Coinbase Director: GDAX Will Not Support Bitcoin Unlimited Code "As Is"

Charlie Lee, the Director of Engineering of Coinbase and creator of Litecoin, will not allow GDAX, the digital currency exchange arm of Coinbase, to support Bitcoin Unlimited code as is.
Charlie Lee, the Director of Engineering of Coinbase and creator of Litecoin, will not allow GDAX, the digital currency exchange arm of Coinbase, to support Bitcoin Unlimited code as is.

Charlie Lee, the Director of Engineering of Coinbase and creator of Litecoin, recently emphasized that he will not allow GDAX, the digital currency exchange arm of Coinbase, to support Bitcoin Unlimited code as is.

In a statement, Lee wrote:

“It’s extremely risky for GDAX or any exchanges to support a BU chain as is. It is unstable because at any time it can be annihilated. There is absolutely no way I would let GDAX make the reckless decision of supporting BU code as is.”

Lee further discussed the impact a hard fork could have on the long-term development and health of the Bitcoin network, especially if it leads to a split chain in which two Bitcoin networks will coexist, similar to the Ethereum-Ethereum Classic case.

Coinbase and GDAX in Support of Segwit

Over the past few months, Lee and Coinbase CEO Brian Armstrong expressed their support for the Bitcoin Core development team’s scalability and transaction malleability solution Segregated Witness, hereinafter Segwit.

As reported by Cointelegraph in January, Armstrong stated:

“From learning more about SegWit I think we should activate it. It's prob the best path forward for Bitcoin at this point.”

At the time, the community favored Segwit over Bitcoin Unlimited primarily due to the technical difference in the updates. Segwit is a soft fork, which is forward compatible. That means blocks considered valid by the newer version of rules are also valid in the old version. However, hard forks are not forward compatible and present higher security risks due to the possibility of the occurrence of a split chain.

In his recent series of statements, Lee dug in deeper into the security issues of a hard fork and the impact these issues could have on the network as a whole. When Bitcoin Unlimited is executed as a hard fork and the network splits into two, hashrate will be distributed between Bitcoin Core and Unlimited.

However, because Bitcoin Core doesn’t accept any blocks that are larger than one megabyte, Bitcoin Unlimited blocks, which as the name suggests, don’t have specific block sizes, will not be compatible with Bitcoin Core.

In other words, while Bitcoin Unlimited will be able to support Bitcoin Core blocks, Bitcoin Core client will not be able to verify blocks that are larger than one megabyte.

“Bitcoin Unlimited (BU) is a compatible hardfork upgrade to Bitcoin. This means that the BU chain will accept an all one mb-blocks chain. Whereas the current Bitcoin Core (BC) chain will not accept any >1 mb blocks in its chain. What this means is that if at any point, the BC chain grows longer (in PoW) than the BU chain, the BU client will reorg to the BC chain. All existing BU transactions will be dropped in favor for whatever is in the BC chain. And GDAX would be on the hook for everything,” explained Lee.

The China factor

According to the Segwit adoption list on Bitcoincore.org, the Bitcoin industry’s largest Bitcoin companies including Coinbase, Blockchain and Bitfury are Segwit ready. Most wallet platforms and Bitcoin service providers are in support for Segwit, despite the growing fee market.

Bitcoin Unlimited is being pushed and supported by miners, especially by miners in China. There seems to be a consensus amongst the Chinese mining community to support Bitcoin Unlimited in order to increase the block size of Bitcoin and welcome more users and transactions.

It is still difficult to speculate which solutions will be implemented. The arguments are that Segwit alters too much of the Bitcoin network’s layer one and that the Bitcoin Unlimited hard fork could lead to dangerous situations which the community could struggle with.