[erlang-questions] gen_tcp nonblocking send

[Claes Wikström]

Per Hedeland wrote:

> I think this fits a common Erlang solution: "use another process".
> I.e. have a separate "sender" process that just does the potentially
> blocking gen_tcp:send(), ack'ing back to the "real" sender when it

And furthermore if the "intelligent" sender creates the binary() to send
there will be no copying of messages - unless the messages are "small"


/klacke