The block size raises serious concerns regarding OpenBazaar, here is why:
Alipay clears 30 million escrow transactions per day, which works out to be roughly 350 transactions per second
Alipay has at least 800 million people on their platform (i.e. Alibaba, Tmall, Taobao, Aliexpress etc) to conduct e-commerce within China and abroad (hell, I have an Alipay account and I live in Australia; Alipay crushes PayPal btw)
For Bitcoin, 2000 transactions per block is roughly 3-4 transactions per second. This is going to be a major bottleneck for OpenBazaar’s ability to scale… and yes, we’ll need Blockchain IDs for our users on top of regular transactions for goods and services. We can’t even grow beyond 1% of the Aligroup with the current spec (assuming OpenBazaar takes up all the tx volume of the blocks lol).
People have reacted to this by saying that Satoshi was naive if he thought that Bitcoin could scale to this level… but history isn’t too kind on people who say ‘X won’t scale’. Anyway, my point isn’t to argue for or against the block size increase, a BIP proposal, XT etc. Rather, it is to highlight that the change of vision from ‘anyone can make an on-chain transaction’, to only a an elite group who can afford on-chain transactions, will have profound effects on our projects.
OpenBazaar will figure out how to make payments and escrow work if we’re stuck with 1 Mb blocks via the Lightning Network or whatever off-chain payment processor is developed. However, settlement transactions are still required for the Lightning Network, so eventually micro and small transaction settlements will be priced out of the blockchain.
1. A secular trend upwards in the transaction volume.
Assuming that the adoption of Bitcoin rises in the future, more on-chain transactions will be submitted to the network until the block size limit is hit at 1 Mb.
The transaction volume will be *offset by delays in transactions with an insufficient fee and/or low priority, as well as a self-regulatory effect where users do not submit transactions due to the increase in transaction fees.
What is key about this point is that the minimum transaction fee to be included into the block rapidly will be difficult to determine, contributing to a general degree of uncertainty when transactions are submitted to the network
Despite the offset effect described in the previous point, the adoption of Bitcoin continues to rise and the transaction volume follows the secular trend upwards.
While the on-chain transaction volume is hard limited in the protocol, the mempool competition to get your transaction included in the block would persist and possibly increase with time
The much touted fee market will attempt to find a market clearing transaction fee, which will establish the second major assumption below
2. A secular trend upwards in transaction fees
With the overall block real estate artificially suppressed in the protocol, and a secular trends upwards in transaction volume, fees will rise
The transaction fee tends to decrease with an uncapped or high cap in the number of transactions per block… but when the block limit is hit, this trend will reverse
The transaction fees can be offset by transaction bundling (i.e. coordinating users to combine multiple inputs for a single transaction, reducing the overall share of the transaction fee between the users)
This is constrained by the fact that the more inputs/outputs, the larger the transaction, the higher the fee
The lightning network (LN) may have a profound effect on reducing the on-chain transaction volume, but eventually (maybe even years and years into the future, starting from whenever it’s ready) micro and small transactions won’t be safe to make on the LN
From my understanding (and feel free to correct me here), in the event of a dispute in the LN payment circuit, the transaction can be settled on-chain.
The idea is that the LN will take so much transaction volume off the blockchain, as well as offering instant payments etc, that making the occasional on-chain settlement transaction will be feasible.
Problem with this idea is that once space is freed up on the blockchain, people will attempt to send their funds directly rather than going through the potential complexity of the LN.
Very soon we’ll be back to full blocks and a return to the secular trend upwards of transaction fees
At some point, the transaction fees for on-chain transactions will make micro and small transactions economically irrational.
Let’s skip microtransactions for a minute, although keep in mind a Westerners’ microtransaction is a developing world salary for a day or year in some cases.
What is a small transaction? It’s obviously highly subjective but let me keep it on context to OpenBazaar. The value of the average ecommerce transaction is ~$100 USD. Once the transaction fee is conservatively over the $5 mark, there is no point using Bitcoin for ecommerce on-chain transactions.
Over the LN, there is a risk that if something goes wrong, settlement cannot be made.
Now in these points above, I haven’t addressed centralization or counterparty risk with miners and/or LNs if the block size remains the same or increases, I have tried to keep it focused on the practicality of using Bitcoin for ecommerce and more specifically, OpenBazaar.
Very keen on hearing any corrections or alternative considerations from this community.
Eh, it won’t happen like that I don’t think. Instead, I think it makes more sense that the two would happen simultaneously, meaning that the overall effect will simply be a reduced fee than otherwise would be without LN.
So your set of on-chain transactors would be X without LN, and X times some factor with LN.
I’m guessing that LN can be used with combined transactions (that would make sense to me intuitively), but I might be wrong. I will ping Poon about that. (Heh.)
And yes, txn fees for any blockchain are likely to increase with the more users it has, that’s just a fact of life and there’s not much the block size can really do about it (block size is not a sufficient mechanism for scaling blockchains, as many others have pointed out).
Yep I agree, my point here is that users will punt for direct on-chain transactions if the LN takes away a large volume of the on-chain transactions… so you’ll get an equilibrium of people using LNs and on-chain transactions. However, the secular trend upwards of transaction fees continue pricing the value of certain transactions off the blockchain and ultimately off the LN.
Every efficiency gains from some ideas here will offset that that trend, flatting the curve for a time, but constraining the block size will accelerate that trend. But like you and others have mentioned, few if any are against raising the limit, just how and when.
block size is not a sufficient mechanism for scaling blockchains, as many others have pointed out
It’s definitely not, but it is the one with the most profound impact at this stage. Nothing that the Core devs have done so far have drastically impacted the number of transactions per second (they have improved things in other areas of scalability, don’t get me wrong), which is the metric I care the most about when it comes to ecommerce and OpenBazaar.