keep-starknet-strange / backlog

Keep Starknet Strange projects backlog and ideas
MIT License
3 stars 0 forks source link

Starknet Phone #33

Open drspacemn opened 3 months ago

drspacemn commented 3 months ago

Starknet Phone Preface

Category: MVP

Exploration Hypothesis

The current mobile UX is very bad. The way to approve it is to implement a light client and wallet at the OS level.

Scope:

options for development:

options may not be all that different, since the implementation is done with apps, rather than integrated directly in the OS. So we can potentially create light client app and integrate into either OS.

  1. Fork grapheneOS, using ethOS implementation as a reference.

    • pros
      • clean slate: only additive
      • can still use ethOS implementation as reference
      • good documentation
    • cons
      • clean slate: might need to write more code?
  2. Fork ethOS and swap to support Starknet

    • pros
      • might speed up development
    • cons
      • not easily forkable: lot of different repos
      • may have a lot of code to remove to get started
      • poor developer documentation
  3. Simply implement light client + wallet + injected browser on android directly

    • technically, graphneOS isnt needed.
    • pros
      • more support
      • better OS UX compared to GrapheneOS
        • better app support etc.
      • google play services: can integrate if desired
    • cons
      • less secure

Important question: is the priority UX or security?

Chosen solution: Fork grapheneOS and add in starknetOS apps using ethOS as reference why: forking may bring more headache than it solves, easier to add code than remove it, the light client implementation for starknet may not be similar (geth android lib vs. porting beerus rust crate)

Project Continuation

DP from SNF