The necessity of CASE tools for system development is beyond dispute. The current generation of CASE tools, however, is too inflexible to provide adequate modelling support. One of the proposed solutions to this problem is the development of so-called CASE-shells . A CASE shell is a method independent CASE tool, which may be instantiated with a specific method to become a CASE tool supporting that method. As such, a CASE shell provides complete flexibility. This paper does not address the benefits of CASE shells, as they are completely clear, but focuses on the feasibility of this concept from a theoretical as well as a practical point of view. 1 Introduction CASE tools are currently considered to be an indispensable part of the systems engineer 's toolkit. Justification of CASE tools is no longer subject of debate. and [Flo86]). Naturally, the range of the required facilities should be thoroughly understood before using automated tools, see also [BS87]. In this reference an environme...

In this paper Green's (1989) notion of "cognitive dimensions" is used to consider the properties of notations and media for sketching. First, a cultural-cognitive approach to HCI research is briefly outlined which draws on some ideas from "distributed cognition" and Russian psychology. Rather than focussing on the cognitive aspects of an activity or the social aspects (as do cognitive psychology and ethnography, respectively) this approach attempts to describe the artifacts used in an activity in a cognitively and socially relevant way. It is argued that such an approach requires a framework within which to analyse external representations, and that Green's "cognitive dimensions" provide such a framework. The cognitive dimensions framework is described and then used to interpret a study of collaborative idea sketching which used interviewing, observational methods and a questionnaire survey of 128 academics. This analysis resulted in a cohesive and comprehensive characterisati...

The current generation of CASE tools is too inflexible to provide adequate modelling support. One of the proposed solutions to this problem is the development of so-called CASE shells. A CASE shell is a method independent CASE tool, which may be instantiated with a specific method to become a CASE tool supporting that method. As such, a CASE shell provides complete flexibility. This paper does not address the benefits of CASE shells, as they are completely clear, but focuses on the feasibility of this concept from a theoretical as well as a practical point of view. Keywords: Computer-aided software engineering (CASE), meta-modelling, early phases, CASE shell 1 Introduction Nowadays, the use of CASE tools as an indispensable part of the systems engineer's toolkit is common practice. It is believed that CASE tools are becoming the most important contributors to the continuing development in information systems development methods, see [ Avison and Fitzgerald, 1988 ] and [ Bubenko, 1988 ...

this paper, I illustrate this by considering cases in which patients produce unsolicited turns at talk 1 while a prescription is being issued, either to raise issues concerning the prescription per se or to introduce unrelated topics (concerning, for example, other complaints, social issues, or problems encountered by members of their family). Subsequently, I discuss some of the potential design implications of the findings and suggest that the linkage between human-computer interaction and face-to-face interpersonal communication raises important issues for the field of HCI