On 8/16/19 2:21 PM, Ralf Jung wrote: > and I think we could get a decent generalization of auth > without a lot of SProp infrastructure. In that case I presume you would keep the old auth, and make the new version just for the discrete case?