I am trying to provide an AT/Modem-like interface around some hardware. Follwing this post I have the server setting up a pty using openpty().
Now I can communicate with the server as expected with a client app that open the slave and communicates via read() and write() calls.
However I would also like to be able to use either the screen command or minicom to issue commands by hand to the slave. However the server never seems to receive any data when trying to do this. Is there something significant I am missing with this approach?
Can read(2) return zero when not at EOF?