What are some situations in which melt from restructuring over ddply from plyr might be preferred? I try to study them and they look alike

It seems that the melt will change your data frame using id columns and fit the measured variables, after which the listing will allow you to perform aggregation. ddply, from the plyr package it seems very similar ... you give it a data frame, a couple of variable columns to group and an aggregation function, and you return the data frame ... the way they are different and there are any good resources / links for sharing these tools besides their documentation (which, especially for the change, is a little difficult to accomplish)

thank