What would make a good “polycrypto” project? (I.e., a crypto-related research question that people could successfully work on in a distributed fashion.) I thought about it a bit, but nothing struck me. Some musings, though:
- It would have to be a problem that’s neither too easy nor too hard; that is sufficiently compelling to sufficiently many people; and where even partial progress would be interesting.
- What kinds of problems are best suited to having large numbers of people contribute in relatively small increments on a wiki? I guess it would be problems that require expertise in different areas; that have no clear path to a solution; and where the solution is likely to be comprised of many small steps rather than one key breakthrough. But I’m really guessing here.
- Is crypto deep enough to have problems difficult enough for such a project? Let me explain what I mean. In mathematics, people are generally experts in their own (sub)-sub-area and are less up-to-date on material outside their own expertise; moreover, some of the deepest math problems seem to require techniques from one or more areas. On the other hand (and generalizing a bit), people in crypto are usually familiar with several areas, and there are not many problems that “straddle” two or more areas.
- Maybe it’s just my perception, but I feel like the general attitude in crypto is less receptive to sharing problems/ideas/credit than in math. Thus, people might be unwilling to propose their “favorite problem” for a polycrypto project, and people may be unwilling to contribute to a paper if their name will not appear on it. (For what it’s worth, I would rather have the names of everyone who contributed to the solution appear on the published paper, rather than have the paper published under a pseudonym [as was done for the polymath project].)
We can test the last statement now: please post if you have any ideas about a good candidate problem!