It's true traditional databases require you to think through your solution before you go implementing it, but this is not necessarily a bad thing.
NoSQL is great for prototyping and being productive fast, but that productivity comes with a lot of pitfalls. Pitfalls by the way that many devs have been blissfully ignorant of for decades because they have been insulated from them by normalized relational databases.
Unless you are experienced with/aware of those pitfalls, I would almost make a blanket statement that a traditional SQL database should be the default solution for the backend until you can explain why NoSQL is a better fit for a given project.
Also, what makes you think PostgreSQL is slow? What is "slow" to you?
Thank you for the explanation. And I meant just the opposite with postgresql. I use it as a sort of data warehouse with a lot of work in a really modest VPS and it works perfectly.
NoSQL is great for prototyping and being productive fast, but that productivity comes with a lot of pitfalls. Pitfalls by the way that many devs have been blissfully ignorant of for decades because they have been insulated from them by normalized relational databases.
Unless you are experienced with/aware of those pitfalls, I would almost make a blanket statement that a traditional SQL database should be the default solution for the backend until you can explain why NoSQL is a better fit for a given project.
Also, what makes you think PostgreSQL is slow? What is "slow" to you?