Sequential IDs have pros and cons. UUIDs have pros and cons too. Why not use both for each table? (Let's say I have -and will have- no problem about storage space)
I could use:
- UUID for exposition, display, export, API, etc.
- Sequential ID for join, internal queries, convenience and debugging.
Is it a bad design idea? What are the pitfalls?
The benefit of using both would be to get the "pros" of each. There is an infinite discussion about "use seq id VS use uuid" and both have good arguments, so maybe I can use both and take benefits of each (depending on use case).