I suppose such a discussion could spiral into semantics so I'll only say this: Yes, in theory we have a lot of choice but in practice you start a project and y

I suppose such a discussion could spiral into semantics so I'll only say this: Yes, in theory we have a lot of choice but in practice you start a project and you need a good ORM / data-mapper, you need a good HTTP/Web/WebSockets framework, you need good DB migration solution, you need SSO for 5 services, and you need 20+ more things and when you start eliminating stuff there are barely any programming languages left. If even 20 are left at the end of such an evaluation that would be still pretty good (but in my practice you usually end up with maximum 7-8 viable choices). You don't have to agree with me on this, or I with you -- it's my empirical observation which is prone to bias and filter bubbles (of course).