Managing Tasks in the O ce Environment

First we describe our experiences with document processing support. Then we consider the similarities between document processing and work-ow management and the relation of both to ooce information systems. In many cases ooce tasks create, manipulate and disseminate documents. While the evolution of modern ooce technology enables people to process documents more eeciently, the same technology creates new problems and challenges. Electronic documents are generated electronically using word processors or graphics programs, distributed electronically across local and wide area networks, led electronically in le systems or data bases. An example of the dissemination of electronic documents is World Wide Web. At the Xerox Webster Research Center we designed and implemented the experimental document management system FolioPub 1]. It is based on a distributed architecture supporting a client/service model, with FolioPub servers providing document management services. FolioPub supports document dee-nition, version and access control, sequencing of documents through multiple text and image processing steps, and communications support for distributed document deenition and processing. FolioPub focuses on document deenition tasks and on automatic execution of system tasks in a distributed environment. The extensive internal use demonstrated where FolioPub goals were achieved and where the system had shortfalls. One of the main negative experiences was the use of a single, tightly bound tree structure which captured both: document content and document processing. A tree node has a twofold semantics: a structural element of the document, e.g. a chapter or a section, and a document processing task, e.g. the execution of the text formatting tool. Changes to the document content structure cause modiications in the document processing tasks. Also, the integration of additional tasks creates artiicial nodes in the document content structure. A separation provides greater exibility in applying diierent types of document processing to the same document without modifying the document structure stored in the data base. Associated with one document may be one or several task networks which deene the processing that may be applied to the document.