Position Paper: Joint W3C/OMG Workshop on Distributed Objects and Mobile Code Bill Janssen, Xerox PARC (janssen@parc.xerox.com) The ILU project at PARC is an outgrowth of earlier research into distributed document systems, in particular the notions behind construction of universal browsers that could examine any document space, regardless of document type. It became clear that in order to do this, a clear language-independent notion of object types would be needed, supported by an underlying technology of distributed objects. We feel that this need is also present in the current notion of the World Wide Web, and that the notion of object types present in ILU, and the distributed object systems embodied in CORBA orbs, are reasonable responses to these needs. ILU (ftp://ftp.parc.xerox.com/pub/ilu/ilu.html) is a multi-language object interface system. The object interfaces provided by ILU hide implementation distinctions between different languages, between different address spaces, and between operating system types. ILU can be used to build multi-lingual object-oriented libraries ("class libraries") with well-specified language-independent interfaces. It can also be used as a distributed object system. ILU interfaces are specified in ILU's Interface Specification Language. ISL is not quite OMG IDL, as it provides several extensions to OMG IDL which we feel are crucial to providing effective implementations of things like the World Wide Web on top of a distributed object system. We see a number of opportunities for a CORBA-like object system to play an important role in the World Wide Web. Links between Web servers and specific services can be made faster and more flexible by replacing the current CGI API with a method invocation on an ILU object, as Digital Creations' ILU Requester has done (http://www.digicool.com/releases/ilurequester/). Browsers can export ILU interfaces to allow manipulation by other applications in the user's space. Proxy servers and cache servers can coordinate their databases using ILU-based protocols similar to those described by the Bayou project (http://www.parc.xerox.com/csl/projects/bayou/). Perhaps most interestingly, current Web usage could be radically improved by augmenting the current HTTP protocol with a carefully designed set of object types, and an optimized wire protocol for conveying method invocations and responses. We are currently participating in an experiment by the World Wide Web Consortium to devise such a set of types, and such a wire protocol (http://www.w3.org/pub/WWW/Protocols/HTTP-NG/Overview.html). This experiment, if successful, provides the opportunity to greatly expand the numbers and kinds of objects accessible via the Web, as well as changing our notions of what a ``document'' is. Some of these notions are already being explored by the Digital Library research initiatives, several of which have chosen to use ILU to tie together their systems. We hope to bring some of the ideas realized in this research to the Web, as well.