In general the rejected papers are never discussed by conferences, as the authors will want to keep their from being public until they can publish it.

I don't entirely agree with this, and some subjects are moving away from this way of working, but it is still very common in computing.