Automated manufacturing processes with the ability to translate digital models into physical form promise both an increase in the complexity of what can be built, and through rapid prototyping, a possibility to experiment easily with tangible examples of the evolving design. The increasing literacy of designers in computer languages, on the other hand, offers a new range of techniques through which the models themselves might be generated. This paper reviews the results of an integrated parametric modelling and digital manufacturing workshop combining participants with a background in computer programming with those with a background in fabrication. Its aim was both to encourage collaboration in a domain that overlaps both backgrounds, as well as to explore the ways in which the two working methods naturally extend the boundaries of traditional parametric design. The types of projects chosen by the students, the working methods adopted and progress made will be discussed in light of future educational possibilities, and of the future direction of parametric tools themselves. Where standard CAD constructs isolated geometric primitives, parametric models allow the user to set up a hierarchy of relationships, deferring such details as specific dimension and sometimes quantity to a later point. Usually these are captured by a geometric schema. Many such relationships in real design however, can not be defined in terms of geometry alone. Logical operations, environmental effects such as lighting and air flow, the behaviour of people and the dynamic behaviour of materials are all essential design parameters that require other methods of definition, including the algorithm. It has been our position that the skills of the programmer are necessary in the future of design. Bentley’s Generative Components software was used as the primary vehicle for the workshop design projects. Built within the familiar Microstation framework, it enables the construction of a parametric model at a range of different interfaces, from purely graphic through to entirely code based, thus allowing the manipulation of such non-geometric, algorithmic relationships as described above. Two-dimensional laser cutting was the primary fabrication method, allowing for rapid manufacturing, and in some cases iterative physical testing. The two technologies have led in the workshop to working methods that extend the geometric schema: the first, by forcing an explicit understanding of design as procedural, and the second by encouraging physical experimentation and optimisation. The resulting projects have tended to focus on responsiveness to conditions either coded or incorporated into experimental loop. Examples will be discussed. While programming languages and geometry are universal in intent, their constraints on the design process were still notable. The default data structures of computer languages (in particular the rectangular array) replace one schema limitation with another. The indexing of data in this way is conceptually hard-wired into much of our thinking both in CAD and in code. Thankfully this can be overcome with a bit of programming, but the number of projects which have required this suggests that more intuitive, or spatial methods of data access might be developed in the future.