4 ms·

I understand, however my whole point is that these different other qualities change the shape of the problem (and necessary solutions or impossibilities) comple

I understand, however my whole point is that these different other qualities change the shape of the problem (and necessary solutions or impossibilities) completely, often in ways that make it unrealistic for the real-world problems that the paper appeals to in its abstract as motivation. The word "byzantine" is so over-used as well as abused in this field of research, that for a random paper that mentions that word I can't really tell what actual problem it's solving unless I read through it fully in detail. I suggest that a wider more systematic set of terms would help to avoid this problem. e.g. instead of saying "secure" say "unforgeable against a chosen-message adversary" etc. Lamport's original paper does seem to treat adversaries appropriately, but I've seen many worse papers from recently failing to consider what happens if a malicious peer specifically tries to exploit your "defence" mechanism.

peterwwillis 9y ago
> The word "byzantine" is so over-used as well as abused in this field of research, that for a random paper that mentions that word I can't really tell what actual problem it's solving unless I read through it fully in detail I wouldn't say this is a problem. It's a useful red flag that you should stop reading this paper, or at the very least question the competency of the authors. Could save hours of your life!
- infinity0 9y ago
  It is a massive problem because it wastes a lot of people's time doing research in this area, trying to read up on previously-done work most of which is uninsightful and useless. The whole point is that I can't tell if it's a red flag simply by reading the abstract, I have to read the entire paper. The good papers also use the same vague terminology!