I am going to review the status of loop quantum gravity as it is seen from the point of view of a Lorentz covariant approach to loop quantization. I'll start from a brief review of the standard loop approach based on the SU(2) gauge group, its main results and problems. Then I'll present another approach which is based on a canonical formulation of general relativity which is explicitly covariant under the local Lorentz transformations. It allows to overcome several problems of the standard loop quantization and at the same time shows that the latter breaks the diffeomorphism invariance and is not a correct way for quantizing gravity. In the covariant framework I'll derive a new spectrum of the area operator, both for spacelike and timelike surfaces, and show that its predictions agree with the spin foam quantization.