Take this thinking to the end and realize that it leads to freestanding functions. In general all the context of the program is needed to execute a functionality. It's not like the registration office owns all the students. It's not like a registration wouldn't change the student's context. Students are both an independent and related concept. The proper object to call most things on is a "Global" object. Now instead of
Global.do_some_thing(foo, bar)
just
do_some_thing(foo, bar)
There you have it. OO is an unsound approach which survived so long mainly due to the perceived real world "analogy" and because the Object-Verb-Predicate syntax simplifies code completion.
I like the simplicity of this. If the app is of appreciable size, do_some_thing depends on databases, webservers, external processes, filesystems and configuration. How do you test/debug/explore its functionality without setting all of that up?
I would actually say these all make good "objects", i.e. isolated pieces. On the other hand, no runtime polymorphism is needed. To avoid OOP at a syntactic level, what do you think of the following?
- For tests at a smaller level, decompose the application such that most parts are easily testable in isolation (without external "hard" dependencies).
- For mid-sized tests with external dependencies but mostly unidirectional dataflow, setup a global virtual table with all the mock methods (and instances) that are needed. Alternatively, traditional linking methods.
- For larger "integration" tests there is no substitute to testing the real thing. At some scale and level of interactivity with the database, you just have to talk to the true filesystem, the true database etc. You can still setup a test instance for most cases where there is no interaction with external services.
I believe OO is good for exactly two things: abstract data containers and state machines.
In the former, access hiding cleans up the API and prevents unsafe usages of the container. In the latter, OO enforces a protocol to keep the state machine sealed off and only aware of key inputs and outputs.
And that's it. I've found nothing else. Data itself is better off when strategized to fit in a database, whether off-the-shelf or a custom-tuned, in-memory design. The state machines may need to query a part or all of the database, as well, so their ability to restrict scope only goes so far.
