The introduction of HLA (High Level Architecture) is expected to facilitate the reuse of models and their co-operation in common simulations. That is valuable in general, but in some situations other methods would be better suited. In the choice between distributed HLA federations and monolithic models based on simulation frameworks one has to consider which aspects that are most important in the current modelling case. The main choice guiding aspects are the time scale of the project, the scalability, the validation and verification, the user friendliness, the performance, the security, the potential co-operation between organisations or nations and the long term strategy of the model investments. For certain applications HLA in itself could be seen as too rigid and heavy. A comparative study of the RTI (Runtime Infrastructure) of the HLA and an English prototype of a more dynamic product, GRIDS (Generic Runtime Infrastructure for Distributed Simulation), has been initiated. A co-operation has been initiated with the National University of Singapore with the intention to test the process model for development and execution of HLA federations, FEDEP, in a long distance international co-operation. Besides it will extend the understanding of how interoperability between models of different types will work.