Guess we failed pretty spectacularly in keeping up with the biweekly updates. Oh well.
This post covers everything we've accomplished so far this year, and what we're planning for next year.
We began our year by rewriting our prototype for Wallet Test Framework in TypeScript, to be more accessible to wallet developers. The original implementation was half TypeScript and half Rust, and without a specific reason to use Rust, we felt it more consistent to use TypeScript for everything. Around this time, we also migrated to the Mocha testing framework.
Soon after, we set up our continuous integration pipeline to deploy the framework to Heroku, making it easier to play with without having to build it. We simplified the backend, and created scaffolding for glue implementations.
After some bug fixes and new tests, we migrated from ethers to viem, due to some caching issues. We also introduced support for remote controlling wallets over WebSockets.
After publishing our first Test Report, we moved full steam into developing automation glue for Coinbase and Taho wallets. We've proven that our tests are portable between wallets, and we've reported a handful of issues upstream.
After a fairly successful first year, we're really looking forward to 2024!
Our top priority for the upcoming year is to broaden our support for wallets,
especially mobile. Wallet Test Framework currently depends on
pretty much limiting us to browser extension wallets. To better support the
Ethereum community, we need to integrate technologies like WalletConnect,
EIP-6963, and vendor-specific SDKs.
Wallet Test Framework is only useful if it is run, and can only find bugs for areas it tests. Our second goal for 2024 is to write documentation for wallets, EIP authors, and security experts so that they can all use and contribute easily to the project.
We plan to write integration guides and documentation for wallets to make it as easy as possible to run WTF as part of testing pipelines. This will cover topics like writing the automated glue and adding vendor-specific tests.
Our biggest strength is that once a test is written, it's available to every wallet using the framework. To take full advantage, we need to make it as easy as possible for anyone—be it wallets themselves, EIP authors, or complete outsiders—to write WTF tests. Documenting the test-writing process and having clear guidelines for contributions are necessary to accomplish that goal.
This might be a bit of a stretch goal, but we'd like to start testing how wallets decode the transactions of some popular applications. This is a wide attack surface with plenty of room for quality assurance.
EIP-1193 defines several events (like
aren't covered by the current tests. Emitting these events is essential for
responsive application development, so we'd like to expand our coverage in this
The world of account abstraction wallets has exploded since the introduction of ERC-4337. We'd like to explore expanding our tests to cover them.
We'd like to develop a dashboard showing the scores of wallets over time, as well as automation implementation status, and enabled test suites. These statistics will allow us to gauge how much of an impact Wallet Test Framework has on the ecosystem.
We want to give a huge thank you to the Ethereum Foundation's Ecosystem Support Program and to the donors on Gitcoin for helping to fund our work so far. We'd also like to thank Coinbase for being very supportive and responsive when reporting issues.