Trying to write a current account about exoplanets reminds one of the definition of Post-Modernism: “it's so new that it's out of date already”. Looking for exoplanets around stars used to be an unpopular topic, shunned by professional astronomers; but times change. One is reminded of the comment by Talleyrand that “treason is a matter of dates”. Now a barely concealed excitement suffuses much of the serious scientific literature. Among many quotations the following give a flavor: “the search for extra-terrestrial planets – rocky worlds in orbit around stars other than the Sun – is one of humanity's most exciting science goals” [2], while the real goal seems to be “the detection and characterization of habitable worlds” [3].

The previous section showed what we have attempted to learn about the formation of planets, based strongly on observations of our own system. The discovery of the exoplanets has revealed many surprises and demonstrates once again the triumph of observations over theory. Although “one of the most surprising aspects of the hundreds of known exoplanets is their broad diversity” [2], this was less surprising to students of our own solar system, in which all the planets and satellites display an astonishing range of sizes and compositions. This results in the notorious difficulties of trying to classify or even define planets, exacerbated now that we have such a variety of exoplanets. The long established upper limit for planets of 12 or 13 Jupiter-masses, that marks the lower limit of deuterium burning, has now been extended to at least 25 Jupiter-masses into the brown dwarf desert, and will need to go to higher masses. Nevertheless and consistent with its name, the desert remains thinly populated. Below the 12 Jupiter-mass limit, problems also occur because of the discovery of free-floating objects down to at least 3 Jupiter-masses. Are these bodies that are not linked to any star, “free-floating planets” ejected from the nest, or “sub-brown dwarfs”? We seem to be approaching a classification for planets, brown dwarfs and the like based on where they formed. Truly a nightmare for classifiers.