Ask HN: Test-Driven Physics?
1 behnamoh 1 6/4/2025, 12:41:22 PM
In software engineering we often have a suite of tests and we make sure any changes in the code base still passes all (most) of these tests.
Are these such tests for physics? For example, someone comes up with a new theory (Relativity++) and wants to make sure this is consistent with the rest of our knowledge about physics. Should they make this verification manually, or is there a set of tests we put this new theory into to see if it passes them?
Example: Let's say Relativity++ is formulated in a certain way. Then we would probably want to make sure it doesn't violate the conservation of energy, objects moving ≤ c, etc.
Many theories are also not designed to be consistent across all known constraints and often are approximations for a particular domain. Where that domain begins and ends is also a matter of interpretation, sometimes not quantifiable in rigorous ways.
Doing science is not like writing code and requires human interpretation.