The observed distribution of IMF shapes can be understood as statistical sampling from a universal IMF and variations that
result from stellar-dynamical processes. However, young star clusters appear to have an IMF biased towards low-mass
stars when compared to the Galactic disk IMF, which comprises an average populated by star-formation episodes occurring
Gyrs ago. In addition to this tentative but exciting deduction, this text outlines some of the stumbling blocks hindering the
production of rigorous IMF determinations, and discusses the most robust evidence for structure in it. Also, it is stressed that
an invariant IMF should not be expected.