Arcadia Papers: ABSTRACT
"Type Evolution Support for Complex Type Changes",
by Barbara Staudt Lerner in Technical Report 94-071,
Department of Computer Science, University of Massachusetts at Amherst, 1994.
Abstract
Type evolution is a serious problem for programs that use persistent data.
Type changes are required during software maintenance, but can make persistent
data inaccessible. Existing systems that support type evolution severely
restrict the ways in which a type can be changed, thereby limiting what can
be accomplished during maintenance. Tess is a system to automate type
evolution for a collection of more complex type changes while providing
a more natural editing environment for the programmer. Tess uses a
comparative approach to identify type changes rather than a specialized
editing process. Experimental results on real data indicate that Tess can
accurately identify more complex type changes than those supported by
existing systems relying on specialized editing.
The Arcadia Project
<arcadia-www@ics.uci.edu>
Last modified: Thu Jan 26 11:51:31 1995