1

In (n,n) Shamir secret sharing if n shareholders do not have the public values (X values) can they still obtain the secret with only Y values?

user153465
  • 1,583
  • 12
  • 23
  • There is no clear answer in the link you provided. And there is no proof if it is feasible or not. – user153465 May 27 '14 at 12:11
  • @user153465. I think DrLecter's point is that the question is a duplicate. Whether or not the duplicate question has a correct answer is irrelevant to the discussion here. There are other ways to try to get better answers on a question than to post a duplicate question. For example, you can (once you have more rep points here) offer a bounty. You can comment on the answers to explain why they are lacking. You can post a better answer yourself. – mikeazo May 27 '14 at 12:31
  • @user153465 In other words, you have to remember, no one gets paid to write answers on here. We do it of our own good will. I personally think the answers on the linked question are pretty good (I am biased though since I wrote one). If they aren't good enough for you, though, you can, at the very least, explain exactly why. If all you are looking for is for someone to formally prove it for you, I doubt very many people here will be willing to take the time to do that. Your profile says you are a PhD student. I gave some hints on how to prove it in the comments. That should help a ton. – mikeazo May 27 '14 at 13:00
  • Thanks for your avices. I need a bit more clue that what have been provided there. Moreover, Some people from the other forum said it is feasible. I am expecting more comments if you allow. – user153465 May 27 '14 at 13:05
  • @user153465 Both the answers on the linked question also mention that is is feasible to still have security when keeping the $x$ values private. There are some caveats, however, that I mentioned in my answer. Perhaps the other question is not specific enough? For example, the OP mentioned in a comment that he wanted some multiparty computation features to be preserved. That sort of stuff should have been present in the question (because it dramatically changes the answer). Our answers are a function of the input we receive on here. Overly vague questions will not get the best answers possible. – mikeazo May 27 '14 at 13:10
  • Cool many thanks. Multiparty computation on Y seems interesting too !!! But if the Y holders perform computation on Ys it should not affect the security of the system, if the intgrity of Y is not of interest. Am I right? – user153465 May 27 '14 at 13:23
  • @user153465 you are right. I was talking about the proposed solution I had in the other answer. That wouldn't apply if you want to do MPC. – mikeazo May 27 '14 at 14:06
  • As an expert in this field, do you have any scenario in mind that the Ys holder (even if its only one party) can do anything to learn the value of constant term ( or X values)? Thanks – user153465 May 27 '14 at 14:10
  • Yeah. I wrote about it in my answer on the other question. – mikeazo May 27 '14 at 15:53

0 Answers0