For the full presentation, click here.
An unscientific survey of FAQs, product literature and articles in the popular press about XML, as well as the developer-oriented discussion groups, indicates that XML is perceived in about equal divisions as a tool for data exchange between platforms, or as the basis for a data definition/data manipulation/data repository platform itself. My argument is that truly XML-based processing -- which I will further argue is required to handle all of the problems that arbitrarily extensible markup admits -- must be both integral to the platform and also the form of inter-platform interaction. An IT design which is purely XML wherever possible will, in fact, display four peculiar characteristics: it will follow instance markup in the definition of data structure, in preference to any schema external to that markup -- highly unusual in database or other data manipulation software; it will rely on uniquely client-side definitions of data structure and of processing in order to specify the exact behavior required; it will depend on XML-described structures and links to define processing flow and procedural or conditional logic, in absolute preference to implementing any such in executable code; and it will conduct all interprocess, inter-platform interaction through XML documents which will be uniquely parsed on every occasion within the context receiving them, in preference to relying on defined interfaces or structures. Failing to exhibit these four characteristics, a solution or design falls short of what can be achieved with XML, and as a practical consequence will in some circumstances be unable to handle well-formed documents structured with arbitrarily extensible markup, which I take to be the acid definition of XML processing capability.
I propose to illustrate in detail the nature of each of these four characteristics. I will show in terms both of overall design and of detailed processing problems why each is necessary. I will similarly demonstrate how, both in overall design and in the specific details of processing, how these four interact and are interdependent. I will also illustrate, in the same overall as well as detailed terms, how all four are inherent in the nature of XML, and are inherently the basis for XML processing. I propose then to get into the nitty-gritty of implementing these characteristics and, in the aggregate, implementing a fully XML-capable processing system. The implementation will be done in Java, but some processes will be first blocked out in pseudo-code, and I will make the effort to prevent Java-specific details from obscuring the more general applicability of the implementation.
About the speaker
Walter E. Perry, PhD. is a founder and the CTO of net.uniqueness, Inc., a New York- and London-based firm applying XML to database functions and to databased application support. He has sixteen years experience developing enterprise systems which apply distributed databased solutions in financial settlements and other transnational processing.