I am having some difficulty getting my head round something that should be carried out in the Elaboration phase;
"An executable architecture that realizes architecturally significant use cases" (Taken from Wikipedia: http://en.wikipedia.org/wiki/Rup)
I have done a lot of reading on the Internet as to what the 'Executable architecture' is meant to be. Some papers I have read suggest building a basic version of the system, which will then be built upon later in the Construction phase. Other papers suggest that it refers to the modelling of the system, using diagrams such as a Class Diagram.
The latter makes more sense to me, because otherwise it would seem that if I needed to develop something in the Elaboration phase, it is sort of overlapping with the Construction phase.
Basically, "Executable Architecture" is RUP's parlance for a description of the system's architecture that can be more or less directly converted into code and other low-level artefacts.
This is in opposition to how architectures are sometimes specified, i.e. as informal descriptions or box-and-arrow diagrams that are fine as human communication mechanism, but close to useless as far as the computer is concerned. In a model-driven scenario (which the OMG tries to push, of course), the system architecture must be specified in such a way that code, configuration files, deployment artefacts other lower-level models can be generated in an automated or assisted fashion by the application of tools.