Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

An engineer's observations on Web3 and its possibilities

psl.com

Post reply on HN
Print view View original

101–110 of 319 posts

Re: An engineer's observations on Web3 and its possibilities

#101

Post by xmly »

necubi wrote: post #92
practice9 wrote: post #82

Earlier quoted context omitted.

Blockchain isn't even central in Web3, but it surely helps. Web3 = decentralized identity, full data and connection graph ownership by user, reputation transfer, verification via zero knowledge proofs, support for VR/AR and 3D objects. All of this stuff is either in prototype or draft phase, but it's ok. These things need to be built to stop horror stories on HN about "algorithm justice" and "data kingdoms”: - Google…

And that's an improvement how? At least with Google there's a chance you'll be able to get your account back. If it's "decentralized" (which in crypto land is apparently synonymous with "run by a cabal of miners in China") and you lose your private key, you're completely out of luck.

When will Trump get his chance?

Re: An engineer's observations on Web3 and its possibilities

#102

Post by xmly »

PretzelPirate wrote: post #95
wmf wrote: post #91

Earlier quoted context omitted.

1,000 validators is "centralized"? Don't look at how many Bitcoin pools there are.

1000 validators is very centralized. All you have to do is bribe 2/3rds of them control the chain. This is on top of Solana not having slashing in place to punish bad behavior, and the Solana Foundation hosting validators for people who want to delegate. Solana’s 1000 validators may not even be owned by 1000 different entities, which reduces the attack difficulty. Solana scales through centralization, not new ideas i…

Solana and affiliated organizations probably owns 999 validators, that is my guess

Re: An engineer's observations on Web3 and its possibilities

#103

Post by madrox »

Today, I hopped into a Twitter Space called "web3 for gaming" and from the moment I joined until the moment I left five minutes later was talking about how crypto payments need to be better on mobile.

This feels like when web2 was coalescing as an idea. A lot of people were trying to shoehorn their pet technologies into the discussion. XML was holding on for dear life. It's ridiculous and completely divorced from the topic at hand. My only explanation for the behavior is that it's difficult to see the obvious absurdity when you're deeply invested in a particular outcome.

Re: An engineer's observations on Web3 and its possibilities

#104

Post by KallDrexx »

practice9 wrote: post #82
WastingMyTime89 wrote: post #73

Not going to lie, I thought and still think that the whole Web 2.0 thing was clever marketing and window dressing on top of fairly pedestrian but useful technologies. It was actually less of seismic shock than mobile first which despite being a bigger shift didn't actually get a number. As far as I'm concerned, Web3 is beyond this. It's what happen when you let Ponzi schemers write the marketing material. I can't for…

Blockchain isn't even central in Web3, but it surely helps. Web3 = decentralized identity, full data and connection graph ownership by user, reputation transfer, verification via zero knowledge proofs, support for VR/AR and 3D objects. All of this stuff is either in prototype or draft phase, but it's ok. These things need to be built to stop horror stories on HN about "algorithm justice" and "data kingdoms”: - Google…

Can you explain how web3 solves any of your bullets?

If you are trying to say that web3 will prevent platform holders from banning you from their platform, then good luck with that. Any platform can and will ban you and have complete control over their platform. Thinking web3 is going to prevent that is naive for many many reasons. If Facebook bans me then web3 isn't going to allow me to post on Facebook still. Same with Twitter and Linked In.

If you are saying that web3 solves these by making it so that these platforms aren't in charge of my identity, well they aren't today so I'm not sure what the deal is. If gmail bans my account it's not a huge issue because I use a custom domain for my email, so I can route it to another provider or host my own. If Facebook bans me then I lose contact with family on there but I don't lost my identity because FB isn't in charge of my identity. same with twitter or linked in or whatever.

All these problems you list are solvable today with the decentralized web we've had for decades. There's a reason centralization has happened on these platforms, and it seems like all the web3 talk I see is naive to that.

Edit: Also Keybase is a great example using cryptography to solve the identity problem without a blockchain and with some level of decentralization (keybase's servers are centralized but the actual identity is all based on PGP and GPG that's been around for a long time).

Re: An engineer's observations on Web3 and its possibilities

#105

Post by kall »

evergrande wrote: post #75

Quite possibly the difference between people who are hostile vs favorably disposed to Web3 is whether or not they think traditional orgs work fine or need improvement. Frustration with traditional organizational pathologies is not universal. The right comparison may not be with past tech revolutions. It may be to organizational evolutions like monarchy to democracy, women entering the workplace, the invention of the…

If the tyrannical crowds on Patreon are a problem skin in the game is going to 100x that.

"You have a fiduciary responsibility to every single one of your fans“ sounds like a black mirror episode.

Re: An engineer's observations on Web3 and its possibilities

#106

Post by paul_f »

strogonoff wrote: post #90
practice9 wrote: post #82

Earlier quoted context omitted.

Blockchain isn't even central in Web3, but it surely helps. Web3 = decentralized identity, full data and connection graph ownership by user, reputation transfer, verification via zero knowledge proofs, support for VR/AR and 3D objects. All of this stuff is either in prototype or draft phase, but it's ok. These things need to be built to stop horror stories on HN about "algorithm justice" and "data kingdoms”: - Google…

Lack of human support in the old Web 2.0 is bad, but going full zero-trust in response seems more like throwing a fit instead of improving things (e.g., banning the exploitative ‘free’ ad-supported business models). A world in which you lose some magic alphanumeric combination (or a thief pinches your hardware key, etc.) and you are literally locked out of your identity (together with access to your eternal blockchai…

It does not even require hacker, a thief could show up at your house and demand your keys.

Re: An engineer's observations on Web3 and its possibilities

#107

Post by k__ »

kall wrote: post #99
mattlondon wrote: post #71

So .. TL;Dr web3 is nothing to do with the web , and is just the same old smart contract crypto hype we've all seen before? No mention of the actual distributed web protocols like IPFS or Dat as far as I could tell. Have I massively missed something or has the crypto community just co-opted the name web3 for their own non-web related stuff? How does the web factor in to crypto?

I think the web part is now "log in with your wallet, fetch account state from the blockchain". It all happens on the normal web and the blockchain access is facilitated by "normal" web API gateways like Infura or Cloudflare. This is my view from the outside so it might be off on the details. It seems pretty reasonable, as far as crypto stuff goes, except that there is a massive, expensive machine behind all the HTTP…

Also, many wallets come with a browser extension, to blur the lines between web and crypto even more.

Re: An engineer's observations on Web3 and its possibilities

#108

Post by latchkey »

thebean11 wrote: post #58
v_london wrote: post #16

Thanks for the great article! I wished the section on DeFi addressed the biggest elephant in the room, overcollateralization. A DeFi borrower always needs to lock in more assets as collateral than what they are borrowing, thus defeating the whole purpose for taking a loan aside from financial speculation. And there's no real way to solve it: all mechanisms I've heard of either try to replicate some form of background…

Aren’t mortgages actually also examples of over collateralized loans? You borrow 80% of the homes value, and use the entire home as collateral.

Liquidation is quite different.

Re: An engineer's observations on Web3 and its possibilities

#109

Post by latchkey »

wmf wrote: post #91
PretzelPirate wrote: post #84

One interesting misrepresentation is this: “ Several innovations in distributed consensus design make Solana’s performance possible.” Solana’s consensus protocol isn’t responsible for its performance, it’s the fact that they mandate higher performing machine SKUs and at least 500Mbs (recommending 1Gbs) internet speeds. They also had to largely centralizing and have around 1000 validators. Ethereum’s fees and performa…

1,000 validators is "centralized"? Don't look at how many Bitcoin pools there are.

a) Pools don't do validation, bitcoin nodes do validation.

b) It is easy for miners to switch to another pool in a matter of minutes. What do you do when you want a whole different set of Solana validators?

Re: An engineer's observations on Web3 and its possibilities

#110

Post by latchkey »

xmly wrote: post #102
PretzelPirate wrote: post #95

Earlier quoted context omitted.

1000 validators is very centralized. All you have to do is bribe 2/3rds of them control the chain. This is on top of Solana not having slashing in place to punish bad behavior, and the Solana Foundation hosting validators for people who want to delegate. Solana’s 1000 validators may not even be owned by 1000 different entities, which reduces the attack difficulty. Solana scales through centralization, not new ideas i…

Solana and affiliated organizations probably owns 999 validators, that is my guess

Solana and... EOS and... XRP...
Post reply on HN