xStelea

"Wallets on Sui are terrible, even after 15 months of mainnet" ...and the reason is deeper than just rushed-out apps. @SuiNetwork, like every other blockchain platform, has taken an largely un-opinionated view on how tokens and NFTs should be implemented on Sui. Where should the asset logic live? In the token? In some connected smart contract? Who knows - dealers choice. This is great from the POV that it encourages experimentation, but is quickly coming up on the hard problem of co-ordination around standards. Because there aren't any. This is like when electricity was invented and there were no standardised plugs and plug sockets. Great, you can use this incredible new invention - not so great, no one can plug the damn things in. DeFi runs on standards. These standards are important because they allow everything in the stack - from the user transaction, to the wallet interfaces, to inter-smart-contract interactions, to work together. DeFi works because the "objects" tokens/NFTs/Interfaces are standardised. ERC20, ERC721 etc - but generally poorly when done at the smart contract level. An untold number of hacks have happened just because of the insane "infinite approval" function for ERC20s. However, these standards only happen through a lot of really hard, and pretty tedious work. If it doesn't, one token is implemented in one way, and another in a completely different way - and you get chaos - chaos in wallets, dApps, interfaces and everything else from having to adapt, adopt and adjust to all the competing standards. And when a new one needs to be built - the process for every single dApp and wallet that wants to use it, has to start all over again. Ethereum took years to get to a "working" EIP process, and holy crap is it painful. Sui, by taking an entirely new approach to programming smart contracts, and not being opininated, is now having to work through this pain. The thing is, with some more thought into the architecture of the system, it didn't need to be like this. For Radix, we took an opinionated view on how NFTs, Tokens, Abstracted accounts, auth systems etc were implemented and enshrined them into the layer 1 itself. This is the Radix Engine: radixdlt.com/blog/radix-engi… By doing all this work upfront on defining standards that are enshrined into the network itself, there is no arguments on standards. What is more if a new one needs to be added, it is done at the level of the network - and suddenly everyone is compatible with it - no re-work required - because dApps, wallets etc are implementing the standardised interfaces into the Radix engine, and rely on the enshrined models. This is like inventing electricity, and then also inventing a type of plug that everyone universally uses, but with the magic property of when a new upgrade needs to be pushed, it gets pushed to all of the plugs, simultaneously. No ripping out required. What is the upshot? Every token type that is created on Radix, in whatever permutation...displays immediately in wallets, has complete human readability in transactions, is instantly compatible with custody/exchanges and anyone else that has already integrated Radix and just...works. Was it easy? No, it probably added around 1.5 years to the development time of Radix, and required the input of hundreds of pieces of developer feedback to get right. But was it worth the extra time? Absolutely. Because, while it was boring, it was 100% necessary to make sure that developers could just get on with the important job of making awesome dApps, and not continually have to battle with competing or badly implemented standards from hundreds of people all rushing to get to market. To go deep on this, check out: radixdlt.com/blog/radix-engi…