Isn’t a low number of transactions per second essentially built in to the original ETH spec? If so, surely it’s likely that each transaction should be expensive?
Yes, but that's just because a design goal is for modest computers to be able to fully participate in the protocol, and the low number of transactions was the best they could do. They've also been planning to scale since early days. Here are some of Vitalik's blog posts on scaling back in 2014:
And here's one of his 2021 posts, in which he talks about the importance of regular users being able to run a node, but says a million tx/sec should be achievable:
If I remember correctly, L1 ETH can handle ~13 transactions per second. Meanwhile L1 Bitcoin can handle ~4 transactions per second. At the time it was introduced, one of Ethereum's selling points was that it was able to handle 3x more transactions per second than Bitcoin, IIRC.
300% more is usually considered a good upgrade for most things (Apple is touting a whopping 20% speed increase on their latest iPad, for example). It's just that six years is a long time in this space and they've learned how to make things a lot faster since.
Why use L2 when one can have a better experience with much lower fees (Solana, Avalance, Flow & Near). The new generation Blockchains like Solana and Avalance are far better than Ethereum. They are going throgh some growing pains, other wise, they beat Ethereum to dust in terms of speed, latency and performance. Solana takes a second to confirm a transaction, with 0.00025$ fees and can support more than 100,000 tx/s. While Ethereum take a minute to confirm a transaction and cost about 100$ and can only process about 15 tx/s. Also, Ethereum is about 1000 times more energy hungry than Solana.
Solana and Avalanche are fantastic L1s, and I am optimistic for their future. But I'm skeptical of their ability to replace Ethereum.
To Solana's credit, they opted to create their own smart contract VM based on Rust, rather than build an EVM-compatible VM. It will take them time to get developer mindshare, but there's a real possibility to build dApps on Solana that couldn't exist on Solana. The same is true in some ways for Avalanche, but I think their current EVM-compatible C-chain means most developers will spend their time porting existing Solidity-based dApps, rather than build something new or novel. Look at Fantom and its DAG-based blockchain.
Having used Arbitrum, Boba, and Optimism, I can say that they're very fast and work very well. How they handle load as we scale up remains to be seen.
I still have to spend time with the zkRollup L2s; sad to see that StarkNet gets mentioned before Loopring. Polygon's investment into the space is interesting too.
To me, lower fees are irrelevant if there are no dApps to use, and low liquidity. I'm an avid user of a very promising dApp on Arbitrum that has done very well since launching in August. However, they're looking to launch on Avalanche now due to the low liquidity and uptake on Arbitrum.
What that also tells me is that momentum matters. And right now, Ethereum has got it. You won't be bridging your NFTs from Ethereum into Solana anytime soon. But on the L2s? Yeah, probably.
> To Solana's credit, they opted to create their own smart contract VM based on Rust, rather than build an EVM-compatible VM. It will take them time to get developer mindshare, but there's a real possibility to build dApps on Solana that couldn't exist on Solana.
Good point. What are your thoughts on Cardano, Algorand and Tezos?
Interesting points about AVAX's C-chain and Fantom as well. I have a friend who did a project on Fantom due to lower gas fees and it also being EVM compatible, but yeah, it's using Solidity.
Also, just wondering, do you agree with this quote from the article?
> What most people don’t know is: Gas fees on Ethereum are supposed to be high. The goal for Etheruem is not to be the chain most consumers transact on. It’s to be the settlement layer for a number of other chains sitting on top of it, which can run much faster and much cheaper, because they’re backed by Ethereum’s security and infrastructure. In other words: the consensus layer for a variety of networks.
I'd also be curious to hear once you have time to look into zkRollup L2's.
This is all at the sacrifice of decentralization zkrollups and zkporter (one roll up solution) will eventually be able to handle transactions at that clip while being decentralized. The energy consumption point is incorrect as moving to proof of stake Ethereum will be using a fraction of the energy it uses today, however at the cost of capital concentration in the hands of validators which I actually see as a negative for all of these proof of stake blockchains.
I think the scariest thing that people think of and I originally thought when I heard Ethereum will scale with "Layer 2's" is you are sending your money to a blockchain you do not have the keys to do not actually control the funds. The layer 2 scaling solutions will seamlessly interoperate with the current dapps that will be ported over looking seemless for users. You are essentially holding your capital in a contract you can call on at anytime and retrieve to the settlement layer (Ethereum layer 1).
These are the trade off imo and do not make either system better. I just prefer decentralization if I'm holding capital on a blockchain for an extended period of time.
What makes you think Solana is Centralized? It runs over 2000 nodes. The number is currently is less than that of Ethereum node count, as the demand is less.
It also would be a fair comparison to compare what is existing today. So, please don't say that Ethereum 2.0 will be far more efficient. By that time Solana would have made great strides too. So, let's compare what is existing today. Ethereum 2.0 is being touted for several years and still not ready. We can compare Ethereum 2.0, when it's ready and there, but not now.
It has 1200 validators, compared to 260,000 on Ethereum, and it's basically impossible for anyone to run a node at home because it requires 300Mbps+ internet connection to run one.
The problem with this you can't yourself verify what is happening on the chain, you have to trust that the node operators are acting in good faith and not censoring transactions or forming cartels to manipulate history.
This is ok if you only have a small amount of money or are using it for games or something that doesn't really matter. But it'll never be able to run the global financial system or store things of large value.
The connection bandwidth, and to a lesser extent the computer hardware, push validators to run in data centers. The reliable fat bandwidth is a big requirement which drives centralization.
If most Solana validators run on AWS, then it is as centralized as the AWS datacenters. I think they have 50+ facilities now.
These other chains don't beat Ethereum in terms of decentralization and uptime. It tells you a lot about the tradeoffs made when Solana has gone down several times in the past few months and the solution is for all the validators to get together and collectively 'restart' the network.
Ethereum has proven that many dont care about decentralization in the short term, so ETH will be forked to death over the coming years to achieve lower fees and high throughput instead.
If you scale up Avalanche's fees to ETH's market cap, the fees are essentially the same.
The low fees on Solana make it more prone to spam. The entire chain was down for 17 hours in September and has recently been facing some major congestion issues.
There's no silver bullet here and I think we should be honest about the capabilities and tradeoffs of each approach to scaling.
> The new generation Blockchains like Solana and Avalance are far better than Ethereum.
There is only one metric that matters and that's users on your platform. I like Solana and am somewhat neutral on Avalanche but they just dno't have the users, or developers or ecosystem that Eth has.
If they were truly better then they'd be winning. So far the users have voted with their dollars and developers have voted with their time and claimed Eth is better.
I hold alot of SOL, but atleast ETH is decentralized. SOL is very much centralized both in terms of ownership and in terms of nodes.
I mean, they managed to get all nodes to shutdown in a coordinated manner a month ago. I was fine shutting my node down but it was strange that all node owners could get together in a discord group and make that decisions in under an hour.
I like Solana but it has gone down several times. Not totally an apples to apples comparison to ETH, which I don't think has ever gone down (someone please reply if that is incorrect). ETH did need to be forked, but that's a different story.
Avalanche does seem promising. I like their concept of multiple chains: Exchange Chain (X-Chain), Platform Chain (P-Chain), and Contract Chain (C-Chain) [1].
Did you even read the article? This kind of misunderstand is exactly what it's addressing. L2's give instant confirmation, process 1000+ TPS, and cost far less than the L1 (ZK Rollups being a couple of cents per transaction), AND they are insanely more secure than Solana.
Solana is ok for now if you don't care about decentralization, but it barely handles more than a few thousand TPS without falling over or having degraded performance (as has happened twice in the last week!), and it can't scale much further with it's current structure.
A rollup centric system can scale to 14 million TPS by 2030 once a few core pieces of tech are released (see https://polynya.medium.com/conjecture-how-far-can-rollups-da...) on similar hardware to what Ethereum is running today (raspberry Pi's of 2030)
I only have experience with Flow, but it's not decentralized. The developers claim it is, but they hold master keys to remove contracts from the chain.
For fundamentals, this resource by Varsity (kind of like Robinhood of India) is pretty good. It uses India's stocks as examples and deals in Rupees and not in Dollars.
After learning the basics, I would recommend "The 80/20 Investor" book. The book has very good advice how to buy stocks, market bubbles and building the circle of competence.
They actually pushed ChromeOS and web services (GMail, Inbox, Docs, Sheets, ... ) pretty hard and realized that its not the way to go forward. While for lightweight tasks (like emailing and docs) it works pretty well, for heavy tasks (like video editing) and most importantly, as a development platform, chrome os din't do well.
So they pivoted to Fuchsia OS. A new OS to provide seamless experience across several devices (in home or in vehicle while commuting. Some of the resource can be living in cloud). Its a sort of networked OS. Streaming music to your Google Home, let me fetch that music form the Fuchsia Desktop's cache and stream it to Google Home, rather than fetching it all the way from Cloud.
With out applications, OS is not much of use. Here comes Flutter. Let the developers make apps for Android and iOS in Flutter (currently over 50k apps on Google Play store) and have them run the same app on Fuchsia OS. I believe it will take probably another year or two for Google to bring in Fuchsia to Pixelbook. Let's hope so! I would like an OS as open as Linux, with macOS like user experience.
> and most importantly, as a development platform, chrome os din't do well
Microsoft completely lapped the industry on this with VS Code. It has a mode where it runs the editor backend the cloud, and the frontend in a browser.
It seems like they're experimenting with it, but claim it's just for exploring stuff and is not the "future" yet. I've read on HN that it's much slower than Linux thanks to it's micro-kernel design.
That said, I've been very excited for Fuchsia for a long time.
Or use Google Cloud, which has 90% good parts. Documentation can be a bit pain, but the services themselves are rocksolid. There are no 3/4 queuing services, just one. GKE rocks! Cloud Console is a breath of fresher, compared to AWS. Cloud Shell makes it easy to bypass firewalls for logging into instances and no messing with public keys. It's all managed for you. Use firebase if you are looking specifically for Web and Mobile Apps. Scaling to millions of users or Petabytes of data is no big deal and you don't have to rearchitect everytime your customer base grows by 10x.
I keep reading that Google Cloud is excellent. It is definitely enticing.
That being said, for me, the fact that it's Google is a big detractor. Non existent support, the worst documentation last I checked, and I read a "my account was closed unexpectedly" story every now and then.
Then there's the privacy angle, but the alternative companies are hardly bastions of user privacy themselves.
I worry that I will overcome my fears and start using it one day, only to years later try to undo that decision and wean off it like with Android, Search, Chrome, and (not yet) Gmail/GSuite.
Google's intro/free stuff is by far the best of the lot so def give it a try. e.g. Free VM 24/7 forever.
>Then there's the privacy angle
Don't think that's a big issue at dev level. They're interest in billions of users not tracking thousands of esoteric devs doing work stuff that has limited ad resale value.
I have to use both. I find google easier to do simple things.. but I find it lacks some of the flexibility AWS has for less simple things.
Well, that and Google's managed k8s solution was down for multiple days awhile back when I was doing a comparison. Another reason I use EKS atm.... despite I think GKE is a bit better.
I don't know when this happened. Give GKE a try, it's really amazing. Blows EKS out of water. As per the flexibility is concerned, once you learn how to use Google Services, you can get the flexibility with simplicity. AWS services are too complex and even things like billing require a PhD degree in finance to optimize for anything non-trivial.
What are some good use cases for a .dev domain? For example, .ai for Artificial Intelligence, .io for startups. One thing that comes to my my mind is .dev is good for developer tools and related projects. Any other projects or products that .dev domain can be used for? Thanks!
