Thanks for replying, Trace! Much appreciated.
Just because something is easy to understand doesn't mean it's efficient, sustainable, or the best possible mechanism for accomplishing the task of creating trust.
Why trust that?
Dan Larimer's post on DPOS is the best answer, but I'll summarize with my opinion to say, essentially, the same reason we trust in POW. Those with the most to lose if the network doesn't function properly and those with the most to gain if it does have aligned motivational incentives. In my opinion, the Steemit witnesses represent a far more decentralized structure than the handful of bitcoin mining outfits that control the outcome of the Bitcoin network, but we don't need to get into that as I'm a fan of many different approaches, and don't subscribe to tribalistic maximalism towards any specific cryptocurrency project.
How could the history be changed? How immutable is it? Why?
Using STEEM as an example, the top 20 witnesses by stake-weighted vote determine what version of software is running which validates the blockchain. If they all agreed to run an adjusted version of the software which would replay history and exclude or somehow modify the blockchain, and the stake holders agreed with their decision and didn't vote in new witnesses who plan to act differently, then yes, the history could theoretically change. In my opinion, it would be a change to improve the network, one the network itself has the ability to vote for, unlike miners who are not directly accountable to network users and stake holders.
Some see that built-in option for change and hardforks as a systemic risk. I see it as the complete opposite. All throughout nature, the most dynamic organisms seem to survive, adapt to change, and thrive. If we know this has to happen for this ecosystem to evolve indefinitely, then I'm personally most interested in a system which takes this into account and plans for it through a dynamic process like stake-weighted witness voting. As a species, many will argue, humans have dominated this planet due to our cooperation and communication. Competition, though very helpful in many ways, can also lead to tremendous waste of energy. DPOS uses cooperation for fast transactions and zero fees.
With all that said, it's also open source software. If someone did want to change history (and even got all the stake holders to agree and vote in witnesses who will go along with the fork), the code could be forked and a new chain could be started by those who don't agree, just like we see with Bitcoin Cash, Bitcoin Gold, Bitcoin Diamond, etc. In my view, this system is more dynamic and can retain consensus more effectively because every user has a say.
High transaction fees only hinder some on-chain activities. If we can have a secure blockchain and fast block times (only 3 seconds), I fail to see how that doesn't make the blockchain more valuable for a wider range of use cases.
I hope this didn't come off too much as a "my coin against your coin" argument. I really am interested in knowing more about why you personally think high fees are a good thing if other secure options exist. I also greatly appreciate your time, so thank you again for the great piece you wrote and your reply.
The market seems to be leaning in favor of the higher costs of POW, which lead to higher transaction costs. At least as relate to BTC, as a story of value. For a store of value, no one wants to take any risks with DPOS.
For a blockchain that is supposed to economically competitive with non-blockchain alternatives, lowest cost per transaction is probably a highly valuable attribute.
Hence the decision for ETH to move off POW and move to POS.
In the end we may only have one blockchain as POW, BTC. All the rest are probably going to be better using some variant of POS.
the market is leaning in favour of the token that is changeable for any currency in the world, directly. as soon as a better system comes into play, that is also changeable for any token, and any fiat, btc will falter. they people aren't leaning in favor of high cost POW, they are leaning in favor if liquidity. if you could go to bittrex, or whatever flavor exchange, and turn steem into usd, or litecoin, or ethereum... many more people would use it.