These days, everyone wants to add new features to the bitcoin block chain, but now one team thinks that it's found a way to do it responsibly - by creating "side chains" that can interact with bitcoin.

The process, known as 'two-way pegging', has some hefty backing. It is supported by Adam Back, the developer of the HashCash, the algorithm bitcoin uses to generate and verify currency.

On the business side, it has Austin Hill, a Canadian entrepreneur and angel investor who in 1997 founded Zero-Knowledge Systems, an anonymous networking and privacy technology company. Hill, who raised more than $80m for the endeavor, likes to say it was Tor, before Tor was invented.

Clearly, the money and the technical know-how are there. So, how will it work?

[post-quote]

There have been two different approaches to expanding bitcoin's functionality in the past. One has been to introduce new features that piggyback on it, using the bitcoin block chain's existing resources. Counterparty is one example of this, and Mastercoin is another.

The other is to just create whole new frameworks for next-generation cryptocurrencies outside of bitcoin. Examples of these are NXT and the still-unlaunched Ethereum proposal.

The problem with completely separate altcoins and frameworks, says Hill, is that they they are not interoperable. They create new "races for scarcity", with each one representing a new asset in short supply.

But, there are several problems with bitcoin, too, says Hill, which is why exchanges like Coinbase generally run their own transactions outside the block chain.

Hill told CoinDesk: "There is a practical reason why some of those things can't operate on the block chain, and that's because at peak, they're doing more transactions than bitcoin can support right now."

Instead, Back and Hill want to 'shard' the block chain, creating another block chain, which has a two-way relationship with bitcoin's own.

This will have two major benefits, Hill suggests: functionality, and scalability.

And secondly: "By sharding, you can have orders of magnitude faster transaction processing and you can start to scale bitcoin."

In October, Back had proposed a concept called one-way pegging, in which bitcoins could be 'moved' from the bitcoin block chain to another block chain called a side chain, that was merge-mined with bitcoin. Bitcoins would be marked in the bitcoin block chain as having been transferred. Another coin in the new block chain would be marked as a representation of the transferred bitcoin.

Then last year, bitcoin core developer Greg Maxwell figured out a method for two-way pegging, in which bitcoins could be moved back and forth between the side chain and the block chain.

This would allow them to be transferred at will, so that the bitcoin network wouldn't lose its bitcoins forever when they were transferred, but could get them back if the owner decided to transfer them back into the bitcoin block chain.

Bitcoin would be 'firewalled' from the side chain, meaning that any security issues that arose in a side chain wouldn't affect people who were only involved in the bitcoin block chain.

This opens up some interesting possibilities, says Back:

The advantage here is that, unlike altcoins, they would all be interoperable, with the bitcoin block chain (or, as he calls it, the "main chain", used as a common point of transfer.

So, that takes care of functionality. But what about throughput? The current transaction rate on the bitcoin network is relatively low, at something under seven-to-nine transactions per second, which is why all of those off-chain transactions happen, in exchanges and elsewhere.

The downside of that is that they become difficult (but not impossible) to audit in any publicly visible way.

Says Hill: "You are still having to extend and move your coins off the block chain into a trust me security model. That has led to huge amounts of theft."

So, that's the proposal. What do core developers think?

"In particular the notion that a different chain might be designed to natively support things like smart contracts is funny, given that Bitcoin already supports them," argues Mike Hearn, the brains behind BitcoinJ, and a core developer since bitcoin began (he spoke about them as early as 2012, arguing that all the pieces of the puzzle were in place).

The recent introduction of 40 extra bytes into bitcoin transactions for arbitrary messages (such as those associated with smart property or smart contracts) supports that view, although third parties have asked for more space.

Hearn also doesn't think that alternative side chains can solve the scalability problem, which in any case is solvable purely within the bitcoin network, he argues.

"It's not as if there's some impossible limit beyond which Bitcoin can't go," he protests, pointing to the Bitcoin Wiki's scalability page, which breaks down how much further we can push the network.

Gregory Maxwell, who first proposed two-way pegging, says that it could provide some relief for bitcoin, though:

Bitcoin core developer Jeff Garzik argues that off-chain transactions are an enabler, not a problem.

Garzik describes himself as a fan of off-chain transactions and points to his own side project, payment channels, as an example of setting up high-volume transactions with other systems outside the bitcoin network.

So, some experts believe that two-way pegging could provide some useful respite for bitcoin from a scalability perspective, if side chains were able to connect with audit-able, fraud-proof private chains designed for high throughput. Those side chains could also be many and varied, each designed with their own specific properties for tasks like microtransactions and derivatives, say.

But, for two-way pegging to work, changes have to be made to the Bitcoin protocol. Will Back and Hill's venture be accepted by the bitcoin open source community?

Said Garzik:

"One thing open source developers try to avoid is pre-approval based on some future supposition.  We have no idea what the future will bring, any more than anyone else.

Once a concrete proposal appears in a pull request, then we can give a reasonable comment based on engineering evaluation."

If the changes necessary to allow bitcoin to talk to other side chains are made, then it could relieve the bitcoin developer community of requests to tweak the blockchain to suit third-party purposes.

Maxwell calls it the one change to rule them all.

Hill and Back, who claim to have some core developers supporting their venture, plan to announce the full details - and its official name - by mid-May.

Chains image via Shutterstock