Abstract Modern enterprise systems are likely to have a very long life. Their specifications therefore need to employ mechanisms that allow them to evolve during their lifetime; where they exploit generic components, these must be adaptable for use in novel situations. The paper looks at some of the issues that arise from this requirement, and how the exploitation of domain-specific language technologies in the tool-chain can assist in maintaining consistency of the specification as a whole. First, it reviews the final state of the family of standards supporting the ODP Enterprise Language, which is intended to handle this kind of application. In particular, it looks at the way the framework for defining policies can be used to accommodate changing requirements during the lifetime of an evolving system. It also looks at the way the idea of deontic tokens enables factoring out of the management of obligations from the basic behaviour of interacting system components. It then proposes a roadmap for building tools that can be used to unify the constraints from different areas of concern into a single specification. The approach taken is to exploit the power of domain-specific languages (DSLs) to allow designers in the various areas of concern to provide their input in terms natural to them. Finally, it looks at the way this approach promotes the establishment of a robust tool-chain capable of handling the evolution and scalability of enterprise systems. The paper uses a running example from the e-health domain to show how specific areas identified in the e-health standards can lead to language definitions, and so to tooling, that can be used to manage unified, system-wide specifications.

There has, in recent years, been growing interest in model driven engineering (MDE), in which models are the primary design artifacts and transformations are applied to these models to generate refinements leading to usable implementations over specific platforms. There is also interest in factoring out a number of non-functional aspects, such as security, to provide reusable solutions applicable to a number of different applications. This paper brings these two approaches together, investigating, in particular, the way behaviour from the different sources can be combined and integrated into a single design model. Doing so involves transformations that weave together the constraints from the various aspects and are, as a result, more complex to specify than the linear pipelines of transformations used in most MDE work to date. The approach taken here involves using an aspect model as a template for refining particular patterns in the business model, and the transformations are expressed as graph rewriting rules for both static and behaviour elements of the models.

If e-business contracts are to be widely used, they need to be supported by the IT infrastructure of the organizations concerned. This implies that the interactions between systems in different organizations must be guided by the contract and there must be sufficiently strong checks and balances to ensure that the contract is in fact obeyed. This includes facilities for the unbiased monitoring of correct behaviour and the reporting of exceptions. One of the ways to provide this support is to generate it directly from the agreed contract. This paper considers the steps necessary to provide sufficient automation in the support and checking of e-Business contracts for them to offer efficiency gains and so to become widely used. It focuses on the role of models, taking a model-driven approach to development and discussing both the source and target models and the transformational pathways needed to support the contract-based business processes.

This paper compares two separately developed systems for monitoring activities to business contracts, describes how we integrated them and exploits the lessons learned from this process to identify a core set of requirements for a Business Contract Language (BCL). Concepts in BCL needed for contract monitoring include: the expression of coordinated concurrent actions; obliged, permitted and prohibited actions; rich timeliness expressions such as sliding windows; delegations; policy violations; contract termination/renewal conditions and reference to external data/events such as change in interest rates. The aim of BCL is to provide sufficient expressive power to describe contracts, including conditions which specify real-time processing, yet be simple enough to retain a human-oriented style for expressing contracts.

There has been a growing interest in recent years in the use of abstract building blocks in system specification. Designs based on patterns and communities are two examples. However, these structures are then refined further during design and implementation, and it is often difficult to determine whether the eventual system implementation is a faithful reflection of the original properties of the pattern specified. This is particularly true of patterns used to describe an enterprise view of the system. The paper concentrates on the behavioural aspects of pattern specification, and investigates the way that observation of the system can be interpreted to check that properties of the pattern specification are preserved. It describes a diagnostic tool that checks the actual system behaviour against the pattern specification, and discusses the requirements this places on the form of the specification language and a number of the problems of interpretation that arise in applying such tools.

The ODP Enterprise Language allows the rules and policies that characterize an organization to be brought together and used to guide the various stages of system design, development and operation. UML is one of the leading notations for system design and is likely to be the basis for a wide range of design tools. However, UML has a comparatively weak set of facilities for supporting the combination of existing, parameterized specifications and, in particular, for defining and managing policies. This paper discusses the requirements for defining communities and expressing policies within a UML environment, compares ways in which the existing notation might be used in enterprise specification, and indicates some of the implications this would have for system development tools.

The approach taken by earlier file transfer protocols and the design decisions resulting from them are discussed. A brief review of the history of file transfer protocols and the development of the concepts on which they rely is given. References are made to the seminal protocol descriptions, although not to all subsequent updates of the protocols concerned. The modeling of filestores and file structures which form the basis for the meaning of the protocol exchanges is outlined. The concept of a document type is introduced to bring together the many and various aspects which characterize the definition of a particular class of information. The resultant International Standards Organization protocol provides a range of facilities for different styles of application, covering both simple transfer and access to shared information. Current activities to extend the international standard and identify common functional styles of usage are reviewed, and the current situation is summarized.>