Tendermint consensus protocol

[jaekwon]

Tendermint is a secure cryptocurrency consensus protocol that requires no proof-of-work mining.

It is an adaptation of an existing Byzantine consensus algorithm from academic literature (specifically the DLS protocol http://groups.csail.mit.edu/tds/papers/Lynch/jacm88.pdf).
It solves the "nothing at stake" short range forking problem by requiring validators to post a bond deposit.

* Whitepaper http://tendermint.com/docs/tendermint.pdf
* Forum http://forum.tendermint.com/
* Github http://github.com/tendermint/tendermint/

For more information on the motivations behind its design, check out my first blog post, http://tendermint.com/posts/security-of-cryptocurrency-protocols/.

There is no official testnet yet, but I'll set one up soon.
In the meantime please follow this thread or subscribe to the newsletter on the website for new announcements.

This is not a presale or launch of a currency.  This is a call for developer and general interest.

[var53]

Is it like a PoS coin that can't be forked?

[jaekwon]

Is it like a PoS coin that can't be forked?

Yes.

It's much like a PoS coin but with a lot of coins actually at stake in the case of a "short range" fork.

There is nothing that stops a "long range" fork.  You could let transactions include the last block-hash and measure the economic activity to determine the honest chain, but that's just a heuristic.  The solution is to simply stay in sync with the network.  You are invulnerable to forks as long as you sync with the network periodically.  If the unbonding period is (say) longer than a year, then you just need to sync every year.

[xxxgoodgirls]

cool concept

[Mercator]

Great idea - this POV (proof of Validation) appears to be a real development and something new beyond POW and POS.

I've been looking for an algo which can replace the rewards to a small coterie of miners with a reward to every participant who runs a full node and thus supports the entire system.

Once you have a large network of participators, each getting a regular if small income, then you have the potential for a powerful global crypto which challenges all others.

Can POV do this?

In order to use POV, the participants need coins in order to bond them and participate - how can you do a fair distro of coins in the first place? - has this been considered in this context?

[thelonecrouton]

Sounds uncannily similar to Darkcoin's Proof of Service.

Swap 'validator' for 'Masternode'...

[jaekwon]

Once you have a large network of participators, each getting a regular if small income, then you have the potential for a powerful global crypto which challenges all others.

Can POV do this?

Distribution and consensus are separate problems that mining lumps into one.  With Tenderment (or PoV, I like that) the distribution is left to the implementer while it solves the consensus aspect.  You could tack on an initial mining distribution phase that fades out quickly, or require a proof of burn with SPV built into the protocol as a one way peg, or handle it manually.

I think in general you want to curate the initial seed set of validators.  Otherwise you get the kind of problem that Stellar had where somebody figures out a way to game the system with mechanical Turk or other.  Get a list of people who want to participate in bootstrapping, let them prove somewhat that they're real unique people by ideally posting their pseudonymous online identity (eg bitcointalk identities with minimum number of posts), challenge them to run a node reliably with good uptime, and then they can be part of the bootstrapping team.

[Shinzi Fuki]

So what type of algo would you call a coin that uses this? TCP? like POS? Has this been deployed into any coin yet?

[jaekwon]

So what type of algo would you call a coin that uses this? TCP? like POS? Has this been deployed into any coin yet?

Let's call it PoV (proof of validation), which is a type of PoS, for now anyways.  There are aspects of the protocol that work together as a gestalt, and labeling it as proof-of-X is an oversimplification.

In terms of deployment, there will be.  But lets go to http://forum.tendermint.com/t/the-apps-category/20 to discuss ways to move forward before deciding on a particular application.  The consensus logic is already implemented so it's just a matter of deciding which application is more exciting.

[delulo]

Tendermint is a secure cryptocurrency consensus protocol that requires no proof-of-work mining.

It is an adaptation of an existing Byzantine consensus algorithm from academic literature (specifically the DLS protocol http://groups.csail.mit.edu/tds/papers/Lynch/jacm88.pdf).
It solves the "nothing at stake" short range forking problem by requiring validators to post a bond deposit.

* Whitepaper http://tendermint.com/docs/tendermint.pdf
* Forum http://forum.tendermint.com/
* Github http://github.com/tendermint/tendermint/

For more information on the motivations behind its design, check out my first blog post, http://tendermint.com/posts/security-of-cryptocurrency-protocols/.

There is no official testnet yet, but I'll set one up soon.
In the meantime please follow this thread or subscribe to the newsletter on the website for new announcements.

This is not a presale or launch of a currency.  This is a call for developer and general interest.

Do you have a more or less thought out road map how to move this project forward and make the idea into a product people can use?

[jaekwon]

Do you have a more or less thought out road map how to move this project forward and make the idea into a product people can use?

Hi delulo,

There are many options possible at this point, so my plan now is to create a consensus as to what to do next.  I can't do it on my own.  So I think it's best to get community involvement at the early stages before making any further design decisions.  The immediate purpose of publishing the (draft) whitepaper and prototype implementation was to prove competency and find birds of a feather.  I hope I did the former and it looks like the latter is happening.

Come join the forum!

http://forum.tendermint.com/t/the-apps-category/20

[fartbags]

Is this a thing that you can buy?

[jaekwon]

Not yet, uh, fartbags.  But there will be a testnet up soon.

[tobeaj2mer01]

Any update?

[coinking]

there's a reddit page if you want to stay up to date

https://www.reddit.com/r/tendermint

[TPTB_need_war]

Tendermint can't work. Sorry.

I don't have time to respond to BS shills who can't comprehend the technological issues. Those who want to understand can dig into the following.

8. Conclusion

Tendermint is awesome. The future is now.

Incorrect! Tendermint is flawed and is a divergent non-consensus design that will crash and burn in the real world.

There are a lot of things that can work in the real world but not in the ideal one. For example, Ripple.

Following is written by David Mazières a PhD professor at Stanford who is the Chief Scientist at Stellar:

Still another approach to consensus is Byzantine agreement [Pease et al. 1980; Lam-
port et al. 1982], the best known variant of which is PBFT [Castro and Liskov 1999].
Byzantine agreement ensures consensus despite arbitrary (including non-rational) be-
havior on the part of some fraction of participants. This approach has two appealing
properties. First, consensus can be fast and efficient. Second, trust is entirely decou-
pled from resource ownership, which makes it possible for a small non-profit to help
keep more powerful organizations, such as banks or CAs, honest. Complicating mat-
ters, however, all parties must agree on the the exact list of participants. Moreover,
attackers must be prevented from joining multiple times and exceeding the system’s
failure tolerance, a so-called Sybil attack [Douceur 2002]. BFT-CUP [Alchieri et al.
2008] accommodates unknown participants, but still presupposes a Sybil-proof cen-
tralized admission-control mechanism.
Generally, membership in Byzantine agreement systems is set by a central authority
or closed negotiation. Prior attempts to decentralize admission have given up some of
the benefits. One approach, taken by Ripple, is to publish a “starter” membership list
that participants can edit for themselves, hoping people’s edits are either inconsequen-
tial or reproduced by an overwhelming fraction of participants. Unfortunately, because
divergent lists invalidate safety guarantees [Schwartz et al. 2014], users are reluctant
to edit the list in practice and a great deal of power ends up concentrated in the main-
tainer of the starter list[i.e. centralized!]. Another approach, taken by Tendermint [Kwon 2014], is to
base membership on proof of stake. However, doing so once again ties trust to resource
ownership.

Again I maintain the Iota/DAG will only remain convergent with a centralization to enforce the trust math model. Btw, I will soon argue the same about Stellar as well.  

[CropCircleSystems]

as of today seems reasonably active. does not provide quite the same security guarantees as proof of work or the exact same use cases as bitcoin but very useful and extensible