[namelosw]

As we usually say technologies does not die easily. But they become legacy when no one want to use them for new project. The main problem of Meteor is scaling problem. Due to it's fundamental design and platform it's relying on (Mongo and node.js). It's very hard to solve the problem for current Meteor. But anyway on the other hand Meteor has extremely high velocity, higher than any framework I have seen (Of course even higher then Rails) due to it's isomorphic nature. Nowadays I'm betting on Elixir and Phoenix to achieve both high velocity and performance. The performance is literally crazy since it runs on Erlang/OTP. But frankly MVC + RDBMS is not something comparable to Meteor on velocity. Currently I'm trying to throw databases away and use WebSocket + GraphQL + GenServer(some in memory process storing data) to boost up velocity, only use data base as a process backup (message + process IS transaction. Then dirty write all the way down). Now there's still some boilerplate, but I found it's promising.

[VikR]

The concerns regarding scaling with Meteor are outdated. Last year that was solved. Please see my comment in this thread.

https://news.ycombinator.com/item?id=16787682