Topic: Bitcoin Core Roadmap visualized - page 2. (Read 5713 times)

A fee market already exists without a blocksize limit, the blocksize limit represents a restriction placed on the market by Core, restricting the market is the opposite of a free market. The blocksize should be based on the real supply and demand for blockspace, this is best determined by the miners themselves in combination with the participants of the network. Not a group of technocrats sitting in an ivory tower. History has shown us that centralized economic planning does work, Core can not possible know all of the variables at play, especially considering that the variables are different for each participant. Cvonvert2G36 actually explained this distinction well.

If you artificially raise or lower the limit, the 'free market' is not actually deciding anything.

AFAIK: One of the engineering challenges that the teams are facing now is a decentralized solution to routing. If they get this implemented correctly then LN should look like number 2.

I don't think it is actually possible to open a channel in which there are multiple parties. However part of the lightning network is to use hops through multiple channels to reach another party that you want to pay. The problem with that is timing as I think using that method is rather inefficient, especially for something that needs to happen quickly as it relies on everyone in between to act quickly. Furthermore, if the hops charge fees, then the transaction fee for that transaction could grow very quickly if many hops are involved. It also requires that each hop have an open channel to another hop and that may not always be the case, or at least it would require several hops.

For those one time payments, going through all of those hops would probably be inefficient. For your everyday coffee that may not be purchased from the same merchant, there could be a network of coffee merchants so that you can still pay for your coffee through lightning by routing it through several different merchants but you only need a channel to one merchant.

I support a free market for Bitcoin, not the "fee market" proposed by Core. It is very revealing that you can not tell the difference.

Not quite. "recurring" at least in this context, mean multiple txs with the same merchant (the one you keep an open channel with).

Not going to address the rest of the post (the RasPi thing) as 1) Lauda wants to keep economics out; 2) It's already indirectly addressed in my posts above.

-------------------------

Putting the fees issue aside. Can any of you guys give me a quick explanation, how in laymans terms, would LN work in practice (an ideal scenario, when everything goes to plan). Would it look more like:

1 - I'd have to establish multiple, separate channels with separate parties (ie merchants) to service recurring (defined above) txs, or
2 - I could connect to multiple people/businesses through one channel (and be able to both pay/receive).

I know both aren't exclusive, so more of a question whether option 2 is possible.

the real difference is that one is not retro compatible(hard fork) the other yes

so with an hard fork you need to upgrade otherwise you will be cut off, with a softfork not

I didn't understand completely what a hard fork and a soft fork are. Does a hard fork require a new Blockchain and a soft fork doesn't?

Well that's contradictory, an everyday transaction is, by definition, recurring, is it not? It happens every day.

I would not consider a raspberry pi to be a small purchase. To me small purchases are less than $10. It is also not comparable to the price of a coffee.

Since a raspberry pi is a one time, not really small purchase, it would happen on chain. Again, one time payments would happen on chain. It is not efficient to do those over lightning.

I've replied in a short manner. I did not want to go in-depth about economics. Feel free to start up your own thread on the economics of LN. I'm not forbidding the discussion albeit I'll certainly not take part in it further. Depending on where the discussion goes I might delete replies and quote them all in 1 big post (as done already).

The discussion of (mostly) the technical side of these and similar solutions.

Hmm, no it's not? Unless you have a habit of buying it on regular basis (every day) from the same merchant. To avoid confusion, "coffee" is commonly used as synonym for small to medium everyday transactions (whether recurring or not). Let's use other example, I want to buy Raspberry Pi with bitcoins, what do I do? Will it be possible, will I get penalised with (artificially driven up) high fee?

Increasing block-size could work, but you still have the issues noted above (don't want to repeat myself).


It's equally ridiculous statement as saying that you can take the 'technical' part out in scalability discussion.
Economics board is definitely not the right place (majority of topics there are not even bitcoin related).
Logically, if 'economics' belong to Economics board, then 'technical' belong to Technical Discussion. Then what's the purpose of this thread? To discuss the colours of infographic?

Sure, your thread - your rules. But you're sending mixed messages, you engage in discussion (including 'economics') and say you don't want to discuss. If it's the latter one, I'll just open a new thread, not a problem.

Buying coffee is a small and recurring purchase, is it not? These are the kind of purchases for which LN is intended for. Additionally I can see a block size increase of sorts after IBLT and weak blocks.
Yes you can. This is why there is a Economics section. Besides, this is my self-moderated thread and my rules apply. Similarly I also delete almost all of the posts of people who are on my ignore list.

You can't take the 'economics' part away to have any meaningful discussion about Bitcoin. And if you don't want to discuss then fine, but then why did you reply to my initial post in the first place?


So is "Blockchain for settlements, LN for coffee" a lie?


Anything more than wishful thinking? You don't need some demand. You need high enough demand converting to high enough fees reward. If you cannot define that potential demand even on theoretical level, then you have a flawed model.

Regardless, I do not want to talk about economics here. This is not the right thread nor section for it. I advise you to stop.

It was explained to you. LN is perfect for occurring payments, not for one time transactions. There will always be demand to transact on chain.

---

---

Let's go back to the technical debate.

I didn't make it.

Even a smaller percentage of engineers.

Then I assume you're in favour of allowing the market to alter the 21 million coin limit? Terrible bit of centralised economic planning, isn't it?


NEWSFLASH: software development is centralised. Because a tiny percentage of the planet are capable of coding Bitcoin.


There will always be, and always has been, A MIX of centralised and decentralised forces steering Bitcoin's direction. All you've proven with your statement is that you've not been paying adequate attention to the facts.

The topic doesn't require advanced knowledge of economics, just a common sense really.

There's a (big) uncertainty whether scaling via block size increase will work or not in terms of sufficient fees, but at least you can project theoretical scenario where it could work fine: sufficient adoption rate -> large number of txs with small fees -> big enough rewards for miners. You don't have such mechanism in the Core's roadmap (not long term), instead you have internal conflict.

Why wouldn't you include small, affordable fee to make sure your tx gets included in block? Why do you think vast majority of current txs include fees? You tell me.


Yes, my understanding of LN may be lacking but I was referring to (incorrect) explanation from knightdk.

So again. LN will be cheap (it will have to be). But you still need high on-chain fees to keep things going (maintain relatively small block size + make sure LN gets traction). So who would create high volume of high-fee on-chain txs, when LN provides solid incentive to stay off-chain? How will you make sure block-size is well-adjusted (you need capacity to stay below the demand, but still you need to make sure that all the txs will eventually get confirmed). You would need central planning to make sure there's a balance (push people on or off the blockchain, whatever currently needed) to get this system working.

It looks like I did not understand in then. However, scaling via the block size limit does not ensure adequate fees either. Why should I include a 'decent fee' when there is so much space in the blocks and my transaction is most likely going to be confirmed 'soon-ish' (transactions that aren't urgent)? This is definitely not my area of expertise. You want to talk about economics in the wrong thread.

It is not a straw-man because I'm talking about the block size limit as a way of scaling. I'm not going to explain why Segwit is better than 2 MB block size limit once more.

Right back at you: How do you know the rate of adoption? You don't. I at least on the other have done some research in this field. Unless something new comes up in the next 5-10 years the technological advancement of processing power (as an example) will slow down. This is why Bitcoin needs to be prepared for worst-case scenarios when scaling or pretty much applying any critical upgrades/changes.

Read the first part of my post.

It probably won't because Gavin implemented a 'bad' workaround for the quadratic scaling problem.

He doesn't make sense either way. If we are to believe in the market dynamic then the 'fee market' would develop naturally.

Your definition of 'high fees' is wrong. Your understanding of LN is lacking and you think that LN is expensive. LN is supposed to be both faster and cheaper than transacting on the main chain. Without the LN (or similar solutions) Bitcoin does not really have a 'decentralized global future'.

So LN is not meant to be a scaling solution? If the average Joe can't use blockchain to buy his 'coffee' due to high fees and can't use LN due to high fees x2 (fee for opening and for closing the channel) for one time transaction, then this roadmap is a joke even in the short term and LN will never get any traction among regular users. You would effectively push majority of people away from Bitcoin.

And it wouldn't even work for micro transactions. If faucets are still around they would likely work as they do now, because it's cheaper and simpler than using LN. If the idea is that everyone would have to accumulate more significant amount of BTC in order to make (overpriced) transaction, then, whatever that is, it's not scaling.

I had a mental image of LN channels opened for extensive periods of time (weeks, months or even years), where you can connect to almost everyone and perform all kind of txs with multiple other parties. That would make sense in terms of making Bitcoin usable by masses, but inherently will discourage on-chain transactions, reducing miners fees reward... which you desperately need to keep both blockchain and LN secure and robust.

Again, I can't imagine it possibly working, because for that to work you need to both encourage and discourage on-chain transactions.

Edit: I see on the infograph that LN channels can be opened for indefinite period of time.

In a capitalism based economy, the fee market would be miners deciding the prices of their product, block space.

They used soft limits for years and would do so again. Malicious block construction could be easily prevented, or at least contained to the equivalence of 1MB blocks.

Core devs with their fingers on this dial is central economic planning, something that used to be a good deal less popular around these parts. Sad thing, that.

The above statements directly contradict one another. You are in favour of market dynamics when it comes to one aspect of bitcoin, and yet against it when in respect of another. And yet you constantly cite market dynamics as an overriding principle, and as the impetus for blocksize increases (and for which you provide no meaningful evidence)





Can you explain the reason for the disparity in adherence to your self-professed principles?

The lightning network still requires transactions to occur on the blockchain. There still needs to be transactions to fund the payment channel and to close out the channel so that the Bitcoin can be spent elsewhere. That would be the source of the fees as well as other on chain transactions. From what I understand, lightning is not very useful for one time payments. It is less efficient and effective for me to pay for something through lightning that I only do once every so often, e.g. buying a new computer. Those types of transactions would still occur on chain. What lightning is good for is for recurring payments and microtransactions like those from faucets. Instead of having a bunch of transactions with tiny outputs from faucets, there is instead one transaction to close the payment channel when the faucet user wants to spend his Bitcoins somewhere else.

Furthermore people will have to reload their balance when they run out so they must close the payment channel and open a new one with more Bitcoin to be able to pay more. It basically isn't possible to stay off-chain all the time, you will have to make some transactions that are on chain.

So basically the funding and closing transactions are the sources of fees as well as other one-time payments and transfers and such that do not need lightning. It isn't possible

Not necessarily. It isn't just about hard drive space. There is also the thing with hashing which doesn't scale linearly. That requires computing power and by doubling the block size, the computing power required is quadrupled so the time required to verify the block is quadrupled. That is something like a couple of seconds but in today's fast paced world where everything is about being faster, those few seconds are an extremely long time.

Additionally, there are problems with the size of the blockchain itself The sheer size of the blockchain makes downloading it take an incredibly long time. With larger blocks, the blockchain will grow faster. That means that more time will be needed to download the full thing for full nodes, more bandwidth is required, and more storage space is required. Not everyone has the bandwidth to download the entire blockchain nor able to support the amount of bandwidth that a full node consumes, which will also go up with a larger block size.


I am not entirely opposed to having a small block size increase, but I think that we definitely need segwit for its actual purpose of preventing transaction malleability and the side benefits of enabling other scaling solutions that don't require block size increases. In the long run, I don't think that increasing the blocksize is sustainable.

Also, for everyone complaining about lack of dates, the timeline is listed right here: https://bitcoincore.org/en/2015/12/23/capacity-increases-faq/