> it's easy to throw a frontend away and replace it with a new one - try to do that with a backend - usually much much harder.
Untrue. I work at a company that's swapped out the frontend and then later the backend. I had to work on both of the projects because, and I can tell you without a any doubt in my mind that they were both equally challenging (for different reasons, but that's beyond the point).
In my experience frontend-only (or frontend-focues) devs tend to underestimate backend work, and the the backend devs are equally guilty of doing the same about frontend work. Everyone thinks they've got it the harderst. ¯\_(ツ)_/¯
I'm sure that this also happens sometimes and that there are products where the frontend is the main part and requires more work than the backend.
But I still believe that it's usually easier and more valuable to recreate the frontend, and the reason is that Backend is (on average) more stable, grows over a longer time. The reason is also that the risk of replacing the backend is higher, since for a frontend, you can usually use two of them in parallel (think AB testing) without much problems. But if you do the same with the backend and the new backend writes incorrect data into the db, then the old backend also starts to behave incorrect. Hence harder to change.
Untrue. I work at a company that's swapped out the frontend and then later the backend. I had to work on both of the projects because, and I can tell you without a any doubt in my mind that they were both equally challenging (for different reasons, but that's beyond the point).
In my experience frontend-only (or frontend-focues) devs tend to underestimate backend work, and the the backend devs are equally guilty of doing the same about frontend work. Everyone thinks they've got it the harderst. ¯\_(ツ)_/¯