Proposal: Create a unique source of truth for CAIP-2 chainIds

[vseehausen]

Hey,

we are currently working on implementing the CAIP-2 and CAIP-19 standards at the ITSA.global.

I realized that it is difficult to find out, which chainIds are actually community-approved and which are not.

Could we brainstorm on how to list those standards? We were also not sure about the case-sensitive spelling. For example: is it EIP155 or is it eip155 like in the examples?

When we can agree on a process to approve this and, more importantly, on a place to list the chainIds, the ITSA could reference and circulate this.

[ukstv]

I guess, you should have a look at https://github.com/ChainAgnostic/namespaces repository. There you could see what CAIPs are available per blockchain = namespace. For example, you could see that for eip155 = Ethereum-based chains 4 CAIPs are standardised, while fil = Filecoin has only CAIP-2.

Everything in that repository is community-approved.

[vseehausen]

Thank you very much, that is perfect!

Two questions:

1. Would it make sense to link these documents prominently in the CAIP-2 and CAIP-19? Our team was using those two documents as a reference and obviously, we failed to find the proper namespace definitions. Maybe this helps other newbies, too.
2. Would it be beneficial, if the namespaces would be referenced on the ITSA docs or the website? Not sure about the form yet, but I think that creating a list with all namespaces, that gets updated regularly, would be a good resource.

[bumblefudge]

1. Good thinking, I'll run it by the editorial group.
2. Yes, we would appreciate references that bring attention to both repos, /CAIPs/ and /namespaces/. They are both slow-moving but ever-growing references. I think my personal preference would be for links to root rather than lists of links to what's there at any one point in time.

[vseehausen]

1. If you need help, let me know. I am new (to CASA) and happy to help. I will also be in Berlin on Sept. 15th.
2. a) Agree. I envision something like a "list" in this repo. This "list" is the unique source of truth (at least it tries). Other sources link /reference these. ITSA is one of the first institute to link/reference this.
3. b) Maybe there should be some kind of versioning for the "list". Could be as easy as a last updated at timestamp. This would allow institutes to put the data in their own design and logic.

My rational why this could be beneficial, please correct if I am wrong: In the end CASA defines standards. Only a wide spread adoption defines the success of the standard, if I get it right. So making it easy to non-techy people to access the data is IMHO an important step. I can tell from experience of multiple discussions that even the different labels (namespace / id / type) are complicated to understand. So this really is my approach to boild down the whole CAIP-2 / CAIP-19 identification to an API that is easily understood by business people without technical background.

https://xkcd.com/927/

[bmann]

Chainlist https://chainlist.org/ is the published source of truth for ChainIDs. I forget if there is again / still a JSON version of that list as we once had.

I know there was an earlier thread around that domain / website, I'll just re-iterate what needs to be done:

• [ ] verify / secure ownership of the chainlist.org domain, and build process
• [ ] add CASA logo and link back to github org
• [ ] add JSON version so that chainlist.org/chainlist.json or similar is a machine readable version
• [ ] (optional) IPFS publishing flow to get versioned links and/or ENS path

For CAIP 19, we need to also publish those identifiers to a site following a similar pattern.

[vseehausen]

Ok, nice.

So do you want chainlist.org to be the source for the CAIP-2 identified chainIds? Right now, chainlist.org only contains EVM chains.