> the developers of popular Bitcoin software have tremendous power over Bitcoin (e.g. they can trigger block chain forks).
If Bitcoin developers do something that miners don't like, they lose their base. Developers are as invested as miners are and don't want to risk a drop in the value of their bitcoins.
The balance is that developers are accountable to miners and users, unlike central banks that aren't accountable to us.
> Your point was that miners' cooperation is required to successfully upgrade
I didn't say that. My point is that miners will decide which chain gets the bigger hashrate and that chain is likely to be adopted as the "real" chain by users, if it's the vast majority of the miners. Furthermore, the prospect that there is no miner support would make users reluctant to upgrade in the first place.
> My point is that an upgrade can be implemented without any cooperation from miners. A change to PoS would be the exemplar of that.
What do you mean by "upgrade" then? Bitcoin was "upgraded" to Bitcoin Cash, an arguably better blockchain. Some people even consider it the "real" Bitcoin, but that's a minority opinion.
> A change to PoS would be the exemplar of that.
You can't use an event that hasn't happened yet as evidence to support your argument. Maybe there will be a flawless transition to PoS in Ethereum, due to influence of developers and the prospect of lower transaction fees. Maybe there will be ETH2 and ETH at the end, traded as distinct assets.
> This upgrade has been agreed to by >90% of Bitcoin miners
90% of hash power, which is an important difference.
The key is that the current bitcoin developers have rejected this and are not supporting it. And yes, some of them founded Blockstream, but the majority are not involved with Blockstream, nor have we taken a company position on this.
Wake me up when it happens. The Bitcoin project's developers still get the benefit of the doubt because unlike Ethereum, they haven't betrayed the trust its users placed in them. Also, network upgrades happen through a miner voting process, which checks the power of the project's developers. Ethereum provides no such check.
> But hey, looks like the market thus far agrees with my position on this.
Bitcoin is worth considerably more than Ethereum, and that will likely be the case for the foreseeable future in part because in each Ethereum hard fork, the monetary policy changes. Why would anyone park capital in Ethereum for the long term if they can't even be sure what their dilution will be next year?
He misses the point that the developers do not control bitcoin; the miners control bitcoin. The developers suggest improvements which will only be adopted if the miners support those improvements.
The miner's interests are purely profit based, so they will not adopt changes which reduce their chance of profit.
>>My point is that miners will decide which chain gets the bigger hashrate and that chain is likely to be adopted as the "real" chain by users, if it's the vast majority of the miners.
If that were true, then an upgrade that switches to PoS, i.e. a hashrate of zero, couldn't succeed, and it's safe to Ethereum's upgrade will succeed.
>>There's no scenario in which you can force miners to adopt some change.
Your point was that miners' cooperation is required to successfully upgrade:
>>One part of the users can upgrade and wait for new blocks for a long time, while the users that don't upgrade can actually use the system as if nothing happened
My point is that an upgrade can be implemented without any cooperation from miners. A change to PoS would be the exemplar of that.
Bear with me here, but I strongly suspect that this is not the case. Because PoS doesn't need miners, it seems the system can transition to PoS at any time and there's nothing miners can do to stop it. So unless the miners also happen to control over 50% of the staked ETH supply, miners don't have a say at all.
EDIT: Don't get me wrong -- I wish Ethereum could only upgrade with overwhelming miner consent. It would check the power of developers to alter the monetary policy as capriciously as they have done so far.
> However, this upgrade is very contentious with the miners who are entrusted with securing the blockchain against attackers
The whole value proposition of mining is that rewards would shift over time from newly-issued coins to transaction fees. Any solution which offers an alternative for transaction settlement and reduces the demand for normal on-chain transactions and the fees from them is, then, an attack on the basic value proposition of Bitcoin mining.
> Developers and miners can't force users to switch to another version.
Miners can, but it’s costly. By a majority mining empty blocks on the chain they wish users to switch away from they make that chain useless. However, they need to sacrifice profit on some other chain (by not mining on that) to do so.
There was a fierce debate that lead to a hard fork over increasing transaction throughput by raising the block size, so I think it's really not a given that BTC will change to increase transaction speed.
Also, one problem people have with Bitcoin is that the idea of every node of the network storing and processing every transaction is fundamentally not scalabe. Sure BTC can change and adapt, but it's hard to overcome a fundamental design shortcoming.
> any changes require consensus from the miners who create Bitcoins and process transactions, and because it's not in their best incentive to do anything to reduce those transaction fees
Can you explain how this jives with miners supporting Segwit2X, which brings both segwit and bigger blocks?
> There are some Bitcoin maximalists that think Satoshi was a god who figured everything out perfectly from inception. To me their insistance on never changing seems more religious than technical
It's not about getting it perfect. It's about getting it good enough and stable.
Bitcoin had its forks that introduced change. People voted with their legs and chose the stability. What matters is that some developers attempted a change and ultimately people decided they prefer original set of rules.
Not sure what happened with ETH but I feel like it might work a bit differently there.
> How can Ethereum be decentralized if they can so easily keep changing it?
Decentralization doesn't mean it needs to be immutable. It just means that power and responsibility is distriubted to many.
> I'm a total dilettante in crypto but if something is decentralized you can't make large-scale changes to it at the drop of a hat.
Exactly, you're correct. It's not easy for these changes to actually take effect. It takes a majority of miners to vote for and adopt any change/fork. The miners are the ones who are running the decentralized network. Even miners who vote for the change can choose to not upgrade their software to the new version.
> Imagine how much coordination effort it takes to change TCP/IP or HTTP
This is why I'm amazed at the skill of the Bitcoin & Ethereum team's ability to continue upgrading a platform that has no central control.
>Developer centralization is another important point. BTC arguably fails at this when a small group, combined with extensive propaganda campaigns, artificially constrains Bitcoin's throughput. There are no valid arguments against raising the blocksize to 2MB for example.
If a handful of people could effectively destroy a “currency” at any point... it’s not decentralized in any sense of the word that is important for trustworthyness or longevity.
A government could crash its own money, but there is a strong incentive to not do that. Bitcoin or Eth has nothing tied to it, it’s entitely at the whim of some programmers who as I understand it, aren’t economics experts.
That’s the point I made above. You want to say the blockchain and hashing is decentralized, great. But when people use that word, they’re talking about (imo) how it’s “safe” to put money into because “no one owns the crypto coin” and if anyone still believes that latter part, well, they probably aren’t paying attention.
> They can't implement those changes uni-laterally, it would cause a hard-fork. In addition to devs and miners, you need users, network nodes and exchanges to agree to it as well.
Sure, but the devs do have the "Bitcoin" name, which would leave them well-positioned to market their changes. Bitcoin Classic may stick around, but like Etherium Classic, it may not be very healthy: https://www.coindesk.com/ethereum-classic-blockchain-subject...
> Miner agreed that software is not a constant of nature. If I want to fork bc at current block and change bc supply from 21mm to 10 billion and offer miners 30% premium over classic bc ... Nothing stops me.
To make a change to the actual behavior (such as to make new money or move money without authorization)--as opposed to merely the order of events (which is important, of course: it allows you to change which of two parties got money if someone tried to send money to both and only had enough for one)--requires the users to all buy into your new fork, not merely the miners; this is something many people seem to get wrong, but it is the most important property to understand: the rules are the rules and no one can change them without pretty extreme buy-in (which does happen, but is quite difficult to pull off).
The case for BTC in this regard is no different from ETH. In both cases developers put together an upgrade, then miners can either upgrade or not.
reply