305 karma · joined January 27, 2023
I would assume to use data they would have to filter it a lot and correlate between many users.
You can detect if the user is the real one and trust their other chats "a bit more".
I feel this is so misleading. For example, by default after spawning, Erlang would have some memory preallocated for each process, so they don't need to ask the operation system for new allocations (and if you want to shrink it, you call hibernate).
Do something more real, like message passing with one million processes or websockets. Or 1M tcp connections. Because, the moment you send messages, here is when the magic happens (and memory would grow, the delay when each message is processed would be different in different languages).
Oh, and btw, if you want to do THAT in erlang, use timer:apply_after(Time, Module, Function, Arguments). Which would not spawn an erlang process, just would put the task to the timer scheduling table.
And Elixir was in the old article, and they implemented it all wrong. Sad.
So, thanks for the charity, but I would rather prefer them to pay that as taxes.
But the main issue is that just googling stuff and checking the results is a pretty miserable experience, because of quality of search results (they would give you CEO optimized trash 90%).