Errata'd EIPs

Question

Errata'd EIPs

Pandapip1 opened this issue 2 years ago · comments

Proposed Change

I propose that with the approval of an original non-errata'd Final EIP's author, a new EIP can be created with the EIP "number" <original-eip-number>-e<number-erratas>, where <original-eip-number> is the original Final EIP's number, and <number-erratas> is a number that starts at 1 and increments by 1 every time an errata EIP for the given original EIP number reaches final, withdrawn, or stagnant.

So, instead of modifying EIP-721, a new EIP would be created with the number 721-e1, referenced as "EIP-721-e1".

Gavin John commented a year ago

Bump.

Sam Wilson · Answer 1 · Wed Sep 21 2022 22:28:46 GMT+0800 (China Standard Time)

If we decide the changes are errata, we can change a final EIP (and git keeps the history.)

If we decide it isn't errata, it's a new EIP and I think it deserves its own number.

I'm also concerned that complicating the EIP number namespace will just make things worse.

Gavin John · Answer 2 · Thu Sep 22 2022 07:00:50 GMT+0800 (China Standard Time)

If we decide the changes are errata, we can change a final EIP (and git keeps the history.)
If we decide it isn't errata, it's a new EIP and I think it deserves its own number.

The point here is that you might have something like EIP-712 where there are valid arguments for and against "it's an errata" vs "it's actually a major change." Skipping the discussion and letting the authors make that judgment call saves EIP editor brain bandwidth.

Dennis von der Bey · Answer 3 · Sat Dec 03 2022 17:23:34 GMT+0800 (China Standard Time)

I love the idea. There will always be a chance for small formal errors in EIPs as we're all humans after all and EIP numbers are already being recognized in the heads of the people. Due to this a sudden change of the number for a new version of a well known standard might be damaging the communication.

github-actions · Answer 4 · Thu Jan 26 2023 08:18:30 GMT+0800 (China Standard Time)