Thread by @mallowsxyz

Decentralized apps are a virtuous talking point, but largely a myth as they are designed today.

The chart below shows the current state of ethereum "dApps" and "defi", and areas where we excel and others where points of failure exist.

Let's dive in! 🧵

We excel, particularly in ethereum, with truly decentralized smart contracts that are censorship resistant, and with provable trust in what you're running through being able to see verified open source code for those smart contracts.

👇

We do pretty well in that many front-ends for these dApps are published immutably to IPFS. And we do well in that what is published is open source code viewable through git/github.

👇

Downside is that

1) what source code is published to IPFS is out of your control - meaning able to be influenced by governments (cough Tornado Cash), and

2) reliance on DNS provides a middle man attack vector.

👇

This leads to a dApp experience where you don't know what you're running, it could be censored at any time, it's actually mutable, and you aren't sure if you can trust it.

👇

So where we started with a great example of decentralization with the ethereum smart contracts, we end up with something that is very much not decentralized and hard to trust.

👇

With the recent events related to Tornado Cash, it's become very clear that the need for more front end decentralization of apps is of importance.

Now what can we, or should we do about the current paradigm of decentralized applications?

👇

Let's fix it! We start to rearrange some blocks and change the locus of ownership around them.

We'll now package the IPFS applications in a more user-centric view of the world, such that a user might be interested in a specific version and want to lock that down.

👇

We'll now mint an NFT that pins its metadata on IPFS immutably and that metadata is pinned immutably to a specific version of the app which the user trusts.

And, then we'll use our wallet to fetch the details of the NFT and run just that front end.

👇

We've now created a much better trust chain where you are controlling each layer of what you are running!

With the bring your own application (byoa) paradigm of decentralization, we have all the building blocks in place!

Original byoa whitepaper: github.com/web3-byoa/whit…

👇

Caveat: It may still be susceptible to censorship at the source code layer, but you are able to decide what version you are pinning to IPFS and able to get around any potential censorship as you see fit, instead of trying to run the latest version that DNS tells you to run.

👇

Ideally, In a truly decentralized version of this, your app would communicate with the IPFS chain directly instead of even using an IPFS gateway resolver to fetch data.

👇

What we've built here are a few key building blocks:
byoa-sdk (which Mallows pioneered)
L1 Global App Registry Smart Contract
L2 (StarkNet) Personal App Registry Contract
decentralized IPFS L1 app store
decentralized IPFS L1 byoa-sdk impl.
decentralized IPFS L2 byoa-sdk impl.

👇

Above this manifests itself in:
store.byoa.org installing Apps on L1
home.byoa.org installing Apps to your L2 (@StarkWareLtd) wallet
appsafe.eth.limo run IPFS impl of byoa-sdk for L2 apps
l1.appsafe.eth.limo run IPFS impl of byoa-sdk for L1 apps

👇

With these tools you can either choose to mint your byoa apps as NFTs and run on L1 with l1.appsafe.eth or use the personal registry version on L2 with StarkNet and run on appsafe.eth.

👇

L2 is far cheaper. Both are secure. But some may like the paradigm of the "physical" nft being in their wallet versus a smart contract registry only.

👇

What we have here are all the building blocks for a simple decentralized app store which is far more censorship resistant than what we have today.

Go get started! Let's build the future of true end-to-end decentralization!

👇

We've been building on the concept of "Bring your own Application" for true decentralization and ownership for over a year now.

Led by @TheCryptoPeter and @ryanberckmans we've laid a foundation of technical and philosophical approach to decentralized apps.

👇

There is still more to research, build, and explore in this space. We haven't touched on fair governance for such an "app store" yet, or how to mitigate more DNS related issues one could have. But, we'll keep researching and building.

👇

The original idea of byoa was to push forward the idea that we had the technology and tools to create experiences we truly owned and trusted. We debuted that with Mallows which were byoa apps which modified the environmental substrate around them.

👇

What we've seen in the past year was the need less so for just environmental modification, but more so for full environmental control.

Our tools are now open source, so let's build it!
/end 🧵