Thread by @alonsodegortari

1: storing data on a blockchain has lots of funky economic implications. who pays for it? how? when? for how long?

here's some thoughts on this thorny 🌵 challenge and how we deal with storage #tokenomics @Mysten_Labs, introducing the #Sui storage fund 💽🫙🥤

2: like other databases, blockchains are useful because you can write stuff into them and read later. ✍️📖

the challenge is what happens in between? someone has to store the data from the moment it's written to the moment it's read. this has a financial cost. 🤔

3: why is figuring out the tokenomics of storage a new problem? 💵💴💶💷

well, because back in the "old days of blockchain", it was so expensive to write anything on-chain that nobody cared too much about the economics of storage.

4: writing into gold 🥇 vs clay 🧱 tablets is a useful analogy.

it's so expensive to write on gold, that nobody will care about the costs of storing those tablets. but once you have clay, everyone can write tablets and storage quickly becomes a financial burden.

5: until very recently, blockchain space was so precious, that storage was an afterthought. on Ethereum, writing 1MB of data can easily cost you tens to hundreds of thousands of dollars! 😳

but now, blockchain space has evolved from being a luxury good to a commodity. 👜➡️🌽

6: what happens if the tokenomics of storage are left unaddressed? 👽

blockchain storage is a market failure. validators operate the network and need viable business models. so if validators need to store lots of data to operate, they will pass these costs on to the end user💲🙄

7: but remember that the whole issue of storage is one of timing. if i write data into storage and don't pay for it, then validators will pass on those future storage costs to future users. ⏲️⌛️

8: this is what economists call an externality. users who write data into storage today don't internalize its cost, while future users are forced to pay for past storage that they don't own.

when blockchains become like clay tablets, these externalities become BIG 🦣🐳🗻

9: externalities occur everywhere, blockchain storage is kind of similar to climate change🌡️

current generations (us!) don't internalize that the planet will need to store all our carbon emissions. future generations will face the costs of living with them... that's not fair🌍

10: how to solve the tokenomics of storage?

the first step is easy, make sure that if you use storage you pay for it. just like you pay for dropbox or other cloud storage. this shifts the cost of storage to the user who writes data and eliminates the externality. ✅🎉

11: in theory everything makes sense. but the tricky details are always in implementation. 👩‍💻

since storage is all about timing, the key tokenomics challenge is: if you're going to charge for storage, for how long? ⏳🧵

12: there's two broad ways to deal with storage: for a FINITE or INFINITE amount of time.

FINITE means you pay storage upfront for a fixed period of time and then you either pay more fees or the data is deleted. ⏱️

INFINITE means you pay for storage once and forever. ♾

13: the INFINITE approach is very hard to design because INFINITE time is a veeeery long time.

how could you possibly ever charge enough fees to cover storage costs foreeeever?? 🫤😳🧐

14: introducing #Sui storage fund:

in Sui, users pay for storage upfront and these fees are deposited in a fund. the fund's size is taken into account when distributing stake rewards and validators receive relatively more rewards than SUI delegators when the fund is larger🧧🧧

15: users who store data today pay for it. validators who store the data tomorrow are rewarded for it.
this solves the storage externality. future Sui users are happy, because they need not pay for the storage costs of past users. 🍾🎀🪅🎉

16: a simple interpretation: it is as if validators borrowed SUI tokens from the storage fund at zero interest, while borrowing SUI tokens from delegators at a positive rate. hence, validators earn relatively more rewards when the storage fund is large. 📈💹

17: the storage fund design has one key property: it never pays out tokens directly out of its capital!

instead, by indirectly distributing tokens from stake rewards, it can fund storage costs in perpetuity. #Sui's storage fund deals with the storage externality forever. 🎉🪅🎀

18: we designed #Sui's tokenomics to address storage costs forever because we thought it would deliver much better user experience than finite rent models. doing the mental accounting of preparing liquidity and paying for storage on an ongoing basis is very taxing. #relief 😮‍💨

19: the cherry 🍒 on top is that we've also designed an economic mechanism to incentivize people to delete data: rebates! 🆒😎

once on-chain data is no longer useful, users can tell Sui to delete it and they get a partial rebate of their fees in return. 🙃

20: new storage writes and rebates of old transactions imply the storage fund's size moves in sync with the amount of data held in on-chain storage. thus ensuring validators are compensated in proportion to how much storage they hold! 🍀🦸‍♀️

21: in sum, #Sui's storage fund is at the frontier of blockchain storage tokenomics and optimized for user experience. it eliminates the externality of mis-pricing storage costs and creates incentives to delete stale data. 💧💧💧

You can read more here: docs.sui.io/learn/tokenomi…