FLIP 66 - Revisiting Flow storage minimum account balance

Governance FLIPs
#21

Thanks for your reply and I totally understand that we need to figure out how to make the network as a whole more sustainable on the long run and I am sure it will benefit everyone on the long run and if we look at the entire ecosystem health.
My problem is with the execution and what will happen with all the accounts that don’t have enough storage. Will they just be blocked until they receive the necessary funds from someone else? Will Dapper increase the refill amount accordingly as well to avoid incurring in such problems (I know already of a few people that can’t transfer Flovatars sometimes because they are too big and the refill bot is not topping it up enough). My fear is that if these types of things are not ironed out in every detail, it will be very easy to create lot of chaos and ultimately damage the image of the blockchain. So the question might be: how do we really ensure the transition will happen smoothly for everyone?

Thanks!

2 Likes
#23

Good and valid points @stwh i was just illustrating the effects the change will bring.

#24

Apologies for the newbie question, but I don’t understand the token hold for storage model in the first place…

Do I understand correctly that the true cost of storage is paid for by node operators? But this amount of held FLOW is decoupled from their reward rate (see unanswered questions in FLIP-1)? How does the held FLOW pay for anything?

I can see how the held FLOW is a security measure, but not how it’s a business model.

[EDIT: I am excited to see discussions like this taking place. I was discouraged when FLIP-1 was a community echo chamber and I hope this is a sign of better things to come.]

#25

The default storage of 10k for s new account would cause even more severe storage issues for dapper wallet. A doodles wearables is like 4k ish. You cannot even do a dooplication without needing more storage.

For me the fact that dapper will sponsor this for dapper wallet users makes this change even harder to swallow. How are non-dapper-wallet ecosystem going to compete with that?

Versus nfts store entire binary images in the nft. They are megabytes.

2 Likes
#26

Also if Dapper sponsors it with an automated top-off, how does it force developers to be more selective with their storage use?

#27

so you see that 2165 is which has 40% of accounts, is something like this: f.dnz.dev : a Flow View Source fork ?

This account has just one FT vault other than standard setup, which is RLYVault. Has no NFT, nothing.

you have 40% accounts like this, those are what @bjartek means ‘vacant’, If you generate one USDC vault, one FUSD vault, you will be probably at 70% percentile. Without a single NFT.

#28

You have no idea what you are talking about sorry, average is not important here, I get user’s from US with even 100 USD per user, but my median CAC is maybe even less than 10 cent.

There is no chance a wallet can come to flow, acquire users with marketing + pay 10 cent to wallet + pay tx fees and be profitable, that’s why we are losing blocto etc.

1 Like
#29

One more thing @kshitij.chaudhary, NEAR and SOL examples, does storage cost is charged to the user or to the dapp (contract) owner?

if dapp (contract) owner is paying, we are making a apples to oranges comparison, don’t you agree?

then maybe we can make some model like:

  • when you mint ( create an object ), you have to burn ( or lock ) some FLOW, and we don’t charge storage fee to the user. They can include this fee to the price of NFT ( or whatever resource created )

Burning ( indefinite staking, or covering inflation with those fees ) can be good mechanism for controlling inflation and Flow token price.

As we burn and Flow price increase, cost for node operators will be less even data is more.

Cause in this FLIP, there is no plus side for inflation. according to your calculation ( and proposal ) developers will hold less data on chain, Flow locked will be minimal versus 1.3 million Flows minted every week.

67.6 million Flow per year, with 0.1 Flow requires 676 million new accounts per year to offset inflation.

2 Likes
#30

I think that adding the necessary FLOW into the NFT itself would definitely be the best solution, but ideally it should happen almost automatically out of the transaction fees without the user and the developers having to worry about it at all.

Right now anyone can produce tons of garbage NFTs that use a lot of storage and spam people by sending it to them and having them to cover for the cost of it (unless they burn it or pass it on to someone else). This doesn’t really prevent for a better storage usage because it’s not the creator of the garbage that is ultimately paying for it.
If we’d make it instead that the storage fee is paid when the new NFT resource is created, then we would prevent people from producing low quality content, and the users wouldn’t have to worry about how big an NFT is before buying it.
For example I have 50.5 FLOW in my wallet and the NFT costs 50 FLOW. Can I actually buy it or not? will that 0.5 FLOW be sufficient for the storage?
I think this type of uncertainty is not good from a user perspective and having to explain each single user about the storage system is not a viable solution if we’re thinking about the billions of users out there.

But again, not sure how complicated it would be to implement of course, so just trying to share a practical point of view from a project like Flovatar that uses storage relatively heavily (50-100kb each NFT because of the SVG stored on-chain).

2 Likes
#31

agreed putting into transaction fees is the best solution, it also prevents the inflation.

if I am a dapp (like flovatar) I should have option to make something innovative (like flovatar) and store data on chain. Paying this in advance solves a lot of problems.

#32

As stated by @luca and @bluesign adding the needed FLOW into the NFT might be the best solution for custodial and non-custodial wallets. there will be a need for some education for the users, but I think it is unavoidable. At some point, the strategy of sponsoring the storage costs and transaction costs by the wallets themselves (Dapper, Blocto, …) is not sustainable in the long run. It is a good strategy to bootstrap the network in the early days, but if you want a healthy ecosystem, all the parties (users, wallets, dApps, node operators, etc) need to contribute in a balanced way

1 Like
#33

Totally agree with this proposal

@bluesign “you don’t predict Flow price going to 10$ ranges anytime soon ?”
No one can predict the price. If the price rises to $10, another proposal adjustment can be initiated at that time.

@luca "it would definitely be a big deterrent for mass adoption … "
web2 users also need to pay for using Email Cloud (when the storage content exceeds the free plan)

@austin How does the flow team plan to address the possibility of attacking someone’s account if it becomes so much more expensive?
This is a funny question, do you think Satoshi is responsible for anyone bitcoin wallet attack?

#34

But here is the main selling point is preventing attacks on chain. web2 users don’t pay for storage, they pay for a company to pay for their storage.

web2 users also need to pay for using Email Cloud (when the storage content exceeds the free plan)

Imagine your email provider is working on AWS ( or your dropbox ) , you pay them, they pay for storage. It is not like you getting an invoice from AWS every-month for your usage.

As a Flow user, if I am using TopShot, I am paying to TopShot, not to Flow Blockchain. TopShot should pay the storage cost on my behalf.

No one can predict the price. If the price rises to $10, another proposal adjustment can be initiated at that time.

yeah this can be very unreliable though, flow price fluctuated 50x-100x in the past. With this kind of big numbers, even 4x-5x change can make a wallet unprofitable. It is not feasible to change this frequently.

#35

This is a good point, which lead to something I was thinking about for a while: Flow is claiming itself as the blockchain with the best user experience (UX). When talking about prices in general, everybody at the end refers to the equivalent in Fiat (USD). So why don’t fix a price for storage in USD (show USD to the end user), make it paid in FLOW, and bring the complexity back to the development level?
This will provide a clear/non fluctuating pricing and a better UX

By the way same could be done for transaction fees

#36

Hey, seems like you could be referring to cost-per-install (top-of-the-funnel customers) while I’m referring to CAC for paying users which is what businesses usually measure against CLTV. You’re right though, installation is what we should consider here as that would cause account creation and require 0.1+ FLOW to be loaded. For my understanding do you know how the ratios of storage costs to CPI would look like for games building on other chains, because if CPI is really 10c the problem would be bigger in sol, near, etc.

#37

Just want to chime in a little bit here. I’m not an economist, so I’m not very familiar with all the economics, but the increase does seem pretty high to me. I would be in support of increasing the price, but maybe reducing the proposal by an order of magnitude or something. @Redallica’s suggestion about locking the pricing to USD would be ideal, but that would require us to have a price oracle that everyone would trust, which we don’t have yet unfortunately. Definitely something to consider for the future though.

Also, we’re here to discuss the proposal and help support each other, so lets please be respectful in our comments. thank you! :slight_smile:

1 Like
#38

I don’t know about sol and near, (correct me if I am wrong) but from what I read they are more like charging the dapp than individual user for their setup. Like we are talking about 4k (doodle) - 1MB (flovatar) NFTs on flow, I don’t see how we can expect user to manage funds.

It feels like dapper wallet will manage for all dw users, rest will be not able to compete, which is disturbing.

#39

Yeah sorry I am a bit triggered on these stuff lately, just I want good for flow, and seeing everybody is pulling to some different direction is little disappointing.

One day I am talking with people (flow people), server-less dapps on flow etc, next day this flip. But ofc I apologize to @kshitij.chaudhary , I am sure he is also trying to find best for flow.

#40

Again for me keep sponsoring these fees is not healty for the ecosystem in the long run. Now if the decision is to keep sponsoring the fees and Dapper Wallet is the only one being able to do it, it’s not going to be healthy for the growth of the Flow ecosystem. Imagine the only wallet thriving in Flow blockchain under this system is the Dapper Wallet. The best argument for a centralized chain !

#41

Is it worth diving into who should actually pay these storage fees? I’m guessing it would be quite complex to change, but ideally there is a way to decouple who has the resource itself, and who created it (and thus caused more storage space to be taken on-chain)