Simplify Process 2021 (candidate|proposed) (change|addition|correction) terminology

[msporny]

The Process 2021 document introduces the following terminology:

• candidate addition, in §6.1.5
• candidate change, in §6.1.5
• candidate correction, in §6.1.5
• proposed addition, in §6.2.11.5
• proposed changes, in §6.2.11.5
• proposed corrections, in §6.2.11.5
• candidate Amended Recommendation, in §6.2.11.3
• Amended Recommendation, in §6.2.1
• Last Call for Review of Proposed Corrections, in §6.2.11.5
• Last Call for Review of Proposed Additions, in §6.2.11.5
• Last Call for Review of Proposed Corrections and Additions, in §6.2.11.5

I have read through the sections detailing the differences between all the terminology at least six times now. I am being literal, that is not an exaggeration, and I still have a hard time keeping all variations in my head when trying to understand what I need to do for each as an Editor of multiple specs at W3C. Others are probably going to have the same difficulty that I do in understanding why all of that nuance is needed.

Is it possible for us to simplify the "Amended Recommendation" terminology down to the following?

• editorial erratum change
• substantive erratum change (includes new feature additions)
• Proposed Amended Recommendation
• Amended Recommendation
• Last Call for Review of Amended Recommendation

[jeffjaffe]

Linking this to the previous attempt: #358

[frivoal]

This is complex, and simplifications would indeed be good. But there are underlying reasons for this complexity. If we want to simplify, we need to deal with those first, and not just treat it as a terminology question.

editorial erratum change
substantive erratum change (includes new feature additions)

We specifically need to distinguish between substantive changes that are additions, and those that are not. Previously issued recommendations explicitly disallowed to former, but not the later. Either we distinguish, or we allow ourselves to break a multi-decade old promise, or we just disallow feature additions in RECs ever. Besides, allowing new features in a REC is something that some groups, but not all, want. Being able to tell the two apart remains useful even besides historical concerns.

We could reduce the vocabulary by one third by dropping the word "change", since it merely refers to the superset of additions and corrections, but then we'd have to say "additions and/or corrections" instead every time we talk about rules that apply to both, which doesn't seem like an improvement to me (though I agree that is subjective).

You are also suggesting to merge the Candidate and Proposed stage. This too is not just a matter of terminology. A Candidate change is one that has the blessing of the WG. A Proposed change is one upon which AC Review and Patent review is being called upon. It's not reasonable to do an AC review + Patent review for every tentative update of a spec. And if we ask the group to batch all their changes in some unofficial place until they have enough for a review, we're back to the earlier monolithic update mechanism that this patch-in-place process is intended to alleviate.

Merging the three types of "Last Call for Review of *** " is indeed just a matter of terminology. We could merge them without changes to what the Process does, but so long as different types of change exist, it seemed more useful to have 3 terms in the process to refer to them rather than need to inline this definition / differentiation in every communication that talks about them.

[frivoal]

https://github.com/w3c/AB-memberonly/issues/244 would inform whether we can simplify here