My senior systems development manager created a pull request for the API. The PR has some database changes such as new columns, as I was testing it locally, I found out that those new columns don't have any migration file (which we're always doing when doing database changes).

So I asked why we don't have any migration for those new columns.

Then he answered that I should run some SQL script to add the columns and he doesn't have any migration and is proud to say he's using the central DB (test and live).

I also checked the Live database and was so surprised to see those columns...

Then again, I asked why we already have the columns in the Live database where in the PR isn't merged yet.

Then he answered again, I should think ahead. there are many ways to add the columns.

Like WTF??? Don't we need any migration?

I asked, if we use a fresh DB, what happens if there is no record of those columns in the migration files?