Laserfiche WebLink
<br />Table 10: Alternative Architectures Summary <br /> <br />e <br /> <br />Function Architecture/ Distributed Centralized <br />Organization Structure <br />Dedicated NO! recommended Feasible <br />Dala Centered Recommended Recommended <br /> <br />Table 10 contains the only other relevant interaction. namely that between the <br />orgllJ1izational ..component of the access architectures and the functional <br />architecture of CRDSS, From Table 10 it is evident that a disoibuted <br />organizational structure does not satisfactorily combine with a Dedicated DSS <br />architecture. The reason for this is that the solution is simply impractical for <br />logistic reasons. Dedicated systems are typically stand-alone systems, <br />Distributing their databases or system updates involves a lot of maintenance <br />and logistic management. Dedicated DSS and Disoibuted organizational <br />structure therefore do not go together well. <br /> <br />Summarizing. a Data Centered Architecture is recommended over a Dedicated <br />DSS approach. The Data Centered approach allows far more flexibility <br />regarding extensibility of the system and is characterized by a high modularity <br />of implementation, We recommend the use of forms-based interfaces for status <br />and reporting on platforms which do not support interactive graphics interfaces <br />and the use of interactive graphics in all other cases. Both centralized and <br />distributed access modes are feasible. the choice being more of an <br />organizational than a software development issue. However. although both <br />dedicated DSS and Data Centered systems go weJl with a centralized approach, <br />we do not recommend a Dedicated DSS architecture in a disoibuted access <br />context. <br /> <br />e <br /> <br />Alternative System Development Cost Estimates <br /> <br />For this task the costs of alternative architectures were systematically <br />estimated. These estimates are based on the results of Tasks 2 through 5 (e.g. <br />model evaluation. data determination. evaluation of alternative user interfaces, <br />and evaluation of alternative architectures), <br /> <br />The total cost of the various architectures is the sum of the estimated coSts for <br />six development tasks: <br /> <br />1. Development of the database. Since we recommend a Data Centered <br />Architecture for CRDSS, considerable effort must be spent on the <br />development of the central database. <br /> <br />2. Development/enhancements or models. From the models inventory <br />(Task 2) it can be concluded that the inclusion of models into CRDSS <br /> <br />e <br /> <br />DAMES& M OORE/CADSWES-45 <br />