Given "BSD systems already prohibit an empty argv" it certainly seems unlikely that it's a standards related issue in practice. So (as they appear to be doing

Given "BSD systems already prohibit an empty argv" it certainly seems unlikely that it's a standards related issue in practice. So (as they appear to be doing if I read correctly) keeping the patch in the tree for the moment, seeing what if anything actually breaks in practice, and then trying to figure out what the best answer is from there, is probably the only thing to do. It's always a horrible day for a developer when they discover they have a bugwards compatibility problem and while it's, I guess, debatable if that's exactly what we're seeing here, it's close enough to the pain I know well that I have much sympathy for everybody involved here.