On 3/1/22 2:19 PM, Bob Ham wrote: > On 01/03/2022 13:07, Christian Grothoff wrote: >> a Master's thesis > >> Florian's Bachelor's thesis > > I cannot believe you would seriously suggest that academic theses > constitute software documentation.
I don't. But: if someone wants to get a first grip on what these subsystems do, they can be a starting point on the theory. Like, when someone wants to understand a TLS library, getting some background on crypto first is a good idea. It's not enough, but at the same time some of the problem I believe we have is that we also need to explain the fundamentals. For example, I had repeated conversations with DNS experts about GNS. They really understand DNS, and TLS. But getting them to understand the cryptography behind GNS, even though some are quite interested, has been surprisingly hard. And I think that is largely because it is non-standard. So if you want to understand how secretsharing, e-voting or SMC auctions work, I think the answer is that no amount of handbook editing will do the trick unless the reader is willing to get some of the academic background first. Oh, and to stress this point: those chapters in the handbook are for _developers_. None of this is something users are expected to understand. But you cannot effectively work on those subsystems without some academic background that IMO really doesn't belong into a handbook either. At least I'd not expect group theory to be covered in a TLS library handbook.
