An important aspect of identifiers is to not leak any information in the identifier. In some scenarios a prefix might be fine but less important things have been blocked by our dpo department.
Yeah and Twitter IDs used to be 32-bit integers. In this case it is kind of obvious that the IDs leaks information about internal data types. Which is not ok at many companies.
Yeah, this idea makes me uneasy as well for the same reason. The natural conclusion would be to have DB frameworks automatically do this based upon code type names and that definitely feels like a leak, handing the world at large a roadmap to your internal system architecture.
