9 ms·

Scaling: Ethereum’s make or break challenge [video]

103 points by dsr12 8y ago

obituary_latte 8y ago
Off topic but I am so entirely sick of amp/mobile pages like this that just fail so hard at delivering the advertised content. Multiple reloads, pop-up conditionals, auto-scrolling...yuck. Clicking the link briefly flashes the title of the topic followed by a blank square with a “close” X then flashes a couple of times and I land on a listing of videos none of which are the op topic. I’m not using ie4 or something - this is the current version of Safari on the latest mobile device from one of the biggest phone manufacturers that exist! All on a “tech” website. First world problems I know but goddamn is it frustrating. Sorry for rant.
- dvh 8y ago
  Same on Android Edit: it works now but it is badly broken video player, stopped when I tried to seek forward
- plasma 8y ago
  I hate AMP as well. They’re less maintained, feature less pages (see reddit), and they have obnoxious script that prevents hold-tap from allowing “open in new window” from working.
- niksakl 8y ago
  here is a youtube link to the video https://www.youtube.com/watch?v=976Q2cu1_DE https://www.youtube.com/watch?v=976Q2cu1_DE
SwingingShips 8y ago
Ethereum is talking about 2020 as its target for scaling upgrades. There are other projects that haven't launched yet that probably have an advantage over them, as so long as it can integrate with Ethereum rather than claim it can totally supplant it. There are projects like Cardano and Zilliqa, but then there are some like Orbs that are using blockchain virtualization claiming it can work in tandem with Ethereum. I assume Ethereum will implement some of its solutions before that 2020 date, but they will probably need to rely on more partnerships to do it.
- haasted 8y ago
  https://cosmos.network/ https://cosmos.network/ will also have an Ethereum bridge, as well as proof-of-stake.
- StavrosK 8y ago
  Cardano has stalled for so long that I stopped waiting. I hear great things about EOS from people I trust, though, I need to look into that.
  - JohnJamesRambo 8y ago
    EOS is like a really expensive AWS. There is nothing revolutionary or decentralized about that. I don’t know what you mean about Cardano, they have a very detailed roadmap and haven’t missed any major milestones I am aware of. https://cardanoroadmap.com https://cardanoroadmap.com
  - shinryuu 8y ago
    The launch of EOS has not been very well received at all. Lots of mistakes and the governance is a mess.
    - mihaifm 8y ago
      Indeed, EOS relies on a constitution with human interpretable rules [1]. Not sure how that fits into the crypto sphere, where governance should be done be algorithms and code. [1] https://github.com/EOSIO/eos/blob/5068823fbc8a8f7d29733309c0496438c339f7dc/constitution.md https://github.com/EOSIO/eos/blob/5068823fbc8a8f7d29733309c0...
      - rhyzom 8y ago
        and it's also supposed to be hashed in each "legitimate" transaction. while network resources are so scarce as to stir rampant speculation. idk - the only thing about EOS that i have noticed is how overtly it screams of greed. the people that jumped in on that bandwagon did so for no other reason but hoping it would make them rich (cos Daniel Larimer said "you can run Ethereum on a single EOS smart contract", lol).
Egidius 8y ago
Great to see these young guys attempting to tackle these huge problems out of their believe that they can make "the world a better place". Not sure DLT will live up to its promise, but these guys truly believe the world is better off with trustless systems. And I cannot blame them, intermediaries, wether it be Google or Facebook, standing in between the user and the content creators, or banks, standing between creditors and debitors, haven't got the best reputation. A lot has gone wrong overthere, but I'm afraid DLT will bring entirely new problems to the table that we cannot forsee at this point. So I'm curious what the future will bring.
- pepesza 8y ago
  Offtopic. DLT is basically a marketing fraud. Both "permissioned" and "permissionless" DLT are packed into one term. Problem is - they are vastly different. Permissioned DLT is not interesting at all from technical and social point of view. We knew how to do them for a long long time. Openness, permissionless is the key innovation introduced with Bitcoin. "Permissioned blockchains" are just repackaged tech with lot of marketing sauce.
  - Egidius 8y ago
    I use the term DLT as, for example, DAGs are not exactly blockchains if I'm not mistaken. They have no blocks, just transactions. It had nothing to do with permission.
    - EthanHeilman 8y ago
      Not all DAGs have blocks however many DAGs do have blocks. BlockDAGs are usually considered blockchains.
      - Egidius 8y ago
        I wasn't aware of that. Thanks for that insight!!
  - mkirklions 8y ago
cslarson 8y ago
Ethereum is novel and ambitious technology. In my mind, and I think in the minds of most in the community, scaling had always been the big, important challenge on the horizon - but it builds confidence to see just how far the devs (and community) have brought it till now. It's like everything is in place for that final push over the mountain.
- baby 8y ago
  Another problem is the size of the blockchain
  - pas 8y ago
    https://github.com/ethereum/wiki/wiki/Sharding-roadmap https://github.com/ethereum/wiki/wiki/Sharding-roadmap
- mkirklions 8y ago
  I think people are overtly positive about ETH because they missed out on BTC and believed in ETH. But dApps are unused. Customers are not paying 7x the price for decentralized verification. Centralized servers can be tampered with, but no one cares about your 4th of july pictures or saved games. Unless currency, voting, or a timestamp is involved, you dont use expensive verification methods. The sheer cost of blockchain makes it insane to imagine apps being built on it. Currency has typically required firepower to move money and different ways to verify money is real money. This is a real case for using blockchain.
  - rhyzom 8y ago
    plenty of what you say is just not right (or plain wrong). btc = monetary/value transactions. i read the paper as early and 2008 and didn't miss out on it, i just didn't care then same as i don't care now. eth = dapps. just because dapps didn't take off immediately doesn't mean they aren't beginning to get used, increasingly so, and that there isn't a lot of valuable, useful, interesting and innovative applications being built and experimented with. btc and eth are two different things. if how much something can possibly enrich you is all you care about, that doesn't mean that everybody else is necessarily on that page. that being said, it's true - ethereum is basically useless for a lot of otherwise really interesting possible use cases and applications (which is why we got things like IOTA, something that i'm sure everybody here has nothing but hate towards, for whatever reasons). then ethereum classic, too, is something just as different. and i couldn't help but notice how most innovation is being done by people of russian or eastern european origin, or israelis, while the vast majority of silicon valley only chase the dollar bill and base their entire business models on how they can possibly screw the user. much as i sometimes find valuable threads here on hackernews, i don't really take much of what you people say seriously.
AltVanilla 8y ago
I think for a cryptocoin to be useful in the real world, scaling and quick transaction times has to be a priority from the start of the design. Its not something you can fix, as an afterthought. That's why I like xrp. Low transaction times, and it scales well. It is permissioned in the sense that every validator-node needs to keep a list of all the other validator nodes it trusts. Which means validators without a reputatable company behind them, might find it hard to become trusted by anyone else. Thats a price worth paying for scalability, in my opinion.
- leppr 8y ago
  What a Crypto Currency needs too is price stability, and that isn't tackled by XRP, nor can it be added to it since it has no smart contract capability. So you sacrifice decentralization for only one of the necessary characteristics of money. Not worth it, PayPal works fine for now.
- mihaifm 8y ago
  It's easy to scale when your transactions are basically only tcp/ip messages between trusted nodes. There has to be a bettwer way...I'm personally in favor of dPoS (delegated proof of stake) networks, that can have high throughput without sacrificing decentralization, with consensus achieved by representatives that can be changed by everyone in the network. Still, xrp might have applications in the banking world, time will tell.
  - cdiddy2 8y ago
    I am concerned that dPOS will lead to cartel like organizations in the long run as the rich get richer and can vote for themselves
    - rhyzom 8y ago
      AFAIK, all dPoS systems have so far proven to fairly quickly spiral into what you just described. Waves uses a leased PoS that might be worth exploring in more detail (or I've been meaning to anyway).
- 8y ago
foobarbecue 8y ago
Man the TC website sure is terrible to use on mobile. Everything keeps flying around, there's a white banner in the way of the video with an X that appears to behave as a back button when you click it??
jmp_0xdeadbeef 8y ago
Shouldn't an engineer be able to predict how a system will scale ahead of time? asking for a friend
nsxwolf 8y ago
Is scaling the real reason there are no useful applications?
- guiomie 8y ago
  Fungible and Unfungible ERC token are used by a lot of people.
asgoniio 8y ago
An important thing to remember is how jaw-droppingly terrible Ethereum is. The cost to perform one million adds is about $30. Storage costs $5 million / GB. That's on the order of 10^9 times worse than a normal system. If these people managed to make Ethereum three orders of magnitude more efficient without any compromise (which will never happen), it would still be so inefficient that I would struggle to think of a use case. Ethereum isn't merely useless. It's so terrible that it transcends the human brain's sense of scale. If I were a malicious person spreading misinformation I wouldn't have the audacity to make up numbers as bad as the reality. [1] https://hackernoon.com/ether-purchase-power-df40a38c5a2f https://hackernoon.com/ether-purchase-power-df40a38c5a2f

103 points by dsr12 8y ago

Off topic but I am so entirely sick of amp/mobile pages like this that just fail so hard at delivering the advertised content. Multiple reloads, pop-up conditionals, auto-scrolling...yuck. Clicking the link briefly flashes the title of the topic followed by a blank square with a “close” X then flashes a couple of times and I land on a listing of videos none of which are the op topic. I’m not using ie4 or something - this is the current version of Safari on the latest mobile device from one of the biggest phone manufacturers that exist! All on a “tech” website. First world problems I know but goddamn is it frustrating. Sorry for rant.

dvh 8y ago
Same on Android Edit: it works now but it is badly broken video player, stopped when I tried to seek forward
plasma 8y ago
I hate AMP as well. They’re less maintained, feature less pages (see reddit), and they have obnoxious script that prevents hold-tap from allowing “open in new window” from working.
niksakl 8y ago
here is a youtube link to the video https://www.youtube.com/watch?v=976Q2cu1_DE https://www.youtube.com/watch?v=976Q2cu1_DE

SwingingShips 8y ago

Ethereum is talking about 2020 as its target for scaling upgrades. There are other projects that haven't launched yet that probably have an advantage over them, as so long as it can integrate with Ethereum rather than claim it can totally supplant it. There are projects like Cardano and Zilliqa, but then there are some like Orbs that are using blockchain virtualization claiming it can work in tandem with Ethereum. I assume Ethereum will implement some of its solutions before that 2020 date, but they will probably need to rely on more partnerships to do it.

haasted 8y ago
https://cosmos.network/ https://cosmos.network/ will also have an Ethereum bridge, as well as proof-of-stake.
StavrosK 8y ago
Cardano has stalled for so long that I stopped waiting. I hear great things about EOS from people I trust, though, I need to look into that.
- JohnJamesRambo 8y ago
  EOS is like a really expensive AWS. There is nothing revolutionary or decentralized about that. I don’t know what you mean about Cardano, they have a very detailed roadmap and haven’t missed any major milestones I am aware of. https://cardanoroadmap.com https://cardanoroadmap.com
- shinryuu 8y ago
  The launch of EOS has not been very well received at all. Lots of mistakes and the governance is a mess.
  - mihaifm 8y ago
    Indeed, EOS relies on a constitution with human interpretable rules [1]. Not sure how that fits into the crypto sphere, where governance should be done be algorithms and code. [1] https://github.com/EOSIO/eos/blob/5068823fbc8a8f7d29733309c0496438c339f7dc/constitution.md https://github.com/EOSIO/eos/blob/5068823fbc8a8f7d29733309c0...
    - rhyzom 8y ago
      and it's also supposed to be hashed in each "legitimate" transaction. while network resources are so scarce as to stir rampant speculation. idk - the only thing about EOS that i have noticed is how overtly it screams of greed. the people that jumped in on that bandwagon did so for no other reason but hoping it would make them rich (cos Daniel Larimer said "you can run Ethereum on a single EOS smart contract", lol).

Egidius 8y ago

Great to see these young guys attempting to tackle these huge problems out of their believe that they can make "the world a better place". Not sure DLT will live up to its promise, but these guys truly believe the world is better off with trustless systems. And I cannot blame them, intermediaries, wether it be Google or Facebook, standing in between the user and the content creators, or banks, standing between creditors and debitors, haven't got the best reputation. A lot has gone wrong overthere, but I'm afraid DLT will bring entirely new problems to the table that we cannot forsee at this point. So I'm curious what the future will bring.

pepesza 8y ago
Offtopic. DLT is basically a marketing fraud. Both "permissioned" and "permissionless" DLT are packed into one term. Problem is - they are vastly different. Permissioned DLT is not interesting at all from technical and social point of view. We knew how to do them for a long long time. Openness, permissionless is the key innovation introduced with Bitcoin. "Permissioned blockchains" are just repackaged tech with lot of marketing sauce.
- Egidius 8y ago
  I use the term DLT as, for example, DAGs are not exactly blockchains if I'm not mistaken. They have no blocks, just transactions. It had nothing to do with permission.
  - EthanHeilman 8y ago
    Not all DAGs have blocks however many DAGs do have blocks. BlockDAGs are usually considered blockchains.
    - Egidius 8y ago
      I wasn't aware of that. Thanks for that insight!!
- mkirklions 8y ago

cslarson 8y ago

Ethereum is novel and ambitious technology. In my mind, and I think in the minds of most in the community, scaling had always been the big, important challenge on the horizon - but it builds confidence to see just how far the devs (and community) have brought it till now. It's like everything is in place for that final push over the mountain.

baby 8y ago
Another problem is the size of the blockchain
- pas 8y ago
  https://github.com/ethereum/wiki/wiki/Sharding-roadmap https://github.com/ethereum/wiki/wiki/Sharding-roadmap
mkirklions 8y ago
I think people are overtly positive about ETH because they missed out on BTC and believed in ETH. But dApps are unused. Customers are not paying 7x the price for decentralized verification. Centralized servers can be tampered with, but no one cares about your 4th of july pictures or saved games. Unless currency, voting, or a timestamp is involved, you dont use expensive verification methods. The sheer cost of blockchain makes it insane to imagine apps being built on it. Currency has typically required firepower to move money and different ways to verify money is real money. This is a real case for using blockchain.
- rhyzom 8y ago
  plenty of what you say is just not right (or plain wrong). btc = monetary/value transactions. i read the paper as early and 2008 and didn't miss out on it, i just didn't care then same as i don't care now. eth = dapps. just because dapps didn't take off immediately doesn't mean they aren't beginning to get used, increasingly so, and that there isn't a lot of valuable, useful, interesting and innovative applications being built and experimented with. btc and eth are two different things. if how much something can possibly enrich you is all you care about, that doesn't mean that everybody else is necessarily on that page. that being said, it's true - ethereum is basically useless for a lot of otherwise really interesting possible use cases and applications (which is why we got things like IOTA, something that i'm sure everybody here has nothing but hate towards, for whatever reasons). then ethereum classic, too, is something just as different. and i couldn't help but notice how most innovation is being done by people of russian or eastern european origin, or israelis, while the vast majority of silicon valley only chase the dollar bill and base their entire business models on how they can possibly screw the user. much as i sometimes find valuable threads here on hackernews, i don't really take much of what you people say seriously.

AltVanilla 8y ago

I think for a cryptocoin to be useful in the real world, scaling and quick transaction times has to be a priority from the start of the design. Its not something you can fix, as an afterthought. That's why I like xrp. Low transaction times, and it scales well. It is permissioned in the sense that every validator-node needs to keep a list of all the other validator nodes it trusts. Which means validators without a reputatable company behind them, might find it hard to become trusted by anyone else. Thats a price worth paying for scalability, in my opinion.

leppr 8y ago
What a Crypto Currency needs too is price stability, and that isn't tackled by XRP, nor can it be added to it since it has no smart contract capability. So you sacrifice decentralization for only one of the necessary characteristics of money. Not worth it, PayPal works fine for now.
mihaifm 8y ago
It's easy to scale when your transactions are basically only tcp/ip messages between trusted nodes. There has to be a bettwer way...I'm personally in favor of dPoS (delegated proof of stake) networks, that can have high throughput without sacrificing decentralization, with consensus achieved by representatives that can be changed by everyone in the network. Still, xrp might have applications in the banking world, time will tell.
- cdiddy2 8y ago
  I am concerned that dPOS will lead to cartel like organizations in the long run as the rich get richer and can vote for themselves
  - rhyzom 8y ago
    AFAIK, all dPoS systems have so far proven to fairly quickly spiral into what you just described. Waves uses a leased PoS that might be worth exploring in more detail (or I've been meaning to anyway).
8y ago

foobarbecue 8y ago

Man the TC website sure is terrible to use on mobile. Everything keeps flying around, there's a white banner in the way of the video with an X that appears to behave as a back button when you click it??

jmp_0xdeadbeef 8y ago

Shouldn't an engineer be able to predict how a system will scale ahead of time? asking for a friend

nsxwolf 8y ago

Is scaling the real reason there are no useful applications?

guiomie 8y ago
Fungible and Unfungible ERC token are used by a lot of people.

asgoniio 8y ago

An important thing to remember is how jaw-droppingly terrible Ethereum is. The cost to perform one million adds is about $30. Storage costs $5 million / GB. That's on the order of 10^9 times worse than a normal system. If these people managed to make Ethereum three orders of magnitude more efficient without any compromise (which will never happen), it would still be so inefficient that I would struggle to think of a use case. Ethereum isn't merely useless. It's so terrible that it transcends the human brain's sense of scale. If I were a malicious person spreading misinformation I wouldn't have the audacity to make up numbers as bad as the reality. [1] https://hackernoon.com/ether-purchase-power-df40a38c5a2f https://hackernoon.com/ether-purchase-power-df40a38c5a2f