Ethereum

Recasting: 𝗙𝗲𝗲𝗱𝗯𝗮𝗰𝗸 𝗿𝗲𝗾𝘂𝗲𝘀𝘁: What is "Ethereum alignment" for wallets?

My take: It is a set of attributes around:
Security, privacy, self-sovereignty, user experience, and transparency.

Breakdown as a thread. 🧵

It is no coincidence that these echo the values sought in other parts of the Ethereum stack.
@vitalik.eth talks about this in his blog post on making Ethereum alignment legible:
https://vitalik.eth.limo/general/2024/09/28/alignment.html
Goal: Keep all parts of the Ethereum stack true to the original Ethereum mission.

Wannabe contributor for /walletbeat.

Censorship resistance requires privacy.

This can't be done by any central actor; it must be bottoms-up.
L2Beat is the best implementation of this in Ethereum, focused on L2s - a critical part of Ethereum today, but not the only one.
Other parts must be held up too: MEV, transaction censorship, staker centralization, and... wallets.
Who watches the wallets?

walletbeat is to my knowledge the best but only effort to date so far: https://walletbeat.fyi

It was created by @moritz and already categorizes wallets along several important features and dimensions.

My goal is to extend it. I am looking for feedback as to how.

So what attributes are desirable in a wallet?

- ⚙️ Standards adherence
- 🔒 Security
- 🕵️ Privacy
- 👑 Self-sovereignty
- 🧈 User experience
- 🔎 Transparency

(in no particular order)

This list is focused on values-aligned attributes, not nice-to-have UX bells and whistles, or things like "features are shipped fast".

These are important too, but competitive pressures already uphold these.

The list intentionally narrows down on attributes that pure market forces may not converge on their own.

⚙️ 𝗦𝘁𝗮𝗻𝗱𝗮𝗿𝗱𝘀 𝗮𝗱𝗵𝗲𝗿𝗲𝗻𝗰𝗲 is the obvious one. It means support for various standards that make wallets work and interoperate with dapps.

- 𝗘𝗢𝗔 𝗰𝗼𝗺𝗽𝗮𝘁𝗶𝗯𝗶𝗹𝗶𝘁𝘆: Does the wallet support EOAs?
- 𝗔𝗰𝗰𝗼𝘂𝗻𝘁 𝗮𝗯𝘀𝘁𝗿𝗮𝗰𝘁𝗶𝗼𝗻: Does the wallet support insert-your-favorite-AA-ERC-here wallets?
- 𝗔𝗱𝗱𝗿𝗲𝘀𝘀 𝘀𝘁𝗮𝗻𝗱𝗮𝗿𝗱𝘀: Does the wallet support ERC-3770 addresses?
- 𝗟𝗼𝗴𝗶𝗻 𝘀𝘁𝗮𝗻𝗱𝗮𝗿𝗱𝘀: Does the wallet support ERC-4361 (Sign-In with Ethereum)?
- 𝗧𝗼𝗸𝗲𝗻 𝘀𝘁𝗮𝗻𝗱𝗮𝗿𝗱𝘀: Does the wallet support and display ERC-20, ERC-721, and ERC-1155 tokens?

Are there more?

🔒 𝗦𝗲𝗰𝘂𝗿𝗶𝘁𝘆 is also obviously critical, but there are different aspects to it:

- 𝗖𝗵𝗮𝗶𝗻 𝘃𝗲𝗿𝗶𝗳𝗶𝗰𝗮𝘁𝗶𝗼𝗻: Does the wallet bundle a light client to verify L1 state?
- 𝗧𝗿𝗮𝗻𝘀𝗮𝗰𝘁𝗶𝗼𝗻 𝘀𝗶𝗺𝘂𝗹𝗮𝘁𝗶𝗼𝗻: Can the wallet simulate transactions and show the effect of a user’s transaction prior to being signed?
- 𝗦𝗰𝗮𝗺 𝗮𝗹𝗲𝗿𝘁𝘀: Does the wallet alert the user about potentially fraudulent transactions they are about to sign?
- 𝗩𝗲𝗿𝗶𝗳𝗶𝗲𝗱 𝗯𝗮𝗰𝗸𝘂𝗽𝘀: Does the wallet verify that the user understands and has backed up the factors used to sign transactions?
- 𝗔𝗰𝗰𝗼𝘂𝗻𝘁 𝗿𝗲𝗰𝗼𝘃𝗲𝗿𝘆: Can users recover their account if they forget one of the factors used to sign transactions?
- 𝗔𝘂𝗱𝗶𝘁𝘀: Has the source code recently been audited?

Others I've considered adding to 🔒 𝗦𝗲𝗰𝘂𝗿𝗶𝘁𝘆:
- Software supply chain security: Reproducible builds, auditable dependencies
- Protection against malicious updates - e.g. from a compromised dev adding private key exfiltration code - but this is very difficult to require out of wallet software.
Would welcome additional thoughts here.

🕵️ 𝗣𝗿𝗶𝘃𝗮𝗰𝘆 is a big one and something most wallets do pretty bad on today (with notable exceptions!). I am hoping that public scrutiny will help make meaningful progress here.

Ethereum privacy as a whole desperately needs more research across the whole stack, but the following are concrete improvements that today's wallet software can improve upon immediately: (👇)

- 𝗣𝗿𝗶𝘃𝗮𝘁𝗲 𝘁𝗿𝗮𝗻𝘀𝗮𝗰𝘁𝗶𝗼𝗻𝘀: Does the wallet support sending funds privately? Receiving funds privately? Spending privately-received funds without leaking information? Very few wallets integrate privacy solutions today.
- 𝗜𝗣 𝗮𝗱𝗱𝗿𝗲𝘀𝘀 𝗹𝗲𝗮𝗸𝗮𝗴𝗲: Can a third party learn associations between IP addresses and Ethereum addresses?
- 𝗜𝗱𝗲𝗻𝘁𝗶𝗳𝘆𝗶𝗻𝗴 𝘂𝘀𝗲𝗿 𝗶𝗻𝗳𝗼𝗿𝗺𝗮𝘁𝗶𝗼𝗻: Does the wallet transmit user information to third parties?
- 𝗠𝘂𝗹𝘁𝗶-𝗮𝗱𝗱𝗿𝗲𝘀𝘀 𝗰𝗼𝗿𝗿𝗲𝗹𝗮𝘁𝗶𝗼𝗻: When configured with multiple Ethereum addresses, can a third party learn that these addresses collectively belong to the same person? 

There may be more here, such as private transaction broadcasting that need more fleshing out.

👑 𝗦𝗲𝗹𝗳-𝘀𝗼𝘃𝗲𝗿𝗲𝗶𝗴𝗻𝘁𝘆 is the desirable property of properly-decentralized system, and one way to encode the "walk away" test: how dependent is a wallet on third parties?

- 𝗦𝗲𝗹𝗳-𝗵𝗼𝘀𝘁𝗲𝗱 𝗻𝗼𝗱𝗲: Can users use their own Ethereum node for L1 interactions?
- 𝗧𝗿𝗮𝗻𝘀𝗮𝗰𝘁𝗶𝗼𝗻 𝗰𝗲𝗻𝘀𝗼𝗿𝘀𝗵𝗶𝗽: Can users self-broadcast a transaction without being blocked by a third party?
- 𝗟𝟮 𝘄𝗶𝘁𝗵𝗱𝗿𝗮𝘄𝗮𝗹𝘀: Can users withdraw their funds from L2s trustlessly?
- 𝗧𝗿𝘂𝘀𝘁𝗹𝗲𝘀𝘀 𝗳𝗿𝗼𝗻𝘁𝗲𝗻𝗱𝘀: Can users use popular dapps trustlessly? (e.g. via IPFS or ERC-6860 frontends)

Today's wallets depend a lot on third parties; even if untrusted via light clients, they are still dependent on their 𝗮𝘃𝗮𝗶𝗹𝗮𝗯𝗶𝗹𝗶𝘁𝘆. This makes them a censorship vector.

🧈 𝗨𝘀𝗲𝗿 𝗲𝘅𝗽𝗲𝗿𝗶𝗲𝗻𝗰𝗲 has been a focus over the last year. L2 fragmentation has been a pain point, and wallets have one of the largest roles to play in addressing it.

- 𝗖𝗵𝗮𝗶𝗻 𝗮𝘂𝘁𝗼-𝘀𝘄𝗶𝘁𝗰𝗵𝗶𝗻𝗴: Does the wallet automatically switch to the correct chain when apppropriate?
- 𝗖𝗵𝗮𝗶𝗻 𝗮𝗯𝘀𝘁𝗿𝗮𝗰𝘁𝗶𝗼𝗻: Does the wallet assist the user in (transparently) bridging assets (L1-to-L2, L2-to-L1, L2-to-L2)?
- 𝗔𝗱𝗱𝗿𝗲𝘀𝘀 𝗿𝗲𝘀𝗼𝗹𝘂𝘁𝗶𝗼𝗻: Does the wallet resolve named addresses (e.g. ENS) when sending assets?
- 𝗡𝗼𝗻-𝗻𝗮𝘁𝗶𝘃𝗲 𝗴𝗮𝘀 𝗽𝗮𝘆𝗺𝗲𝗻𝘁𝘀: Can users pay transaction fees in assets other than the native gas token?
- 𝗣𝗲𝗿𝗺𝗶𝘀𝘀𝗶𝗼𝗻 𝗿𝗲𝘃𝗼𝗸𝗶𝗻𝗴: Does the wallet assist in monitoring and revoking token permissions?

I'm sure there could be more here!

Lastly, 🔎 𝘁𝗿𝗮𝗻𝘀𝗽𝗮𝗿𝗲𝗻𝗰𝘆 in the way a wallet works is essential to being trustworthy by its users:

- 𝗦𝗼𝘂𝗿𝗰𝗲 𝘃𝗶𝘀𝗶𝗯𝗶𝗹𝗶𝘁𝘆: Is the source code visible to the public?
- 𝐎𝐩𝐞𝐧-𝐬𝐨𝐮𝐫𝐜𝐞 𝐥𝐢𝐜𝐞𝐧𝐬𝐢𝐧𝐠: Is the source code licensed under an open-source license?
- 𝗙𝘂𝗻𝗱𝗶𝗻𝗴 𝘁𝗿𝗮𝗻𝘀𝗽𝗮𝗿𝗲𝗻𝗰𝘆: Is the wallet transparent about its monetization strategy, if any? (e.g. VC funding, token allocation, user-visible swap/onramp fees)

Perhaps audits would fit here, but they are already under "🔒 𝗦𝗲𝗰𝘂𝗿𝗶𝘁𝘆". Reproducible builds would also be nice to have perhaps? Again, feedback welcome.

This is what I've been able to come up with so far - with credits to @moritz, @v, @harris-, and others who have contributed to the discussion so far.

I have posted this to Ethereum Magicians for feedback as well:

This is what I've been able to come up with so far - with credits to @moritz, @v, @harris-, and others who have contributed to the discussion so far.

I have posted this to Ethereum Magicians for feedback as well:

https://ethereum-magicians.org/t/making-ethereum-alignment-legible-wallets/21841

walletbeat

If this is exciting to you, please follow /walletbeat where I will be posting updates.

One thing I haven't discussed yet for example is how this can all be integrated into the existing walletbeat website (wen walletbeat.eth domain?). But that's for another thread.

Thanks for reading and I welcome your feedback!