@User: The concern is probably only relevant internally to the static timing analysis tool; if vesta is calculating the fastest paths to check for hold timing violations, having apparently non-causal output can produce very weird results. It could be that I'm just handling it the wrong way. Pretty much everything in the vesta tool is just what I pulled off the top of my head from basic principles, and a bit of experience writing viterbi decoders and hidden markov models. I am not actually an expert in static timing analysis. I don't even play one on TV.