CIP-0001 | Add Versioning + other small edits

[Ryun1]

Changes

• Added stipulation that CIPs must include a versioning scheme (Previously no such requirement existed).
• Added that on-chain metadata CIPs must include a CDDL schema (Previously this has been asked for by editors, but was unreflected here).
• Added suggestion that "pre-numbered" CIPs can be named CIP-XXXX or CIP-???? (previously just CIP-???? suggested).

Motivation

• Lack of versioning in CIPs has been a long standing frustration in the community. Here I am trying to focus this conversation to reach a conclusion. I do not believe these first additions are sufficient, rather I want to hear communities opinions and this PR can act as a first step.
• We as CIP editors ask authors for this, thus it needs to be mentioned in CIP-0001.
• For Windows users using ? in filenames is not possible, so offering an standard alternative seems sensible.

TODO:

• Reflect historical versioning frustrations in Rationale section.
  • Won't do - save this for a larger discussion.

[rphair]

I think the versioning issue could be handled & explained a number of ways, but I don't think we have to take all of these steps at once: so having the additional language requiring versioning in the Specification would help get us ahead of the versioning issues with all CIPs submitted from this point forward. ✔️

Points in favour of a quick merge (pending my 1 suggested change below), assuming no disputes from editors:

• Though I think we should announce this PR in the dev community ASAP, I would rather have this language merged immediately than spend another 3 months trying to get full buy-in from everyone about a more elaborate versioning scheme, get the usual difference of opinion over the further details, and then possibly fail to address versioning issues with CIPs and potentially breaking updates that come along in the meantime.
• Since the (thin) feedback we've got from devs so far via Versioning alternatives #520 & elsewhere has been in favour of more versioning rather than less, if we choose to merge this simple requirement without buy-in from the whole community then I don't think we'd be contradicting anyone's already expressed opinion.

If for some reason this language doesn't cover every case (particularly for those CIPs for which versioning doesn't make any practical sense), CIP authors can do either of the following:

• add a statement in their Specification explaining why versioning isn't appropriate for their CIP; or
• offer versioning with perhaps a different scheme than the one we've anticipated, and provide a more detailed justification for it in the Specification & we'll accommodate it somehow in the editorial process.

If anyone thinks any of the above should be added to the CIP then let me know and I'll have a go at it with a code suggestion. Otherwise I think this is ready as offered, with the small change below about a different issue.

N.B. while on the subject of versioning there was a last-minute suggestion from @Crypto2099 (#520 (comment)) which might be debated here:

• ... with a bit more background in his draft CPS here: https://github.com/Crypto2099/CIPs/tree/versionioning-cps/CPS-XXXX#2-framework-for-versioning
• ... though for the reasons above I think either set of changes, as they affect versioning requirements in CIP-0001, would better be put into another PR on top of this one.

[rphair]

OK, with 4877c07 I am in favour of merging this for all the reasons already given.

[Crypto2099]

This seems like a simple first step towards a larger discussion on versioning and organization as a whole and the change to pre-numbered organization (thank you) is also a minor change.

[Ryun1]

@rphair @Crypto2099

Thanks for the suggestions, I like the idea of just merging this quickly as a first step and thus I have updated the PR description to better reflect this.

I think we should wait for another author's approval before merging though 🤓

offer versioning with perhaps a different scheme than the one we've anticipated

Good idea - I have adjusted the wording to address this. As well as made the wording a bit more explicit in general.

[rphair]

@Ryun1 @KtorZ if we agree to do this we should also add a "versioning remediation" step to the outline here (at least to consider whether such remediation is possible, especially for the CIPs which have been around for a while): #389 (comment)