Client/Server based statistical computing (Q1855643)

The authors propose a client/server architecture for statistical computing similar to client/server architecture of database applications. This leads to architecture, where new statistical methods are stored in a method-base on the server. In addition the authors propose to extend the server by the ability to execute the methods. The main feature of this approach is the possibility to connect various client programs via a TCP/IP connection to a powerful statistical engine. For large flexibility a middleware program managing the communications between server and client is added. This general structure is illustrated by implementation of the XploRe Quantlet Client and XploRe Quantlet Server.