Declare the servers to be spawn in a config file

[kevgo]

Create a file cukelib.yml in features/support with the content:

servers:
  echo:
    command: node foo/index.js --port 3001
    online-text: running at port 3001

Then you can say in Gherkin:

Given a running "echo" server
...

and that step library waits until the server prints running at port 3001 before finishing, so that you know your server is up and running. You can use https://github.com/originate/observable-process for this, it provides functionality for spawning and waiting for output out of the box.

[mjhm]

There was already a parameter .isReady: function(proc) {...}. The function parameter is expected to return a resolved promise when the proc is ready. The default behavior is to wait for any output on "stdout". As you implied this isn't specific enough. I added another parameter .isReadyMatch: <string|RegExp> which indicates the text to watch for in either stdout or stderr.

I'm going to leave this issue open because the yml config file might be a good addition later. But the code is still too new and fluid for now.