首页 | 本学科首页   官方微博 | 高级检索  
     


Automatic deletion of obsolete information
Authors:A.N. Habermann
Affiliation:Carnegie-Mellon University U.S.A.
Abstract:Information generated in a programming environment is often allowed to grow indefinitely. Designer and user alike are counting on standard backup and disc clearing procedures for archiving old data. In this paper we take the view that one should distinguish between relevant old data that is purposely archived and obsolete information that should automatically be deleted. The two main topics of the paper are the strategies and mechanisms for deleting information and the facilities available to designers of programming environments to specify deletion strategies. Information can be deleted applying a passive or an active strategy. With the passive strategy, information will not actually be deleted until it is certain that there is no interest in it any longer. With the active strategy, an object is immediately deleted when it becomes obsolete, while users of the object are notified of the deletion event. The paper discusses various implementations for these two strategies and shows when they apply. Taking the view that it must be easy to modify and fine tune programming environments, much attention must be given to the designer's support environment for generating programming environments. The paper discusses in particular the facilities for expressing the semantics of names in an environment. Various naming modes are useful for a designer to specify the deletion strategies for his target programming environment. Details are illustrated by applying the ideas to an environment for software development and maintenance.
Keywords:Address Correspondence to Dr. A. Nico Habermann. Computer Science Department   Carnegie-Mellon University   Pittsburgh   PA. 15213.
本文献已被 ScienceDirect 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号