Résumé: Text documents are a valuable resource for virtually any enterprise and organization, and in many cases, document processing is a highly collaborative process. Thus documents have to be stored, managed and handled with the utmost care and efficiency. Strangely enough, while other data ( most prominently customer, product, .nance, and the like) have long been supported by sophisticated database and work .ow technologies, until now, documents have typically been treated as second- class citizens, where mostly ad- hoc, home- grown, and often intricate solutions are being used. As a consequence, many of the achievements ( with respect to data organization and querying, recovery, integrity and security enforcement, multi- user operation, distribution management, uniform tool access, and similar) are not easily available for documents. We propose a radically di .erent approach, centered on natively representing text in fully- .edged data-bases, and incorporating all necessary collaboration support. This paper presents the overall concept and details of the approach ’s implementation. It also shows that such an approach ––against .rst- glance scepticism, something often used as a killer argument ––is indeed feasible with respect to efficiency.

Langue: Anglais