W record types
were always referenced with pointers, hence mutable.  I think it progress
that in Pascal the notion of record has been separated from that of pointer.

For that matter, I think early SETL (BALM-SETL) had this problem.  The wary
user had to do lots of copy().