Proof-of-Stake, including delegated, has serious flaws which makes me doubt its viability, even if you guys do shoe-horn in more features.
Reference - https://medium.com/@hugonguyen/proof-of-stake-the-wrong-engineering-mindset-15e641ab65a2
One of the reasons you have to devise all these complex systems sitting on top of your original scaling assumption is because the basic premise is deeply flawed.
Hardfork away, it won't make much of a difference in the long run - especially if you get any real usage numbers. Current estimates are around 45,000 active accounts. (Hat tip to @paulag for that.)
I don't see how you're going to make it work, frankly.
Steem is not PoS, it is DPoS. There are some technical differences.
The artile is flawed, but I don't feel like opening a debate.. I guess the main thing that it doesn't take into account is that new block producers can be voted in at any time. Also, in terms of the geographic concern - when there was a recent issue with the Steem blockchain, there was no issue with a sufficient number of witnesses being pulled in to deal with it. There were lots of people woken up at 4:00am, pulled home from errands, and lots of other stuff. We still got enough people there to deal with it in a timely manner.
You didn't read it - it clearly talks about PoS AND Delegated Proof-of-(mis)Stake.
Listen, even with checkpoints you guys are playing with fire.
But you know what? Go ahead, build your sandcastles in the air, just don't complain when it gets taken down.
I tried, and if you can't see it -- then that's your problem, not mine.
I did read it. The first two concerns assume that there is no mechanism to establish the longest chain and two forks will just continue on. In DPoS, 2/3 of the block producers are needed to establish an irreversible block. If the network split there would not be concensus to continue producing blocks. The network would halt - which is still bad, but not as bad as the scenario outlined in the post. The double spending scenario also does not take into account that 2/3 are needed for an irreversible block. I’m not an expert, but afaik there is not a way to make two irreversible blocks with conflicting data when both would need 2/3 of the block producers to create a block that agrees.
Let me lay this out for you - there are two scenarios for Steemit:
User retention slips or stays at a paltry 13%, and it gradually joins the long-tail of technological failures from lack of interest.
By some miracle users come to the platform, where it promptly seizes up after a frenzy of activity.
Hoping, wishing, promoting or nit-picking technical details isn't going to save it.
You disagree, fine. That changes nothing.
We are both entitled to our opinions, which is really all these latest comments are. Maybe you are right. Maybe not.. It is really speculation either way, so I don’t have anything else to say. I guess we will just have to wait and see.
Neither of them have anything to do with the security of DPOS though, which is what our original conversation was about.
And where are those scenarios mentioned in the article?
What have those two scenarios to do with DPoS?
DPoS is a flawed engineering solution. Just because we're shambling along with 45k users doesn't mean it will handle order-of-magnitude more.
Look at my prior posts, there is a comprehensive technical series about why.
Maybe or maybe not... but if DPoS is flawed and is way ahead of the number of users that any other blockchain out there can handle... Are you saying the BC technology is fundamentally flawed??
POS may have a lot of flaws but it does make it up with the fast it can process more tps.
POW is still the best Blockchain algorithm I seen. POC not bad but could be put into better use.
I wish we could have useful POW, primecoin was an excellent example of this.
"Useful" PoW.
A secured global network isn't "useful" enough for you?
I'm tired of people railing that Proof-of-Work as "useless". It secures a staggering amount of wealth, much more than this shambling Steemit wreck of Delegated-Proof-of-(mis)Stake.
You didn't read the article, so fine -- go down with the ship.
When did i say POW was useless? Look at all my posts, I defend POW to my dear life. Hell I got a comment from Dan about my view and wrote a post countering his point. I'll give you 1 sbd if you can find where i did not say it. And for a record I read that post in June lol
Of course a secure global network is useful but useful POW would also be cool too.
If you have 1 pant that gives your protection from alot of things..
Wouldn't a 2nd pant also help just incase?
I love POW but primecoin did show that POW could be used more efficiently.
Hell POW is still more efficient that banks:
https://steemit.com/bitcoin/@cicbar/bitcoin-energy-consumption
I want more research into useful POW when a scaling solution has matured aka the LN and sidechains. A blockchain need to scale for it to support the world.
Right there, buddy.
If you missed one word, such as "more useful" then that changes things slightly.
You don't have to convince me about PoW, its the only thing that has worked consistently at a global scale without centralized "witnesses" or other garbage.
And? I still want useful POW. The POW we have now is amazing and better than POS shitcoins.
But I did not say POW was useless. I want scaling of the blockchain before looking for useful POW first though.
Actually the overwhelming reason why we have to keep improving our protocol is because so many people use it. The rate at which Steem dapps are coming online continues to increase. The rate of user growth continues to increase. Our "real usage" is still orders of magnitude beyond any other blockchain. Look at how many transactions are being performed by the top Ethereum dapp. If Ethereum has so many real users, why are none of them using their dapps? Even you are using our blockchain to claim it doesn't work. And that's totally fine. Criticism and negative feedback are vital components of any healthy ecosystem and DPoS does count on people constantly "kicking the tires." Personally, I welcome everyone who believes DPoS doesn't work to keep using Steem to complain about how DPoS can't work, thereby proving that it does in fact work. So thanks!
So when the Steem blockchain seized up - that was just an "oopsie"?
You've got to be kidding me that you think ETH is a good example, in fact, they're about to head for their own version of PoS hell with sharding -- https://hackernoon.com/the-ethereum-blockchain-size-has-exceeded-1tb-and-yes-its-an-issue-2b650b5f4f62
And you prove my point by saying "why we have to keep improving our protocol is because so many people use it.".
It doesn't scale at the outset, so you have to bolt more things on to keep up, except those "exceptions" aren't going to help in the long term.
Every PoS coin thinks they have it together, but then when there's actual usage -- the flaws become realized and they are systemic.
But you know what? Go ahead, keep shoveling people on here and see just how badly it won't scale.
The problem with this platform is the slightest suggestion that it isn't manna from heaven is shot down and dismissed.
Enjoy your future implosion.
And then... What is your suggestion??
Halt everything until we have a guarantied way to scale up to 1000000 TPS?
If you are at the top of the wave and you are able to keep improving regularly...
Doesn't it paint a bright future to you??
If you were paying attention, you'd know that what I'm saying is Steemit is fundamentally flawed.
No amount of hard-forking, glossy features or rampant promotion is going to fix it. Unless perhaps they gut the central premise and rearchitect the centralized system they have.
I doubt they'd do that, so it seems failure is assured.
Infinite loop commence
3
2
1