Tonge, Rzepa, Yoshida, page 1

Proofs to: H. S. Rzepa, Dept Chemistry, Imperial College, London, SW7 2AY.
------------------------------------

Authentication of Internet-based Distributed Computing Resources in Chemistry.

Alan P. Tonge,

^aHenry S. Rzepa^a*and Hiroshi Yoshida^b

(a) Department of Chemistry, Imperial College of Science, Technology and Medicine, London
SW7 2AY. (b) Department of Chemistry, Faculty of Science, Hiroshima University,
Higashi-Hiroshima 739-8526, Japan

Abstract:The evolution of the World Wide Web from a model allowing only the public and open exchange of information on the global Internet viasimple HTML-based pages, to one involving the additional development of Intranets or Extranets with secure interactive client- server processes is discussed. These client-server processes can be customised on the client browser by the use of embedded applications such as Java Applets that allow the client to communicate with other remote applications or databases via a distributed computing architecture. We show¹how software authentication using digital object signing procedures in conjunction with X.509 certificates was used to develop three authenticated distributed chemical applications. COS (Chemical Object Store) is a database application based on Java Remote Method Invocation, MoldaNet invokes Java3D to create a molecular visualisation tool and JSpec delivers analytical spectral data. We argue that such authentication of chemical resources on the Internet provides one mechanism for increasing the perception of quality and integrity of molecular information disseminated within the chemical community and creates an architecture for electronic commerce to develop in the molecular science community.

Introduction
The dominant computing model during the 1970s and 1980s was of a mainframe computer with a centralised CPU located in a special room and providing computing services to local dumb display screens. This model was superseded by the advent of powerful and relatively inexpensive desktop workstations in the late 1980s, which permitted the introduction of the so-called interactive client-server or network distributed computing model. This allowed applications to be split amongst multiple processors, with actions on one local machine creating processes on a second remote machine, for example a computationally intensive scientific calculation or a database query. The separate widespread deployment of Internet technology using TCP/IP networking protocols in the mid 1990's saw the introduction of another model, involving addressable information pages marked up using HTML (Hypertext Markup Language) and served by a HTTP (Hypertext Transport Protocol) server, the so- called World-Wide Web model.²The server in this architecture is a networked computer, which can receive HTTP requests from any TCP/IP compliant computer in the world, and can returns appropriate pages to be accessed and viewed by a suitable Web browser at the remote client.

These two distributed computing and publishing models are now fusing in the late 1990s into a more closely integrated system. Thus Web browsers have developed from being capable of simply displaying static HTML pages to also having the ability to become interactive clients