Engineering flexible World Wide Web services

World Wide Web service and application engineering is a complex task, comparable to the software engineering process. Similar to CASE environments in software engineering, web service editors and web document designers are successful products on the market. Most of the tools concentrate on single web page creation; only a few manage the organization of complete web services. Large web applications containing hundreds or thousands of documents and complex interactive services need a more sophisticated engineering approach. Basic requirements for a web application management system are proper organizations of both the data and the navigation model. Consistent interface design and integration facilities for static information and dynamic interactive services identify a good engineering toolset. A real challenge to the engineering task is the introduction of flexibility to the web applications in terms of content changes, layout design updates, temporal information changes, multilinguai support and online content management. We discuss the benefits of databases to keep the web service information and the use of the HTML++ toolset to map the database contents to World Wide Web documents. Databases are used to introduce an abstract level of data management for web applications. The object-based HTML++ toolset is used to engineer complex web applications like the Vienna International Festival WWW presence. This case study contains 300+ multilingual pages and several interactive services. The combination of database support and HTML++ introduces a high level of flexibility to the festival's web service engineering. 1. I N T R O D U C T I O N Since the development of the lntemet. Ihe World Wide Web is the most popular application using ~his global Permission to make digital/hard copy of all or part of this work tbr personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage, the copyright notice, the title of the publication and its date appear, and notice is given that copying is by permission of ACM, Inc. To copy otherwise, to republish, to post on servers or tO redistribute to li~t~_ reolllre~ nrictr ~n~ . t ' i f i r n a t m i © e ; a n a n , 4 / n . ~ te~* information infrastructure. Due to its immense growth rate, millions of computers are currently connected and myriades of information packets are shared amongst the distributed systems and their users. Hundreds of thousands web servers offer interesting multimedia information for all kinds of user interests and publicly available browser allow anyone to connect to those sources. Web surfers download data directly or employ search engines that provide lists of information sources containing the individually requested data. The available information is that rich that service providers are forced to use innovative technologies and ideas to stand out of the competition. When Tim Berners-Lee[3] introduced his HTrP protocol and the first WWW servers in 1993, page after page of new information was added to a rapidly growing number of hosts. The first generation of web services, simply showing static HTML pages[2] is already outdated and multimedia contents and active components are manifested as state-of-the-art. Database interaction with the WWW provides a proper technology basis for active and dynamic information services that are able to show up-todate news and temporally changing data without loosing the manageability of the contents. Managing complex web applications is not well supported by current WWW development tools. The generation of static HTML pages with commonly available web editors or program implementations to create pages on the fly J from external sources like files and databases or as a reply to user defined form input parameters is a rather time consuming, but conceptually minor part of WWW service engineering. In the hypertext research area web se~'ice information needs to be well organized and managed ideally following a methodology such as RMM[8.10], Tomas lsakowitz et. al, describe in their methodology how to organize the data set, portion the information, design the navigation model and implement the web service. Parts of the steps proposed in the RMM methodology like the strueturization of the data and the service implementation can be solved with current web editors. There also exist tools [11,17] that integrate information from database~, h,, interpreting retrieving commands based on extended HTML • This work was supported by Hewlett Packard Laboratories. P.~l,,h A l t n a n , t lb .* V i P n n a F e ~ : l i v : l l i "~orr l f ' t l i l r t '¢ • pages. These tools provide the presentation of database contents whereas the online management of the data is rarely solved through proprietary systems. In this paper we present a flexible approach to engineering complex web applications: an object-based langnage is used together with latest database technology to provide easy to manage web services. As opposed to related WWW applications our system provides online content management and high flexibility to information changes. As an example for a modern service we discuss the web presence of the Wiener Festwochen 97112] (Vienna International Festival), implemented by the Distributed Systems Group at TU Vienna (DSG). The innovation used in the engineering process of this website is the combination of relational database technology and a web service engineering tool also developed by DSG that introduce a new degree of flexibility to web service engineering. Aside the beneficial employment of the HTML++[I] toolset for managing the hypertext information system, this web application explains the strength of databases for data management and online content manipulation. The interface design of the service can be updated quickly and easily and content changes can be done online by accessing database contents from a restricted management platform using standard WWW browser. Layout changes are applied consistently throughout the whole information set and hypertext links are implicitly handled in the abstract object definition language of our engineering system. This paper is structured as follows: Section 2 describes the benefits of using databases for the data organization and content management of WWW services. Section 3 provides a discussion of our web service engineering tool used to implement the Wiener Festwochen web application (WFA). In Section 4 we concentrate on the benefits that are gained by combining a database and HTML++, the web service engineering toolset. In the conclusion we discuss the benefits of our approach as well as future plans to improve HTML++.