Don't worry: this isn't another pop science post about space Bitcoin -- it's about scaling and soft forks. Please forgive me if you find the Martian conceit tiresome.

Light takes from 3 to 22 minutes to travel between the two planets, depending on where they are in their orbits. If you're trying to mine BTC on Mars, any block you are mining on is at least three minutes old, probably older -- possibly as many as 22 minutes old. And if you happen to find a block, you're going to experience the same delay before you can tell the rest of the network about it. Odds are, Martian miners aren't going to be adding blocks to the chain. But this is old news.

Mining BTC is out, but what about using it? Given the amount of back and forth communication between Lightning nodes, it seems unlikely that you could use Lightning between planets. At the least, the Lightning nodes on Mars would risk being partitioned off from nodes on Earth -- there might be lots of Martians-to-Martian channels but a channel between Mars and Earth seems a bit risky. But perhaps there could be a Lightning network local to Mars, just with some adjustments for the communication delay.

It's possible Martians will keep using BTC, just with longer confirmation times and greater risk that their broadcast could be embargoed (who knows how centralized communication between the two planets will be). In the worst case scenario, their view of the chain is five or more blocks behind the people on Earth. The Martian portion of the network might get partitioned off if it doesn't have a healthy number of connections to the BTC network on Earth. And if all the miners are on Earth, there is a clear risk that miners could act to censor Martian transactions.

So there are a lot of complications to making Bitcoin work with interplanetary latency, and it's worth asking the question: why not just take the same code as BTC and start a new Bitcoin chain on Mars? Even the Redditers got this far.

Perhaps, we could change a few things: like address encoding to make sure that people don't accidentally lock BTC from Earth to Martian BTC addresses -- just like we already do with testnets -- and possibly fix the 32-bit timestamp problem or maybe adopt a quantum-resistant signature scheme (if it isn't already part of BTC), and probably change the mining algorithm so the mining superpowers on Earth can't 51% attack the Mars chain.

Now Mars has MBTC and Earth has BTC -- both follow the description of Bitcoin in the whitepaper, but they are different chains and have different histories. Are both of them Bitcoin? If they are, did the Martians just scale Bitcoin?

The noob question: "What stops someone from just starting a new Bitcoin?" isn't as stupid as it sounds. Arguably this is exactly what happened shortly after Bitcoin got going.

By the end of 2011, we already had BitDNS, Namecoin, Multicoin, Ixcoin, Devcoin, Solidcoin, Geist Geld, Tenebrix, and Litecoin.

In general, early Bitcoiners seemed to be excited about these new chains. Some of them changed very little from Satoshi's original client. Others altered so much that they clearly didn't even fit the somewhat broad description of Bitcoin laid out in the whitepaper. There were a few chains that were obviously scams, but the most frequent complaint about these new chains was the premines.

Bitcoin is famously bad at scaling. Bitcoin blocks are mined every 10 minutes on average. Because of the difficulty adjustment, nothing speeds this up. Increases in the number of miners, advances in ASIC efficiency, and access to cheaper power all get leveled every 2016 blocks when the difficulty adjustment irons out the wrinkles and gives us a new epoch of 10 minute blocks.

Since we can't speed Bitcoin up, the most common scaling proposal is that we should increase the size of blocks. Bitcoiners seem to have generally accepted the conclusion that this doesn't result in much scaling either because it comes at the cost of reduced security (which is changing the product, not scaling) and it seems dubious that there is some optimal blocksize beyond which we will never need to increase.

If increasing the amount of blockspace that has similar security assurances to BTC is the goal, it seems that a possible solution would be starting another chain that has similar security assurances to BTC. Afterall, if MBTC gets more people using Bitcoin, isn't that scaling?

What do we mean when we say Bitcoin Only? Do we mean this particular chain with exactly these proofs of work and exactly this transaction history or do we mean exactly these consensus rules? Mining is probabalistic and so clearly the blocks that get mined could have a different proof of work and still be considered "real" Bitcoin blocks. And whether you choose to send a transaction today or not doesn't change the reality that the BTC chain is Bitcoin. All of which leads to the conclusion that it's the consensus rules that make the BTC chain Bitcoin.

But if it is a coin's consensus rules that make that particular coin Bitcoin, we certainly can entertain multiple Bitcoin chains with the same rules.

The soft fork conversation is often synonymous with the scaling conversation. There are many proposals that will make it easier to use Lightning or to run Ark, but making those things easier is something that should be evaluated on their own merits.¹ But as far as changing Bitcoin in order to allow more people to use Bitcoin, soft forks are talking about scaling.

I recently read Antoine Poinsot's article on soft forks and noted his emphasis on being cautious when it comes to soft forks:

This is a good approach to keep in mind, especially when the justifications for changing consensus are mostly what-ifs: what if fees get too high, what if government tries to enforce KYC-AML onchain -- perhaps hypotheticals aren't reason enough to change consensus rules.

It also makes me wonder, is there a world where a new chain is a more reasonable scaling solution than a soft fork?