Topic: Bitcoin Core Roadmap visualized - page 3. (Read 5713 times)

Lauda, I do like the visualization you made, not its content obviously but its design is very good, that is well done.

I am talking about two megabytes specifically. Stop creating straw man arguments of doomsday scenarios if we increase the blocksize, we can increase the blocksize to two megabytes and we should.

How do you know the rate of adoption, or for that matter the rate of technological progress. I would say that this is a unqualified statement, I am saying increase the blocksize today according to the technological limits that exist today.

First of all we can definitely afford to increase the blocksize to two megabytes in terms of hardrive space today, you should know better then to use that argument, and again I am saying that it can reflect the technological limits of today.

This idea that we can not scale Bitcoin to the whole world today therefore we should not scale Bitcoin any more directly today, allowing a "fee market" to develop, I think is intrinsically flawed. Such a dramatic change to the economic policy has profound effects, the more conservative route would be to continue the course, if we increase the blocksize and the majority chooses to use off chain solutions over using the Bitcoin blockchain directly then that is fine, then we would not need to increase the blocksize as much again and it might stay in step with technological growth. I do not see any reason to restrict the capacity of the network using a centralized economic policy tool, in order to "incentivize" people to move off chain. This seems counter intuitive to me, even wrong in the way that it is being done now, if anything we should encourage more on chain transactions for the health and good of the network, allowing the free market to develop.

The Core guys are very good at what they do, however I think they lack insight in the macro economics of Bitcoin, that is fine. Bitcoin has a governance mechanism however, maybe we could just have the best of both worlds. A hard fork to two megabytes blocksize limit soon, then segwit can be implemented when it is ready, fully polished and reviewed.

I do not think it makes any sense to endanger the economics of Bitcoin over a two megabyte blocksize limit increase. Core might not understand how important the timing of such things are and expect people to pay higher fees and maybe even go through a "economic change event" until their technology is ready, when they could just simply increase the blocksize now and avoid all of this potential damage.

Recognize that there is a governance mechanism involved here, and that fundamentally changing the economics of Bitcoin is not something that should be taken lightly, beyond all technical arguments of the superiority of segwit Core does not intend to increase the blocksize limit significantly over the long term at least not enough for the limit to be above the average transaction volume, they have explicitly stated their intention in this regard as well. They might be experts in IT and Engineering but I think they are seriously lacking in economic knowledge and real world pragmatism and politics. Fortunately Bitcoin relies on the economic self-interest of the masses to govern consensus, and is not reliant or beholden to a group of technocrats in an ivory tower.

Increasing the blocksize limit to two megabytes is possible, and will not cause any sort of doomsday for Bitcoin. If Core refuses to do so the community, miners and business will do so themselves. Bitcoin is freedom.

Yes, it could work. It's not a pretty solution and will probably face more challenges later on, but at least there's a realistic scenario where everything could work out fine. I can't really see it for Core roadmap (which could work truly amazing, but not the long term).


Personally I don't see Satoshi as a god. I don't have any problem with people suggesting to depart from his vision, but only as long as they're open about it and seek for community support, rather than using mental gymnastics and pretending that complete change of how Bitcoin should function is somehow in line with original design.


I don't think you got my point.
First of all, lets ignore potential high BTC price, I'm specifically referring to situation where block subsidies are irrelevant or non existent.

This system is internally conflicted. To dumb it down: to keep small blocks (and for LN to get/keep traction) you need high fees (the smaller the block-size, the higher the fees). While users need low fees (otherwise BTC becomes completely unattractive and will lose to altcoins/fiat).

For that to work you would need to have some sort of constant balance, basically you'll need to find a way to keep average tx fee on optimal level (if such can be determined), keep number of on-chain txs on high enough level, prevent the channels to be open for too long (to enforce open/close on-chain fees) but let them be open long enough (to keep LN usable). On top of that you would need to be able to adjust the max block size ad hoc (both increase or decrease if needed) to adjust to current needs. Essentially you'd need central planning to get this thing somewhat working. Also, you would definitely need much bigger blocks.

The above would also require 'fee market' (keeping txs/sec capacity below the demand) which is a flaw on its own. If you had constant demand of 10 tx/sec and capacity of 7 tx/sec, that means even if all of the 10 senders put all of their wealth on paying tx fee, at least 3 txs don't go through. That's not what reliable, sustainable model looks like.

That's not exactly my point not was I talking about 2 MB in particular. In the case of wider adoption scaling primarily via the block size limit will never be able to accommodate enough users without sacrificing decentralization. Technology is currently in a 'difficult' position. On one hand we have fiber internet which is a huge improvement over the current world average, on the other hand we are witnessing the death of Moore's law and problems with HDD's. It is quite possible that technology advancement in some fields become really slow (look at the battery technology), however it is also possible that this won't happen. You can't work towards a best case scenario here when scaling large scale systems.

I've told you this a number of times and I'll tell you this again. I'm a very open minded person, especially when it comes to this debate. There was a time that I've agreed to a bigger block size limit; I'm supportive of various proposals such as LN, sidechains, dynamic blocks and even a hard fork (I do think that the system could use the experience in addition to giving us the opportunity to apply certain fixes/improvements). However, in the case of Segwit vs. 2 MB block size limit I'm fully supportive of Segwit. Additionally I highly disagree with the HF rules set by Gavin and the 'workaround' used to avoid the problem of quadratic scaling.

You don't even know how to use a fallacy unless someone uses it beforehand. I'm not appealing to anything. I have at least stated on one occasion that the Core developers did not handle this 'block size debate' situation as best as they could.

That is not necessarily the case as technology improves. We can even just scale Bitcoin directly as technology improves, decide on a lower bound, lets say the average desktop computer in the developed world for instance, even if that was our lower bound we could still justify increasing it to two megabyte once like we are now. It is not my preferred position but I would compromise there, however it seems like you are not even open to that possibility, using the blocksize limit as an arbitrary economic policy tool instead, while using IT and engineering knowledge to justify such a position, using Core as an appeal to authority.

I've revised this statement. There are various ways in which it could work. In general there are going to be more on-chain transactions regardless of off-chain solutions. The users need to open/close their payment channels as well. Additionally the price of Bitcoin could pick up the 'pace' where we need not worry about fees for a considerable amount of time. Increased block size limit does not guarantee an adequate amount of fees in the long-term future either.

This would effectively mean the end of Bitcoin as we know it today.

Because it will never be able to scale that way to accommodate enough users without sacrificing decentralization. Each time you raise this limit by a considerable amount the problem becomes worse. Besides, unlike Classic the developers behind Core are working on infrastructural improvement that should make such 'upgrades' less 'heavy'.

Nobody cares what you believe in. Stop appealing to authority, this argument is without merit.

I do not see any reason why the "Classic Solution" can not work. Just increase the blocksize, increase it again when we need to, the market will find its own natural equilibrium.

Fortunately the solution already exist, we do not need to reinvent Bitcoin, Bitcoin is already the solution. We have multiple alternative implementations who are staying true to the original vision of Satoshi who are prepared to scale Bitcoin directly now. The way it was always intended, we should allow the experiment to continue, I do still believe in its original vision.

So are you against Core map or pro?
Blockchain is a backbone of LN, with no incentive to miners both blockchain and LN fail.


Great. But how?

Again, there's a conflict in design, you need both average users to stay on-chain and off-chain. I don't think even central planning would work.

So again, I hope I'm missing something here, but if not, that means this roadmap would put 'expiration date' on Bitcoin.
And to make it clear, it's not about "something can go wrong" but about "it cannot possibly go right", well, unless you make some significant changes in entire Bitcoin design.

The only solutions that comes to my mind (at this moment) are:

- lifting 21 million cap
- switching to different algo (PoS?) but that would affect LN
- increasing block size and making txs cheap again (back to Classic solution), but hoping that users would use both, on-chain + LN (probably not likely to succeed)
- hoping that few BTC businesses would take on mining operations on themselves (mine at loss), but so much for decentralisation
- hoping that BTC businesses will be making regular charity donations by voluntarily paying high enough fees.

Does any of the above sound attractive?

I'm bit surprised to see any serious members in full support of this roadmap without having a clue how the above could possibly be sorted out. Or is it just "who gives a f**k, this problem not gonna happen anytime soon" kind of logic?

I think that you are suffering from the engineers nirvana fallacy. I think that the logic that we can not scale to global levels today, therefore we should not scale bitcoin directly at all is flawed. Technology improves and we can increase the limit as we need, restricting Bitcoin now at one megabyte does not make any sense, we can easily increase it to two megabytes without massive node centralization or whatever. Which would avoid transacting on the Bitcoin network directly becoming more expensive and less reliable, which would not be good for adoption.

If you think that Bitcoin has no competition then you also have your head in the sand, Dash for instance can already do fifty six transactions per second, quickly, cheaply and directly without the inconvenience of having to use the lighting network or any other payment channel build on top of the blockchain.

Bitcoin's predecessors where completely centralized, not a good example at all and that is not what I meant. In regards to the internet, think about it, what it would it take to destroy the internet today? Impossible right? Security through mass adoption, not obscurity is the path forward for Bitcoin and this was always the intention.

The internet has given us so much, freedom of information, communication even Bitcoin I find it hard to understand why you think the internet was a mistake and why Bitcoin should not make these same mistakes again. I am inclined towards the opposite believe actually, we should not use blocksize limit as a economic policy tool, it was never intended for that, the limit should be significantly above the average transaction volume, this way it serves its original purpose as a anti spam measure.

Arbitrarily restricting the organic growth of the network should not be considered a good thing.

LN remains the best way to scale Bitcoin worldwide without ruining node decentralization in the process. Until the day a cryptocurrency allows for global level transaction volume, all on-chain without massive node centralization, that is fact. But I don't think that's possible, therefore Bitcoin has 0 competition, realistically.

It should work.

That's not what I said. I said if they could rebuild Bitcoin from scratch they would, but it is not possible right now. What I meant was improve the underlying infrastructure from scratch. The internet is not really decentralized.

Core did not create the infographic. Anyhow this is all much better than Classic regardless of when it comes in 2016.

Transacting via the LN will be cheaper and faster. You don't know what you're talking about. If Bitcoin goes down, so do almost all of the cryptocurrencies.

"As decentralized as internet", so you mean as the other endless iterations of "internet money" that have always end up getting raided by the authorities?

That's nonsense, the first priority should be decentralization of the nodes, "as decentralized as internet" in terms of nodes means datacenters running nodes, which means Bitcoin turns into Paypal 2.0 in other words useless, since we already have Paypal.
I don't care what Satoshi said, the current Core view on reality is the most realistic one and the best compromise in terms of pro privacy and pro decentralization.

I can turn that around actually, who is going to bother with LN and SD when everyone can just transact directly, easily and cheaply using real cryptocurrencies. I would hope that Bitcoin chooses to remain a real cryptocurrency otherwise I am sure it will be obsoleted and outcompeted.

I think that you are spot on, it does not work. Lauda even said it himself, Core wants to rebuild Bitcoin. According to them Satoshi's vision of Bitcoin is not good enough. They are diverging from the original vision of Bitcoin. I think that as decentralized as the internet is good enough. Bitcoin is build on top of the internet after all.

Did you notice how there are no dates in the roadmap whatsoever. Fundamentally changing the economic policy of Bitcoin like this while pretending to intend to scale Bitcoin significantly directly is deceitful. There are fundamentally different visions for the future of Bitcoin that are at odds here.

https://lists.linuxfoundation.org/pipermail/bitcoin-dev/2015-December/011973.html
http://konradsgraf.com/blog1/tag/block-size-debate
https://medium.com/@jgarzik/bitcoin-is-being-hot-wired-for-settlement-a5beb1df223a#.7ek47hakx

I don't think anyone is going to bother with on-chain transations in the future when you can send transactions as safely instantly and cheaply thanks to LN. So the question is: Who is going to use on-chain transactions so the miners get a decent amount of money to justify keep validating transactions?
I mean if LN is as safe as on-chain transactions.. why even use on-chain. Maybe if you are buying something huge like a car or something, but even then, if it's safe to pay with LN and not pay higher fees..

Nice infographic.

What bothers me is the bigger picture (distant future). How would this roadmap work in terms of providing sufficient miners fees when block subsidy becomes irrelevant (or disappear completely).

If I get this correctly, in order for everything to work, we would need to have pretty specific, balanced condition where:

- fees are high enough to prevent average Joe to make his regular txs on the blockchain
- fees are affordable enough to make Joe settle his balance often enough to provide miners with sufficient reward.

I honestly can't imagine what would possibly be the right tx cost. If it's too high, Bitcoin is unattractive to Joe, if too low, Bitcoin is unattractive to miners.

Alternatively:

- Joe stays off-chain (but on LN) his entire life (assuming it's possible) and enjoys cheap txs.
- Miner fees are paid mostly by businesses/institutions.

But then, if it's possible to operate off-chain, why would businesses pay high fees if they could settle balances as rarely as possible (never?) to pay lower fees.

I'm obviously missing something here, there's no way no one thought of that, but can't figure it out. Seems to me like there will have to be an artificial bottle neck and some sort of central planning (by adjusting max-block size) to get it work (but even then it doesn't make sense).

Could anyone be so kind and paint me hypothetical scenario how this could work out?

Bonus question: what % of BTC users would have to start using LN in order for it to get desired traction (I vaguely recall reading somewhere that it's >60%).

The Infographic was updated with corrections from BtcDrak, Adam Back, Eric Lombrozo, and Luke-Jr.

It wasnt relevant to the OP. It was relevant to Carltons post.  He said 'quadratic' was raising to the "fourth power"

Incorrect; how? Provide a source please.
I took it directly from a website long ago; now I've found it again. Take a look for yourself. I'm not sure what you're trying to say here nor how it is relevant to OP?