>>36
It should know, which players are on the network
umena global consensus on cluster state mechanism built into language. not realistic.
Lisp
i didn't see this buzzword is going to pop out
>>37
it sucks at strings and hashes/records
nope it's not.
it does suck at whole bunch of more important things
unityping. in absence of oop support (module extension, parametric module style calls are banned (for good reasons)) this sometimes present big problems. dialyzer is (still, after ~10 years of development (kostis is a hell of a slacker)) broken beyond practical value. not mentioning that it requires (just) 20 minutes of processors time for a small project.
no metaprogramming support - parse transformations are unusable.
being advertised as functional language its capabilities stop right at ``apply'' function.
puke-inducing syntax makes price of abstraction penalty too high. execution model also plays important role here.
speaking of which, just as yale implementation of haskell failed due to gc discrepancy of lisp and haskell, akka and several haskell endeavours failed to provide actors model in other languages. erlang remains unique in this aspect, sadly.
>>39
Elixir
120% hipster shit
irrespective of your taste in languages you'll be more comfy with ruby.
.
gotta check golang, but is seems to be way too low-level.