Noncovariant gauges at zero and nonzero temperature

I review the formalism for gauge-field perturbation theory in noncovariant gauges, particularly the temporal axial gauge. I show that, even at zero temperature, there are complications and it is not known whether a formalism exists for handling these that is correct for all calculations. For thermal field theory in the imaginary time formalism, there are different difficulties, whose solution so far is known only up to lowest order in the coupling $g$.