Floating the idea that in the next budget we should begin allocating a bounty for Breaking Darksend?
Budget Proposal - Breaking DS
- Thread starter Sapereaude
- Start date
AnarchicCluster
Well-known member
I like this idea. I'd get way more motivated to keep translating Hehe.
Regarding DS breaking , I think so too that this is a great idea and we discuss that already, we need a confirmation from Evan and udjin regarding the code (ready or not) and how to approach that
Regarding more translations ... I think the problem is not translating , but maintaining ! As this is an extensive effort I am still vouching for staying in the top 5 languages and having them 100% !
Regarding more translations ... I think the problem is not translating , but maintaining ! As this is an extensive effort I am still vouching for staying in the top 5 languages and having them 100% !
AnarchicCluster
Well-known member
So there are two ways of doing it. Announcing a bounty for breaking DS. If somebody succeed a proposal will get introduced and masternodes will vote on the payout of the bounty.
Another way I can see, is something like this: Masternodes vote now on the payout of the bounty. Bounty is paid out to a prominent member of this community (maybe Evan) who awards the funds in the case somebody successfully break DS. If it doesn't happen within, let's say a year, the contest will get shut down and the bounty will be allocated towards any other good cause e.g. marketing.
Before posting any proposal I would start from defining what "Breaking Darksend" is exactly and realising that we are not living in a perfect world 
1) Following back from specific output?
In some rare edge cases some really unlucky guy with small amounts of funds to mix still could have multiple mixing rounds leaking information. The more rounds the less probability but it still exists until we release a fix https://github.com/dashpay/dash/pull/564 (or smth like that). (note: it says "merged" but was reverted later and most likely will require to be fixed/reimplemented)
2) Breaking during mixing (injecting you inputs to follow back later)?
While there are not so many people mixing and almost no liquidity yet it's not that hard to modify client (or build a tool) to provide multiple 2 inputs for every mixing session so that you'll be 100% sure that 3rd input is another person. Do this 24/7 and that's it - you have all sessions from round X (X - where he started mixing, could be 0 if you are lucky) to whatever number of rounds that person was mixing. That is possible because of the very nature of CoinJoin (and DS inherits this) - it provides you an ability to hide among many others. But there must be "others" AND there must be many of them for you to be able to hide among them. When one of these conditions is false it's highly questionable if the whole method works at all.
PS. Bounty receiver address is in my signature :grin:
1) Following back from specific output?
In some rare edge cases some really unlucky guy with small amounts of funds to mix still could have multiple mixing rounds leaking information. The more rounds the less probability but it still exists until we release a fix https://github.com/dashpay/dash/pull/564 (or smth like that). (note: it says "merged" but was reverted later and most likely will require to be fixed/reimplemented)
2) Breaking during mixing (injecting you inputs to follow back later)?
While there are not so many people mixing and almost no liquidity yet it's not that hard to modify client (or build a tool) to provide multiple 2 inputs for every mixing session so that you'll be 100% sure that 3rd input is another person. Do this 24/7 and that's it - you have all sessions from round X (X - where he started mixing, could be 0 if you are lucky) to whatever number of rounds that person was mixing. That is possible because of the very nature of CoinJoin (and DS inherits this) - it provides you an ability to hide among many others. But there must be "others" AND there must be many of them for you to be able to hide among them. When one of these conditions is false it's highly questionable if the whole method works at all.
PS. Bounty receiver address is in my signature :grin: