@tolitius yea, this is my assumption. but print debugging in the :start and :stop functions shows the the things are happening in the expected order. should be any chance that states state is shared/cached between stop and start?

btw @tolitius your remark on the anonymous functions should make no difference right? because :stop can take a function or a function definition? is it always evaluated in run-time (reading fun_with_values.cljc I understand that it is)?

@ido: > if a test calls (process) that leaves a message in the outgoing channel, the next test fails on this case, what is the scope of the "outgoing channel"? i.e. if you see "things are happening in the expected order", don't you see this channel being closed between :each test?