Use DIDs

[SionoiS]

I think that switching to DIDs instead of, right now, IPNS addresses is a good thing. The DID spec allow for multiple kind of DIDs.

IPNS

https://docs.ipfs.io/concepts/ipns/

• Included in go-ipfs, no other dependency.
• Can use PubSub for real time update.

--- PRO --- CON ---

• Slow
• Managing IPNS records is a mess.
• No easy way to hold your keys securely. Ideally you'd want a browser extension that ask a hardware wallet to sign message so that it's easy to update.

ION

https://identity.foundation/ion/

• Faster resolution? Would have to test that it's actually faster than IPNS via Pubsub.

--- PRO --- CON ---

• No easy way to hold your keys securely. Ideally you'd want a browser extension that ask a hardware wallet to sign message so that it's easy to update.
• Not integrated with IPFS. It add another dependency.

Others

https://w3c.github.io/did-spec-registries/#did-methods

Counter-Argument

https://dwhuseby.medium.com/dids-are-dead-32a5295016c4

[SionoiS]

One could implement an IPNS system useable in browser with an hardware wallet.

[alibama]

what are your thoughts on Ceramic and their DID tools?

[SionoiS]

what are your thoughts on Ceramic and their DID tools?

I do something similar to their "stream" concept, the main problem is adding a dependency.

It add complexity and what do I get out of it? IDK!

Also this about DIDs is good counter-argument.

[alibama]

Thanks man - i'm reading that article now - I am really still trying to get my head around the whole ecosystem. i've also been looking at brightid.org and find it really attractive in more ways for its simplicity and more algorithmic & anonymized so somewhat intrinsically more secure approach

what i see as potential value is the organizing layer of ipfs - since that is something of a mess.

i'm also playing around here https://gitcoin.co/issue/ceramicnetwork/ceramic/80/100026723 and would totally enjoy your perspective if you're interested... perhaps DIT could build out a schema "standard" to simplify integrations and discovery

[SionoiS]

Thanks man - i'm reading that article now - I am really still trying to get my head around the whole ecosystem. i've also been looking at brightid.org and find it really attractive in more ways for its simplicity and more algorithmic & anonymized so somewhat intrinsically more secure approach

what i see as potential value is the organizing layer of ipfs - since that is something of a mess.

i'm also playing around here https://gitcoin.co/issue/ceramicnetwork/ceramic/80/100026723 and would totally enjoy your perspective if you're interested... perhaps DIT could build out a schema "standard" to simplify integrations and discovery

BrightID cannot be used in a decentralized ways It require a central registry for verification. You could look at the Upala Protocol it's more open and would warp around any kind of identity system.

As a general rule if your asked to; sign up, pay something or something is gated, not open source, it's not decentralized even if they say so. Decentralized is just a buzz word at this point you got to read and understand how the system really work.

About organising data, that's what IPLD is for, no reason to use the Ceramic network as far as I know.