I helped build a new team of elixir programmers up to about 8 people in 6 months all in a third tier tech city. Most of that hiring was done before Elixir hit 1.0.
I don't buy the bus factor argument.
In fact, it's quite the contrary. If you want good programmers, "limiting" yourself to something like erlang or elixir (vs a "popular" language like go) works in your favor.
The best engineers know why to choose erlang or elixir over go and will come to you. It works as a filter and makes hiring easier.
Seeking the most popular language is the method of PHBs who want to see programmers as a commodity and want to get the cheapest labor pool from which to work with.
You increase your value by going tup the spectrum into more esoteric and better languages.
That's good and all but you miss a point: Go isn't (still) a popular language (either quoted or not and in the context of Java or C++ being popular languages with vast pools of programmers), it's just slightly more popular than Erlang.
I don't buy the bus factor argument.
In fact, it's quite the contrary. If you want good programmers, "limiting" yourself to something like erlang or elixir (vs a "popular" language like go) works in your favor.
The best engineers know why to choose erlang or elixir over go and will come to you. It works as a filter and makes hiring easier.
Seeking the most popular language is the method of PHBs who want to see programmers as a commodity and want to get the cheapest labor pool from which to work with.
You increase your value by going tup the spectrum into more esoteric and better languages.