search

IBM / ipfs-social-proof

IPFS Social Proof: A decentralized identity and social proof system
MIT License
142 stars 27 forks source link

Proofs class #23

Closed daviddahl closed 5 years ago

daviddahl commented 5 years ago

A proofs class that mints proofs as a more "standard" JSON format, with a version and a JSON schema to validate them is a pretty good idea

tyleryasaka commented 5 years ago

Taking a stab at this

daviddahl commented 5 years ago

I got super carried away today on proof-related stuff and this happened - none of it is tested yet: https://github.com/IBM/ipfs-social-proof/pull/26/files#diff-1ba4c7ddfb9ca96b92ca92976bb89916

daviddahl commented 5 years ago

@tyleryasaka perhaps you can critique the approach I am taking? And - maybe we can discuss the best way to split this up :) its just too big. You pull one thread and the whole sweater turns back into yarn!

I think the way i am splitting it up will make a test suite easier to create

daviddahl commented 5 years ago

Perhaps what you can do is follow the initial comment here and just put some thought into what a proof should look like. I have been looking at Keybase proofs for inspiration. My current proof is rather simple, just a message object to stringify & sign, some ID metadata, public key and the signature.

tyleryasaka commented 5 years ago

Sounds good! Taking a look now.

I actually think it's fine not to worry too much about the proof format for now, in the early stages. Proof of concept that gets something minimal implemented should be fine. But that's just me.

tyleryasaka commented 5 years ago

Glancing over that branch, looks good to me! I like seeing everything getting split into more meaningful and readable chunks.

Yeah I can think about schemas and maybe post a proposal here.

daviddahl commented 5 years ago

fixed in #26