User:Crazynas/rfc

RFC as it currently stands is not a suitable method to reach a consensus agreement, particularly regarding policy pages. It seems that in the course of attempting to regulate and maintain consistency, in a certain sense we've forgotten what the wiki model is all about.(hint: reread WP:Bold) This is a method of continuous revision with regards to policies and guidelines, that would provide a balance between retaining policy stability, continually soliciting wide ranging input, and providing an effortless method to determine if a new consensus has been reached.

Theoretical Overview
Currently, the way we manage and organize changes to major policy pages is fraught with peril. As our polices are effectively our Constitution, changing them should be something that requires wide ranging input. Simultaneously, policies are there to reflect current practice, change must be attainable.

The recent fiasco at changing the lead of WP:Verifiablity demonstrates that our current methods of consensus building are insufficient for those pages. Arguments including: insufficient outside notice of the RFC; non-neutral wording both in the title of the RFC and how it was advertised; that no counterargument to retain the status-quo was prominently displayed;  how it should be closed, by whom, when, where and what constitutes a 'consensus' were all discussed as the RfC was progressing and closed (twice). The Process was as much an issue as the change (although some may argue that process was used for or against change) and a better process specifically devoted to policy may be a remedy.

Unlike article RFC's where outside sources and references to policy can be used, when attempting to change a policy, there is no outside guidance, the multiple points of view we have regarding should be expressed, and if a large enough proportion of the community comes to accept it, it should be adopted.

Proposal
Each page would be under continuous RfC style review, with major changes being implemented only when a plurality endorse that version. Endorse would be the only choice (positive voting), since it's an endorsement of that version; editors would be allowed to add or remove their endorsement at any time. Any combination of versions could be endorsed (including endorsing multiple conflicting wordings), opposition is unnecessary (endorsing every option except one is effectively an opposition).

This would provide stability in that the current wording will generally have a large number of endorsements, making change unlikely but not impossible. I would suggest applying the inactive standard of administrators to endorsements, although not removed they could be separated for purposes of determining current consensus (any editor with one year of no logged actions would be considered inactive). There would also need to be a probationary when this was first implemented, where no changes are made to the policy (30-60 days) to give the current wording a chance to gain a 'head start' consensus (if it has it). I would suggest a liberal application of snow to doomed changes, for the borderline ones (get over say 50% of the current proposal) I would be inclined to leave it, as long as necessary.(something along the lines of 5%-10% within five days or it snows, then it has 30 or 60 days to get to 50% or it gets failed) Additionally this would provide a venue for significant minority viewpoints on wording, while still having one authoritative version. Endorsements would be limited to registered accounts, comments open to all.

Rationales on the RfC page would be limited to consensus based arguments for change(with no signature), discussion would be directed to the talkpage. The primary method of rebuttal would be in the other rationales, (the status-quo would adapt to new proposals with new arguments in favor of itself, and arguments against new proposals, perennial proposals would of necessity have a rebuttal in the status quo).

Since each proposal has a sub page devoted to it, debate and consensus can occur relevant to each idea, a disadvantage of multiple and changing talk page mitigated by all the failed proposals being archived anyhow, and relevant discussion on the viable or minority viewpoints being retained differentiated by topic.

How this is not a standard RfC

 * The RfC is continually open
 * The plurality version is the current version (subject to some conditions)
 * Proposed amendments at introduced as separate sub-pages, may be introduced at any time (in effect creating a wiki-space for changes, rather than one monolithic talk page)
 * once an active editor has endorsed a version, there is no need to ever return to the page to !vote against any change, however, editors may change endorsements at any time

=Example (Verifiability)= This is the Permanent RFC for Verifiability, a process to ensure that changes to our core polices have consensus. (note that for the example I've only recreated two versions of the lead, in actually the entire policy page would be displayed (hatting might be required for the various sub-subpages because of length).