What do throat lozenges and OO modelling have in common?
In particular, what do Juanolas and whole/part relationships in class diagrams have in common? I guess you’ll understand if you follow the links, or if you are familiar with both Juanolas and OO modelling.
Let me give you some background.
I am participating in a project where a formal language for the description of heritage buildings is being developed. The project team is made of two archaeologists, an architect, and me as the software guy. They are the domain specialists; I am the techie. From the very beginning, I started using class diagrams to capture the information that my three colleagues would consider relevant, and they seemed to like them. They got themselves a copy of Visio and started producing their own class diagrams. I gave them a quick description of what specialisation, whole/part relationships and associations mean, and showed them how to use different symbols to depict them. No, I didn’t use UML symbols. UML is broken and invariable drives you to irresoluble quandaries, so I wouldn’t be so mean to expose my poor software-virgin humanities friends to mind-twisting UML. I used OPEN/Metis notation, as I have been doing for almost a decade now.
The outcomes of all this is that, after a few weeks, my team mates are actually class modelling now. Amazing. The only unorthodox bit is that they call whole/part diamonds “Juanolas” for evident reasons.