Blockstack community thoughts on block size / Bitcoin XT debate

[ryan]

@jude

Will making the blocks bigger actually address this in a consequence-free manner, or will it just push the problem down the road, and result in more blockchain bloat and fewer full nodes along the way?

People have done simulations of this but it’s hard to see what will really happen until it’s tested in the wild.

I think it’s natural to conclude that the number of full nodes will decrease, but we might just have to live with that. For scaling, the alternatives to increasing the block size are to hook the blockchain up to off-chain systems (like the lightning network or sidechains), which would arguably lead to more centralization, as fewer people would be directly on the bitcoin blockchain, fees would be really high due to high demand for block space, and only the transaction aggregators could afford such an expensive, low bandwidth system, which makes it a clearing system only.

@jude

Part of the reason I’m skeptical is because distributed systems typically scale by growing sub-linearly with the scaling parameter.

Yes, we’re in a bit of a predicament with the blockchain because everyone has to have a copy of the data

For data notarization (what we’re doing with Blockchain ID) we do have the ability to merkle-tree pack operations, but this leads to it’s own set of centralization concerns.

@jude

So, I need to see some extraordinary results from the XT proposal before I believe that it solves the scalability problem, especially if the proposed solution is to do better by a constant factor.

I don’t think Gavin or Mike purport that the 8MB XT modification is meant to solve the scalability problem once and for all, but rather that it is meant to at least play a part in the solution. Let’s be honest, we’re going to need other solutions in addition to any block size increase that we go with. We probably will need to use some off chain systems like Lightning or side chains. But we should at least do our best to keep the price of transactions down by increasing the supply of blockchain space.

@larry

Specifically, how does block size affect projects like blockstore & Blockchain ID?

Right now, if we assume that every block is filled to the 1MB brim and we break this down to a mean of 2000 transactions per block at 500 bytes each, then that means the blockchain can handle 288,000 transactions per day and 105 million per year.

If 100% of transactions were used to register blockchain IDs, and each user required on average 4 transactions to be onboard-ed and 2 transactions per year to maintain their account, then we’d be able to maintain a max user base of 50 million people.

Now assuming we can constitute 25% of transactions, that leads us to a maximum of 12.5 million users. A blocksize increase of 8x would allow us to have 100M, while 80x over a period of 10 years would allow us to hit 1B users.