Day 1, 1:30 pm - 2:30pm
@benjaminbollen’s session with some distracted interjections from me!
I care about this topic and do want to follow it!
Adding my notes here from this session. @mariapaulafn also has comprehensive notes here
Founding Question: How do we make UX design better for users having access to their keys and their tokens
- Need to find the balance between 12 words on a piece of paper vs. the “giving everything” approach at Coinbase
Also looking at privacy and wallets and login recovery
- Multi signature smart contracts to hold the users tokens for them - use good wallet design, so user doesn’t need to hold the key and a trusted party holds them and when the user has time, then they learn more about how to own and control their keys
If company owns the keys, what are the clear set of values for the companies/Coinbase model?
Private Keys should not migrate - educate users on the experience upfront so they know their value/the experience
Also, worst user experience is to create another private key Can users bring their own private key?
Smart contracts managing keys can help for generalized recovery mechanisms (you can always use those 12 words, lose the keys, etc.)
MP challenged that this brings in the layer of centralized Trust - which may go against trustless systems principle
Follow up reading: Uport white paper for consensus proposals for managing keys
Jason shared that we need to educate users at the point of where they care - doesn’t matter if you educate them upfront, it depends on what they need to use the wallet for and size (of need and transaction) matters.
Boris mentioned how do we segregate wallet concepts as they different b/w blockchains vs. identity and login (reclaiming wallets/keys)
Entity vs. Identity (Kopinski’s whitepapers on user frameworks)
Entity is my legal boundary, money is horizontally relevant (it is relevant to all communities that use that currency)
Identity is the sum of all my entities and attention on aspects of you is what creates identity made by a claim represented by an authoritative party (education, company, etc.); vertically relevant, you cannot take this relevance amongst other groups
Research: Zippy on mobile access and storage, need to tie to mobile hardware and OS for powering mobile blockchain apps (must think beyond just Ethereum community when building wallet structures and must focus on the end user’s goals/resources).
How do we create the dApp userflow for using challenges to educate users to come on board with using their private key?
UX Design for onboarding new users and understanding their wallets/private keys
What technical needs do we have to consider for supporting multiple key management and recovery?
- I really think people just weren’t meant to screw around with private keys. It should be something that should be embedded deeper into the operating system or browser, with more user friendly management scheme’s on top. Casa wallet’s managed wallet service (3 factor seed phrase recovery and social sharing) is really neat. More experimentation should be done in this regard, eventually building systems like @alexvandesande’s personal DAO idea.
- Identity needs a lot more work. I think there’s a huge opportunity here to redesign the account systems of all applications and ecosystems. Kill the password dead beyond all reckoning, and introduce new paradigms that are more secure and natural to users (device based authentication, universal profile, not having to set my profile picture on a new app).
Interesting idea I just had is that you can derive multiple accounts from your seed phrase… One (or more) account could be your personal, one could be your professional, or you could have one for each job you ever have, or every year you use a new one for tax reasons. Play with that a bit more? Seems interesting.
Yes, it is probably too late for me to be rambling.
I’m adding my notes in dialogue format.
Please everyone check I did write your names (or not - clearly i missed some) and comment on the google doc.
Next step is forming a Ring (working group). Except the Ring-formation meta group needs to define the formation process
I’ve heard Mordor has a forge…
uPort White Paper is deprecated, I think this is a pointer to all their stuff: https://medium.com/uport/a-complete-list-of-uports-protocols-libraries-and-solutions-63e9b99b9fd6
@ligi this is partially wallets, but I think logins & identity should be decoupled from wallets?
@benjaminbollen thoughts on naming?
Will there be any meetings about this in Prague? I would be interested to be learn more about what you all have been working on.
It seems unlikely that identity can (or should) be managed via a dApp. Identities should likely exist on the protocol level. Fundamentally each identity might be recognized as an oracle. Does it seem sensible to create a new classification for these?
We created a service (like an oracle) for identity verification of Ethereum address/key owners on https://cryptonomica.net
To verify identity of Ethereum address private key owner user has to sign a simple message with his/her verified OpenPGP key (we have online verification service on our website), and send this message from his/her Ethereum address to our smart contract.
Then user requests our system to pull this signed message from smart contract and check it’s signature. If signature can be verified, our server puts user data (first name, last name, birthdate, nationality, OpenPGP key fingerprint, Ethereum address) to smart contract on https://etherscan.io/address/0x846942953c3b2A898F10DF1e32763A823bf6b27
User data from this smart contract can be read by other smart contracts and decentralized applications. Users can use verified identity to make legal binding and also legally enforceable contracts/documents on the Ethereum blockchain.
Ethereum address can be connected to one Cryptonomica verified OpenPGP key only, if OpenPGP key expires, user have to use another Ethereum address for new OpenPGP public key.
This service created to be as much smart contract developer friendly as possible. See smart contract verified code on etherscan: https://etherscan.io/address/0x846942953c3b2A898F10DF1e32763A823bf6b27f#code