Author Topic: Looking for Truthcoin work  (Read 3778 times)

kvnn

  • Newbie
  • *
  • Posts: 4
    • View Profile
Looking for Truthcoin work
« on: April 17, 2015, 01:42:20 pm »
Hey everyone. I've just found Truthcoin this week, and have read everything possible since then. I want to help realize the system as it is specified in the white paper. I'm an experienced full-stack web developer and would love to work with a company doing this. If you're out there and could use a developer, please reach out.

Furthermore, if there is anyone who needs help but doesn't have funds to hire, I'm still extremely interested in pitching in. Lets make this happen.

Cheers,

- Kevin
- http://kevinriggen.com/
- https://dl.dropboxusercontent.com/u/1283216/KevinRiggen-resume.pdf

zack

  • Global Moderator
  • Sr. Member
  • *****
  • Posts: 384
    • View Profile
Re: Looking for Truthcoin work
« Reply #1 on: April 17, 2015, 03:26:14 pm »
Hello and welcome!
4 blockchains most likely to host Truthcoin: Ethereum, Tendermint, Flyingfox, and NXT.

The team putting it onto ethereum is Augur http://augur.net/ (I used to work for them, and built their prototype)
Tendermint is a proof-of-stake fork of ethereum. I am currently living with the Tendermint team. Their plan is to use Augur. It is written in golang. Tendermint will be secure against up to 1/3rd of the validators attacking.
NXT is a blockchain that already supports trading to some extent, but doesn't have all the Truthcoin features yet. It is in Java. NXT uses a proof of stake algorithm that I don't know very well. I am not sure what percentage of the validators can be attacking. I don't know much about them.

Flyingfox is my project. The earlier versions were in python https://github.com/zack-bitcoin/augur-core but now I am redoing it in elixir: https://github.com/zack-bitcoin/commerce/tree/master/flyingfox Flying Fox proof of stake is secure against up to 1/2 of the validators attacking. Flying fox is not turing complete.
Flying fox is a little different from truthcoin. Instead of LMSR market maker with many participants, it uses tons of CFDs (contract for difference) to make bets between pairs of people. This way is more anonymous and scalable, because the bets don't ever have to be published onto the blockchain http://lightning.network/
Instead of Sztorc's consensus to determine outcomes of events, I will probably use a multi-sig. This is because of p+epsilon attacks. https://blog.ethereum.org/2015/01/28/p-epsilon-attack/

It is possible to be hired by augur or ethereum or tendermint, but I cannot afford to hire anyone for flying fox.
If you contributed ideas or code to flying fox then I will give you some of the coins when it launches.

kvnn

  • Newbie
  • *
  • Posts: 4
    • View Profile
Re: Looking for Truthcoin work
« Reply #2 on: April 17, 2015, 09:32:28 pm »
Thank you, Zack. I hadn't heard of Tendermint - yet another white paper to read this is exhausting but i can't staaaahp

Quote
Instead of Sztorc's consensus to determine outcomes of events, I will probably use a multi-sig. This is because of p+epsilon attacks. https://blog.ethereum.org/2015/01/28/p-epsilon-attack/

So in Truthcoin, when Decisions are mature, "Voters encrypt, sign and broadcast a Ballot which contains their Votes and a new public key". Will you exchange this mechanism for something like "Voters add their signature to a transaction that includes the Outcome they wish to vote for"? Or is the process of voting much different than in Truthcoin?

Do you know if Sztorc responded to the attack presented in that post (I looked but came up empty)?

Why did you move from Python to Elixir?

Thanks again!

zack

  • Global Moderator
  • Sr. Member
  • *****
  • Posts: 384
    • View Profile
Re: Looking for Truthcoin work
« Reply #3 on: April 18, 2015, 12:10:26 am »
Sztorc and Vitalik argued long about P + epsilon. It is one of the longest threads. Sztorc probably still doesn't think this attack matters. The Augur team doesn't think the attack matters.
I think the only way to avoid it is if we do SMPC https://blog.ethereum.org/2014/12/26/secret-sharing-daos-crypto-2-0/
I programmed the multiplication step here: https://github.com/zack-bitcoin/secure-multi-party-computation

So in Truthcoin, when Decisions are mature, "Voters encrypt, sign and broadcast a Ballot which contains their Votes and a new public key". Will you exchange this mechanism for something like "Voters add their signature to a transaction that includes the Outcome they wish to vote for"? Or is the process of voting much different than in Truthcoin?

SMPC is a long-term goal, flying fox wont have it any time soon. Flyingfox uses the trusted feed idea Vitalik described once. decisions are created by oracles, which are groups of addresses. When the decisions expires, the participants in the oracle all sign a transaction saying the outcome. At least N of the M participants have to sign for it to be valid.
It is pretty bad, but I think it is a little better than Sztorc's consensus.

I switched from python to elixir because it was so difficult to write python code that wouldn't crash. Especially the networking aspects were very difficult. Elixir has a set of libraries/principles called OTP. My program is split into 10+ threads, only 1 of which has the blockchain. If any of the threads besides the blockchain crashes or hangs, then it is automatically re-started. OTP is really nice for blockchains. The code ends up being shorter and cheaper to have audited. Out of the 1750 lines of code in flying fox, less than 50 of them are mission critical lines capable of crashing the program. The other 1700 lines could have bugs, it doesn't matter. I never have to catch errors or check types.

psztorc

  • Administrator
  • Sr. Member
  • *****
  • Posts: 464
    • View Profile
Re: Looking for Truthcoin work
« Reply #4 on: April 18, 2015, 03:30:21 pm »
The P + e thread is worth reading, as it is basically a microcosm of "The Truthcoin Forum Experience" so far.

Which is:
[1] zack proposes an attack, claims that it has ruined truthcoin and drastic changes are needed immediately
[2] I point out that the attack has been fully anticipated by me back in 2013 when I drafted whitepaper 1.0
[3] lots of repeating back and forth for a while, tables are usually constructed
[4] vbuterin shows up and either falls back on some old guard game theory (items including non-ordinal utility, failure to emphasize trembling hands, etc), or makes up some brand new crypto-gametheory, crypto-econ, or crypto-statistics.
[5] vitalik and I go back and forth on the differences between our theoretical models
[6] vitalik writes a blog post on https://blog.ethereum.org/
[7] people from the Augur team misunderstand MSRs (which are pretty confusing, sure)

Feel free to email me directly, there will probably be a need for some kind of truthcoin version of blockchain.info , and I know someone who is already building one.
« Last Edit: April 18, 2015, 03:34:09 pm by psztorc »
Nullius In Verba