Some of my colleagues work on CFT's and quantum groups and I hear them talk a lot about $\widehat{su(2)}_k$ algebras. According to them (and the general physics literature) these are what mathematicians would call affine Kac-Moody algebras.
However the defining relations of the $\widehat{su(2)}$ level $k$ algebras are rather those of what a mathematician would call the affine $su(2)$ algebras. To obtain a Kac-Moody algebra one needs to introduce an extra derivation.[$\ast$]
Is there a reason for this (apparent) problem with the nomenclature?
[$\ast$] Also see my other post https://math.stackexchange.com/q/3322944/