BTC, BTU, and You

You may have heard the terms “Hard Fork”, “Bitcoin Core”, and “Bitcoin Unlimited” in the news recently. If you’re like most casual investors, what exactly a “Hard Fork” means probably isn’t clear.

Understanding what a hard fork is and what it means to bitcoin investors shouldn’t take you more than 15 minutes even if your understanding of Bitcoins is at the lower end of the spectrum.

Consensus

Bitcoin works on consensus. The rules aren’t immutable, and that’s what makes a hard fork possible and problematic. More on that later.

What does consensus mean in practical terms? Well, at one level it means we all, as bitcoin adopters, agree that a bitcoin is a store of value and that it is fungible and convertible into products and services. A bitcoin has worth to someone who believes bitcoins are a store of value. A Canadian dollar likewise has a worth to someone who believes they have value. You would probably have limited success bartering using either in a Tibetan monastery though.

On another level, a bitcoin has “authenticity” just like a Canadian dollar does. There are real ones, and there are counterfeit ones. Some are more valuable than others. A legitimate and authentic bill is accepted everywhere at face value whereas a counterfeit bill has a variable value relative to the counterfeit spotting skill of the recipient.

Bitcoin “authenticity” relies on protocol consensus. If we theoretically agree that a Bitcoin block is made of up to 1 million digits and every 20th digit is a 3, then any Bitcoin blocks that don’t fit that description are “counterfeit” blocks and the bitcoins within them are theoretically of limited or no value.

A Difference of Opinion

If you’re old enough to remember that other Trudeau guy, you probably remember that there have been a few versions of “green” (A.K.A. the Canadian $20 bill). Luckily for Canadians, they have up to this point peacefully co-existed. You can use a plastic $20 with a paper $5 to pay for a sandwich at Schwartz’s Deli and almost nobody will complain.

But imagine a Canada divided. Liberals only accept and hold plastic money. Conservatives only use paper bills. The two factions, forgetting the inclusiveness that Canadians hold dear, refuse to entertain the notion that a type of bill other than the one they prefer could represent value. The other party might as well be passing off counterfeit notes.

Welcome to the world of hard forks.

What is a Hard Fork then?

A hard fork is a difference in consensus that reasonable people can’t agree to disagree on. It’s a change so fundamental and uncompromising that it divides a group into multiple factions.

If a block can’t have every 20th digit a 3 AND have every 40th digit a 4, then there are necessarily going to be two definitions of “authentic” and two versions of “counterfeit”. Two factions that believe in different protocol specifications, and probably a third group that thinks all crypto currencies are untrustworthy.

So Why Hard Fork Then?

There was a time where 10 minutes of mining produced a Bitcoin block, and that Bitcoin block contained 10 or few transactions. Actually, there were many such times in the early days. The 1 MB block size limit was so generous that it was of little concern.

Fast forward a few years and fees have increased from optional and usually 0 BTC per byte of transaction to a level where the simplest of fund transfers can and often do cost $0.45 CAD or more. Bitcoin block space (commonly referred to as on chain storage in discussions on the subject) is a precious commodity people are in constant competition for.

So the solution seems obvious: add more!

Well, a surge in paper currency counterfeiting has an obvious solution too: make a more complicated to counterfeit bill that is simultaneously as good as before but no more expensive.

Easy enough to point out, but how easy it is to accomplish?

The warring factions agree that something has to be done, but neither will concede. They have incompatible solutions

Do the programmers decide how big a block can be?

Do the large mining pool operators (who we hope don’t form an oligarchy against bitcoin adopters’ interests) decide how big a block can be, and consequently how much pressure there is to engage in a mining fee bidding war that ultimately only the miners profit from?

Do we have planned growth? Take a large leap now, then smaller predictable leaps in the years to come? Certainly nothing could go wrong with predictable growth that is disconnected from needs (and thus likely too be too big a leap or not sufficient both potentially driving a need for a future hard fork), right?

Static algorithm? Machine learning? Do we let an algorithm learn by example? What if some bad actor teaches it the wrong kinds of things, it’s not smart enough to spot that, and there isn’t enough competent and motivated helpers to give it good examples? What if there’s a cost to teaching the ‘right answer’ but no cost to giving a bad one?

So What is BTU / Bitcoin Unlimited Then?

Basically, it’s a strategy of self-regulation with no fixed limit on Bitcoin block size.

Miners can put as many transactions in a block as they want.

The theory is that miners won’t fill blocks with junk transactions because that could potentially cost them their reward for mining if the block doesn’t get recognized by other miners fast enough.

That’s one risk for miners with Bitcoin Unlimited.

Also, if there are unlimited transactions possible, then why would anyone pay a fee? And if there are no fees, and block rewards keep shrinking, why would anyone mine? And if nobody mines, are there still bitcoins?

That’s another tangible risk.

I propose the following inquiry:

If you sort all the bitcoin blocks mined for the last year based on the internet connection speed of the node that mined it, what’s the lowest speed found for the fastest 10%?

The fastest 20%?

Now, if we look at the nodes with the best connection speed for the top 80% of transactions, what’s the worst case scenario for block propagation?

If 80% of mining success is covered in a given time span, there’s only a 20% chance that a block is orphaned in that span.

Factors include: how many nodes are there in that 80% (only perhaps 15 mega-pools?), how fast can they be propagated to (function of each of their download speeds, whether they are peering with the node that mined, and most importantly the upload speed of the successful miner), and how fast the rest of the network can then get the blocks to cover the other 20%.

If we assume 30 seconds is a reasonable speed for propagation, and assume the successful miner has a gigabit upload speed, and there are 30 nodes that represent an 80% coverage of hashing power, that provides a current upper limit of:

UL /node = 1 Gb/s * 30s / 30 nodes = 1 Gb = 128 MB.

Bitcoin unlimited certainly seems like it could support such a high transaction volume that fees would only be necessary if an oligarchy of mining pools were to force them upon bitcoin adopters.

So, with no fees and diminishing returns, with rising energy prices and steady pricing per mining unit, would there still be a reason to mine?

If you think so, you might be part of the faction that wants to back bitcoin unlimited.

If you don’t, maybe you want to stick with bitcoin core for now.

So What Happens to My Current BTC?

Best case, the hard fork is treated like a stock split. Quebec Corporation A splits into Quebec Corporation B and Quebec Corporation C. 1 share in A grants you 1 share in B and 1 share in C.

They operate independently.

Neither is initially as valuable as the original. The sum of the parts should hopefully equal the whole—or ideally be more than the whole. But maybe they both perform poorly and become, in aggregate, less profitable than they were pre-split.

It is perhaps too early to tell.

If the sharp decrease in bitcoin prices is any indication, current sentiment is that a hard fork is on the way, and it will bring more problems than solutions.

What’s This Replay Thing I Hear About?

That would be the worst case scenario.

You have BTC address 12345. It holds 1 BTC.

Bitcoin Core and Bitcoin Unlimited become separate entities as a result of a hard fork.

You now have 1 BTC and 1 BTU in address 12345.

The BTU is worth $600 now instead of $1400 the BTC was originally worth. You sent it to address 67890 and exchanged it for $600.

All is well because you have your money and spent your BTU.

But wait, there’s more!

BTU transactions would theoretically confirm instantly. You would pay low fees and get quick service.

But then someone copies the transaction from the BTU network, and sends it to the BTC network.

Now BTC miners think you are sending your 1 BTC instead of your 1 BTU to address 67890, and you are paying a tiny fee that does little to encourage quick acceptance.

Hours later, you notice your BTC is in address 67890. You of course hope that the exchange supports both BTC and BTU so you can do something with it—so that it’s not lost in the ether.

That’s the potential danger of transactions being replayed between forked crypto currencies if there isn’t some kind of guarantee that transactions on one currency aren’t valid transactions on the other.

Going back to the earlier example, you can’t have transactions with less than 39 bytes getting broadcast on both chains. You have to be sure that transactions on one chain start with 1, transactions on the other start with 2. Bitcoin adopters MUST have a strong guarantee that the two types of transactions are distinguished from one another for the currency(ies) to survive and remain even minimally viable.