Why should anyone use EOS instead of Ethereum?

[jamesray1]

With Ethereum, you can build Dapps. Why then, would you use EOS?

Update 2018-Feb-6: read https://ethresear.ch/t/against-replacing-transaction-fees-with-deposits/940/2 and https://ethresear.ch/t/initial-explorations-on-full-pos-proposal-mechanisms/925 (for the latter, particularly from "The other important topic that needs to be covered is parametrizing...").

Ethereum 1.0 isn't very scalable, however there are developments in the pipeline such as sharding, eWASM, Casper PoS and zk-STARKs (although zk-STARKs will be a few years away, probably, as they need to be optimized), that will improve its scalability while maintaining or improving security and decentralization. EOS has DPOS, which is too centralized, less secure and transaction amounts are proportional to how much EOS one holds. These trade-offs are not worth better scalability. Ethereum's approach to improving scalability while not making sacrifices is better.

Further reading: https://github.com/ethereum/wiki/wiki/Ethereum-introduction#rd

Granted, Ethereum has issues, but IMO, EOS has more.

[jamesray1]

This post on Steemit compares EOS and Ethereum: https://steemit.com/eos/@trogdor/eos-vs-ethereum-for-dummies.

[jamesray1]

I'm still not fully convinced of the advantage of EOS over Ethereum, with its market cap and roadmap addressing scalability and other issues that EOS also claims to address.

[phroiland]

You're using steemit, and questioning the next iteration of the technology which is EOS? Bravo.

[jamesray1]

Ah, the old personal attack deflection. I'm not very interested in replies of this nature. Having an open discussion of how an alternative blockchain platform to Ethereum is beneficial is of more interest. However, I have confidence in the Ethereum developers, and their ability to address issues with the technology.

[jamesray1]

I fail to think of any problem with Ethereum that is so intractable that it would be better to build another cryptocurrency platform. Raiden and multi-payment channels, plus other functions, can address scalability, which I understand is the main advantage that EOS claims to have. As Buterin points out, not having a gas limit is a bad idea. One reason is that it can make the network prone to attacks or misuse from large transactions.

Update 2018-Feb-6: sharding, eWASM, Casper and other developments in-protocol are more secure than side-channels.

[jamesray1]

And yes, I have used Steemit a little bit. I think doing due diligence is a good idea. Given that EOS documentation does not seem to answer my question well, I have posted it here.

[godsdammit]

Something about no gas required. and way faster.

Maybe Putin's buying for Mastercoin. Loves Vitaluk but who knows. Will the,love last?

[jamesray1]

Yeah you haven't sold me.

[cryptonut]

@jamesray1 you'd rather pay for slow transactions on ETH than get super fast ones for free on EOS?

How about the ability for devs to write in common languages?

eos will be a domain name AND free hosting for that domain, with database, free transactions (transaction capacity of the network is divided amongst number of tokens), you can rent out the token to others who need the transaction capacity...

any of that wet your whistle?

[jamesray1]

Free transactions means no gas which means the network is prone to manipulation from very large transactions.

[ghost]

Dan larimer answered via podcast

https://epicenter.tv/episode/197/

[ghost]

Topics we discussed in this episode

A history of Dan’s previous projects Bitshares and Steemit

How Dan knows when the time has come to move to the next project

Why they describe EOS as a decentralized operating system

The EOS approach to scalability

Benefits and downsides of the EOS fee model EOS’ Delegated Proof of Stake consensus

How EOS handles malicious validators Addressing criticisms around EOS’s crowdsale terms and governance structure

[ghost]

As I remember Dan describe Ethereum as a microcontroller

You can't run Facebook on a microcontroller

[cryptonut]

@jamesray1 transaction volume is regulated in the same way they did on steemit. You can only submit transactions in line with your STAKE. If you own 5% of EOS tokens you can use 5% of the transnational capacity. As such, it would be extremely cost prohibitive for anyone to spam the network...

Anything else I can assist you in understanding? BTW, I think you should read the white paper, it's all in there ;-)

[jamesray1]

I'll peruse the info provided when I have the time, it's on the to-do list. If I have more questions or comments I will say so.

[cryptonut]

Sorry. That's all you'll get from me. I've given you enough to decide for yourself. Good luck

[dskvr]

@jamesray1 This really isn't the place for such questions. If you've read the whitepaper and don't understand the benefits of EOS over Ethereum, then you should read both their respective whitepapers again. Look at the most used Ethereum apps, and then the most used Graphene apps. Compare user adoption. You will see a huge gap between just one Graphene app (Steem) and all the Ethereum apps combinned, one has tens of thousands of active users and high traffic, the others do not.

Ethereum apps require far too much user overhead for adoption. For a user to interface with an Etheruem app, they are generally required to have a plugin and/or a node, or possess crypto currency. The fee model of Ethereum restricts dapps to a very specific business model. Additionally, current throughput of Ethereum limits adoption and general usability of applications. Plasma may or may not address this problem, but is still restricted by gas.

EOS proposes a feeless model that lowers the bar of entry for non-technical users, is far more flexible for businesses, includes native contracts that handle common functions and a toolkit that compliments an architecture that is conducive to rapid development. All the aforementioned can be observed in STEEM, which has the highest female and non-tech user adoption rate for any blockchain based app and went from concept to production in ~6 months. Additionally, Graphene has demonstrated high throughput in both of its high-profile implementations.

EOS proposes to take these lessons from Graphene based app-chains and translate those successes into a platform.

The disadvantages over Ethereum are plainly obvious once you've spent time writing in solidity and attempted to interface with your contract. You'll spend most your time working around it's inconsistencies and throughput issues just to get an MVP. While it's exciting to conceptualize functionality on the Ethereum network, a great deal of friction arises during the actualization phase, where compromises must be made that either (1) cripple user adoption (during quest for ultimate decentralization) or (2) increase centralization (during quest for better usability).

[jamesray1]

It'll be interesting to see how DPOS and Subjective Best Effort Scheduling work in practice, and whether POS can perform as well (or better) in a more decentralized fashion than DPOS.

[jamesray1]

If someone would like to suggest a more appropriate channel for cross-platform discussion comparing the pros and cons of Ethereum vs. EOS then I'm all ears.

[systemsplanet-zz]

This kind of discussion is key to improving adoption, increasing understanding, and helping developers understand the big picture.

How about a trello board with a column for each topic that includes content and links to related articles?

[kunalsharmaau]

I agree this discussion has convinced me that EOS is the best platform to invest in. Good answers btw!

[over9001]

I find this discussion quite helpful. Kudos and thanks to @jamesray1 for sticking with it.

[jamesray1]

Any way with Ethereum you can (hypothetically, it's in the pipeline) have heterogenous sharding and PoS and be able to pick a shard that has two of the following, or a moderate combination of all three, etc.:

• low latency;
• high transaction throughput or low overhead; and
• a high number of validators or decentralization

https://twitter.com/JamesCRay01/status/939455311982731264?s=17

[jamesray1]

With EOS you can have some balance/compromise or imbalance between these three, while it looks to be geared towards low latency and high transaction throughput and a low number of validators. Note that a low number of validators means that you have less attack points, and more generally more attack vectors and a less secure network. E.g. it's easier to have 51% attacks / cartel economics, Byzantine faulty networks, long range attacks, etc.

[randomthrowawaygithubaccount1]

@jamesray1 Try reading @dskvr reply next time. Your question had been addressed in @dskvr 's post above.
In practice ethereum is just as susceptible to 51% attacks? cartel network. in fact it already is run by cartels of mining pools. In the end you ends up with not as much more decentralized, a much harder to scale network with Ethereum.

[jamesray1]

@randomthrowawaygithubaccount1 regardless of the composition of miners in the Ethereum network now, Ethereum is transitioning to sharding with Casper, and will be much more decentralized, and will be scalable with a greater level of security.

[harshitgupta]

first of all, a big thanks to @jamesray1 for starting the thread. I am all up for constructive criticism.

Here are my two cents. I have been following EOS for couple of months now and have listed pros and cons below:

Cons

• Create an account: Hard for a beginner or general public to understand

  1. Need to have an account first to create an account: One would either need to friend, who holds EOS, to create an account for yourself by paying him/her some EOS as its not free or one can use a service like zeos by paying them in crypto. In short, you have to pay to get an account. I don't know how many people would do that.
  2. Need EOS to buy RAM (at-least 4KB): EOS network runs on RAM, so each action you need/want to perform needs RAM. One can create an account with 4KB but you won't be able to do much with it so eventually you would end up buying more.
  3. Need EOS for CPU and Network: Similar to RAM, your actions are gonna need some processing(CPU time) and network bandwidth for the data transfer. Unlike RAM, you don't have to buy but stake some EOS to get these resources.

• RAM Price(Expensive and Volatile) - Although it might only be temporary but its a big issue.

• Key Management - Although it's not a problem of EOS alone but all coins and tokens, that you need to manage your public and private keys very carefully. Can't expect my granny or even my parents to do that. In short, even EOS doesn't solve this problem.

Pros

• No Fee: You can't expect people to pay every time they want to do something e.g. like a post, add a comment like I am doing right now etc. when using an app running on blockchain. Although it costs to have an account on EOS but you can do transactions for free, given that you have enough resources. OK, i know "enough" is subjective but thats exactly the point, for a normal user RAM of 6-8KB would be more than enough, but a heavy user can buy as much as he/she needs.

• Transaction Speed: Transaction are confirmed within seconds as a normal person expects. Here are some facts (at the time of writing) • TPS: EOS -> 3097 whereas ETH -> 15. Also, EOS network claims to be able to do million TPS but we 'll see when that happens. • Network Usage: EOS -> 0.03% whereas ETH -> 100% blockchain capacity + 68,944 Unconfirmed Transactions (Crazy!!!). Can you sell that to a normal person?

• Buy and Sell RAM: so what if you need more RAM for some operation on temporary basis or you have some RAM that you are not using, you can always buy more or sell your RAM.

• Voting/Building a community: As EOS uses DPOS, we have block producers maintaining the blockchain and who gets to be the block producer, the ones which get more votes (top 21). It is creating an awesome feedback loop. So as a block producer if you want the votes, you gotta do something good for the EOS hodlers or the EOS community like build tools or dApps(unless you are a whale 😂). I don't see that happening in ETH because miners won't get anything by doing that.

• Human friendly Account names: It's one of the best features of EOS that it allows properly named accounts. In case of EOS, it's very easy for me to ask my friends, family or clients to send me EOS on harshitgupta. Plus it's easy for them as well. But when it comes to ETH, me telling them to send me some funds on 0xB577F188CfD3Dc1C90497b4ca242752bf7D6ae27 would scare them as its gibberish for one. Second, it's prone to error making. Third, when sent to wrong address, it's gone forever for good.

• Ability to change keys: In case, you know or feel like your private key(s) have been compromised or you have left them exposed somewhere, you can change them right away in EOS, instead of transferring them to new account as one might do in case of ETH.

So above are the pros and cons I have realized so far. I will keep on updating this as I learn more.

Please feel free to criticize. I would really appreciate that and be very much happy to correct myself.

[jamesray1]

Voting/Building a community

This is also a plutocracy which does not necessarily serve best the interests of users. See https://vitalik.ca/general/2018/03/28/plutocracy.html.

Human friendly Account names

This is already doable with ENS.

Ability to change keys

Changing an account with Ethereum is very easy, e.g. with Metamask.

See also https://github.com/ethereum/wiki/wiki/Alternative-blockchains,-randomness,-economics,-and-other-research-topics#eos.

[jamesray1]

Given that this is a repo for documentation, and the documentation does not seem to adequately address the question in the title of this issue, I feel like it's appropriate.