I remember the gist of the critique being that the K&R book was perfect for the assumptions of its era, but can mislead people who aren't experienced enough to apply its lessons to the modern networked world