There are a few exchanges I use for trading cryptocurrencies such as Kraken and Coinbase, so I was interested to receive this email from Coinbase last Friday:
Dear Coinbase Customer,
We are contacting you to make you aware of recent developments in a number of proposals for technical changes to Bitcoin. All BTC stored on Coinbase will remain safe during these events described below.
The User Activated Hard Fork (UAHF) is a proposal to increase the Bitcoin block size scheduled to activate on August 1. The UAHF is incompatible with the current Bitcoin ruleset and will create a separate blockchain. Should UAHF activate on August 1, Coinbase will not support the new blockchain or its associated coin.
The User Activated Soft Fork (UASF) is a proposal to adopt Segregated Witness on the Bitcoin blockchain and could result in network instability. It is scheduled to activate at the same time as the UAHF.
To ensure the safety of customers’ funds, we will temporarily suspend BTC deposits, withdrawals, and buy/sell starting approximately 4 hours before activation of either fork.
- If you do not wish to have access to UAHF coins, and do not wish to access your BTC during the fork, you are not required to take any action.
- If you do wish to have access to UAHF coins or access your BTC during the fork, you should send your BTC from Coinbase to your external address by July 31.
For more information on these potential Bitcoin forks, please refer to this article: https://support.coinbase.com/customer/portal/articles/2844217-uahf-uasf-faq.
Some people are congratulating the firm for warning users that if there is a hard fork on August 1, then at least they have advised you should withdraw your funds now. I disagree. The reason I disagree is that Coinbase have a limit of £15,000 per week for withdrawals so, if you have more than £30,000 in Coinbase bitcoins then your additional funds will be lost if there is hard fork. Or rather, they won’t be lost. They will just be taken by Coinbase. It is yet another sign of how flaky this cryptocurrency market is – the DAO hack, the Mt.Gox debacle, the Bitstamp affair, the Bitfinex loss, the Parity Wallet breach … the list is endless. All it proves is that if you are investing or trading in cryptocurrencies you are as likely to lose your funds as keep them, unless you hold them in a secure wallet of your own, such as Trezor.
Meanwhile, what is this fork all about?
This update from 99bitcoins explains all:
The source of the issue is Bitcoin’s scalability – It’s ability to handle a growing number of transactions. The current protocol limits the size of each block to 1MB (once every 10 minutes on average). This, in turn, limits the growth potential for bitcoin’s usage.
There’s a need to upgrade the protocol, but there is a disagreement about the best way of doing so. This disagreement has been a hot topic for over two years, and has gone beyond just a technical discussion, into a two camps schism about politics, governance, philosophy, identity, side picking, propaganda and more.
One of the consequences of the inability to decisively reach a solution is that it that none of the proposed solutions have been implemented and therefore the network’s capacity is still limited. As a result, there are periods where the transaction fees rise dramatically, due to inability to fit all of the transactions in the upcoming block. In other words, people are paying more in transaction fees in order to “cut in line” and get their transaction confirmed quicker.
Many proposals were made, but two leading solutions have been brought forward. The 2MB Hard Fork and the SegWit Soft Fork.
About them forks…
There are two main ways in which you can upgrade the bitcoin protocol, a hard fork (HF) or a soft fork (SF).
Hard forks loosen up the rules of the protocol. Blocks which were invalid by the old protocol become valid in the new one.
Soft forks tighten the rules of the protocol – Blocks which were valid by the old protocol become invalid by the new one.
A hard fork obligates all of the nodes in the network to upgrade in order to be implemented . A node that does not upgrade to the newer version, will run into blocks which are invalid by his version, will reject them and the rest of the chain, refusing to recognize what’s taking place on the network.
A soft fork doesn’t require all of the nodes to upgrade. The main chain, valid by the strict new rules, is also valid by the older rules enforced by a node that did not upgrade. Therefore it will accept all of the transactions in it.
Having said that, nodes that are taking a part in mining do need to upgrade. This is needed so they won’t mine blocks that are invalid by the new stricter rules, which will be rejected by other miners. Nodes that want to explicitly use new features enabled by the soft fork must also upgrade.
Due to the inability to make sure all of the nodes in the network have upgraded, and the damage done to a user that did not upgrade in a hard fork, hard forks are considered by many to be a riskier solution. Hard forks should be used mainly as a last resort and need to be carefully planned. Soft forks, on the other hand, are considered a safer, successfully tested solution.
Holders of this position, including the community of the reference client (Bitcoin Core) are advocating for a solution called SegWit (Segregated Witness). The witnessrefers to the signature of a transaction, and the segregated refers to the possibility to separate it from the block and keep it in a separate database.
This mechanism has a number of advantages:
- It solves a problem called transaction malleability, which allows for the same transaction to appear with different transaction IDs and confuse the system.
- This fix will allow the use of bitcoin for more advanced transaction types, such as a payment channel network called the lightning network, which aims to dramatically increase Bitcoin’s scalability, allowing for instant, cheap and safe transactions.
- It also gives an immediate effective block size increase, and does it through a soft fork. This is why the development community rather start with SegWit as a solution to the current network congestion, and then consider if and how to implement further solutions.
User Activated Soft Fork (UASF)
Bitcoin’s development team, which supports SegWit, has released a new version of the software that enforces the new SegWit rules after 95% of miners show for it as well. When a miner mines a block he can signal his support for SegWit, and when enough miners do so the soft fork becomes valid.
Unfortunately, Not enough miners have signaled for SegWit support, so the protocol does not change. Some of SegWit’s eager supporters decided to take action despite the miners signaling and developed a procedure called UASF, or User Activated Soft Fork..
The most widely known version of UASF is called BIP148. It’s a protocol change saying that beginning on August 1st, blocks that do not signal for SegWit are invalid.
This change was not merged into the bitcoin core reference client code, but only to an alternative version of it for users that explicitly support a UASF. The idea is to force miners to signal for SegWit support. This is due to the fact that miners would want their blocks to be accepted by nodes that enforce UASF. When miners start signaling support for SegWit, all of the nodes will start enforcing it, even if they don’t explicitly support UASF.
Splits in the Bitcoin network
The problem arises when some of the miners go by the new UASF rules and others don’t.
For UASF enforcing nodes, blocks of non-UASF miners are considered invalid.
For nodes that do not enforce UASF, the UASF blocks will look valid but irrelevant, since they are mined on top of the shorter end of the chain (assuming UASF miners are the minority).
This causes different miners and different nodes to have a different view of which blocks are valid, what the blockchainactually looks like and which transactions are included in it.
Different nodes will give different answers regarding the funds in a given address. For an address that contained funds before the UASF day, there is no problem. Both sides of the split will recognize the validity of blocks, their transactions and the bitcoins in the address. However, a transaction that was conducted a day after UASF was in place may be considered as legitimate by one node and not by the other.
This phenomenon causes bitcoin to actually split into two coins – UASF Bitcoin and Non-UASF Bitcoin, or to be more generic we can call them Bitcoin A and Bitcoin B.
Each coin has its own nodes, its own blockchain and its own balance for each address. Each user that had coins before the split could use these coins separately on the Bitcoin A network or on the Bitcoin B network. This will cause each of the coins to have their own exchange rate as well.
So if someone had X bitcoins before the split, he’ll now have X bitcoins A and X Bitcoins B, which he could do whatever he pleases with.
Possible Consequences of the Split
The only precedent of this sort of split took place with a cryptocurrency called Ethereum, which split into “Ethereum” and “Ethereum Classic”. Let’s analyze the meaning of such a split.
First, lets see what happens to investors that are holding Bitcoin and how does it cope with Bitcoin’s non-inflationary definition. It might look as if there is a contrast, since there should only be 21 million bitcoins, and there will now be 42 million.
However in essence, the purpose of the fixed limit is that when you hold a certain percentage of the total currency base, you’ll keep on having that percentage, and no one can issue more than the total currency base and dilute your holdings. So however you define Bitcoin, as either one of the sides of the split or as both of them, you’re still holding the same percentage of the total currency, because you have X of both.
Of course, the combined dollar value of both coins can drop or rise after the split, but that’s no different than how Bitcoin’s price can change due to changes in demand.
A split could be short, medium or long term. The chain might split and then, after a while, one of the sides will lose support and become abandoned, while the other one rises to be the one and only Bitcoin. Under such a scenario it is like the split never existed.
It could be that it takes a long time until only one side is left, and it could also be that both sides of the chain are long-lasting as independent currencies. In this scenario, a struggle is expected over the Bitcoin brand name. At some point one or both coins will have to change their name to something else.
Such scenarios are referring to a split that’s done in a smooth clean manner. Nevertheless, due to our inexperience with such splits, there could be a few problems, the biggest of which is a Replay Attack. Since both coins are based on the same original protocol, a transaction meant for one of the networks might get processed in the other network.
This means a user that meant to pay for something with Bitcoin A, will accidentally also send his Bitcoin B. This means he might lose his Bitcoin B, depending on who he sent it to. Mechanisms to resolve this will have to be developed. But that’s an issue for a whole other post….
The New York Agreement
To further a solution for the protocol upgrade without causing a split, a compromise has been put forward between the SegWit camp and the 2MB Hard Fork Camp. This compromise has many names – The “New York Agreement”, “Silbert Accord”, “SegWit2x” or “BTC1”.
The idea is simple – Activate SegWit first, and then hard fork to 2MB within a few months. Numerous companies and miners have already signed this compromise.
The mechanism works as follows:
Every miner enforcing the agreement will signal his agreement in the blocks he mines. If more than 80% of the miners support the agreement, there will be a new rule determining that block not signaling for SegWit adoption are invalid. If this happens, all of the miners will start signaling for SegWit in order to not have their blocks ignored.
Meaning – the final state for a node that support the NY Agreement is identical to a node supporting UASF, only the condition for activiation is different. If a big enough majority of miners support the NY agreement, UASF becomes irrelevant – everyone will only mine blocks signaling SegWit and there will be nothing for them to deem invalid.
Signaling for the NY agreement starts about a week before August 1st, by then we’ll know much better where things are headed.
Nevertheless, even if we are spared from this split on August 1st, Miners supporting the NY Agreement have also agreed to do a hard fork early in November. This may reopen the possibility of a split. If the NY agreement fails, there could be a number of different splits possible on August 1st and the following weeks.
So what should you do now?
Stay calm. Right now the most likely scenario is that there will be no split. And even if there will be, you can prepare for it with a few simple steps.
Keep your coins in self hosted wallet with control over your private keys, and back them up. That’s always a best practice, but more so towards a possible split. If you hold your coins in an exchange you can’t tell how it’s going to handle the split or if it can handle a Replay Attack. Examples of self hosted wallets are Electreum (desktop wallet), Ledger (hardware wallet), TREZOR(hardware wallet) and MyCelium (mobile wallet).
If a split seems likely, don’t receive or send any payments from the time of the split until things clear up. The Uncertainty period is likely to be between hours and days, in which the behaviour of even the most recommended wallets is unpredictable. When secure methods to transact without risk of Replay Attacks are introduced, you can return to transact while following the required instructions. If the wallet you’re using is different than the one recommended for use you can, export your backup key and then import it into a recommended wallet.
Only invest in what you understand and believe in. Also invest an amount you can afford to lose. Turbulent times are ahead, with bitcoin’s exchange rate projected to be even more volatile than usual.
Coindesk just created this flow chart summarizing the (current) possible options. Perhaps it will help make things clearer.
Over the weekend, it is notable that BIP21 was activated, making the hard fork option highly unlikely.
Chris M Skinner
Chris Skinner is best known as an independent commentator on the financial markets through his blog, TheFinanser.com, as author of the bestselling book Digital Bank, and Chair of the European networking forum the Financial Services Club. He has been voted one of the most influential people in banking by The Financial Brand (as well as one of the best blogs), a FinTech Titan (Next Bank), one of the Fintech Leaders you need to follow (City AM, Deluxe and Jax Finance), as well as one of the Top 40 most influential people in financial technology by the Wall Street Journal's Financial News. To learn more click here...