Comments: Unfortunately, the main result of the paper should be considered as not proven. Namely, the construction D applied to a code $C_0$ with many light vectors gives a lattice $L$ but there is no proof that the kissing number of $L$ is exponential. The initial impression that this lattice retains all the short vectors (light vectors considered as 0-1 vectors) from $C_0$ is definitely false