We've recently clarified status for CFPs (See #42).

While attempting to apply this standard to existing proposals, I noticed that there may be interest in merging CFPs when they are not considered "Implementable", but the current process largely assumes that this is the minimum bar to merge.

More specifically, there are a couple of cases I'm considering:

Dormant Status

These may require subsequent discussion to identify and resolve key questions prior to marking the CFP as "Implementable". I've already merged #7 under this category, and I think #36 also applies under this category. The status text does not currently really describe one way or another whether such PRs should be merged, but it seems like a reasonable way to "close the loop" on an open PR. The alternative I see would be to just close dormant CFP proposals, and if we wish to adopt them in future then someone would have to resurrect the PR afresh.

Draft Status

The current text prescribes that one of the properties of a draft is that it's not merged. It also describes "Implementable" PRs as being merged. This means that a specific idea needs to be significantly matured before it is merged into the repository. I noticed in #32 that contributors have been managing review discussions in their own public repositories in order to more easily iterate on the discussion. As a pattern, this can work, but it also draws the discussion away from the central official repositories, making them generally less visible to the overall community.

Furthermore when I reviewed the above PR, I had dozens of resulting comments. With so much activity, the GitHub UI tends to begin to inhibit healthy conversation by being slow, making it hard to find comments, and so on. This makes me wonder whether we should redefine the "Draft" status such that we could merge proposals in draft status as long is the core idea has high level agreemeent. We could set the culture to expect "TODOs" to be documented in a draft CFP before merge. Then, before the proposal can graduate to to "Implementable" we would need to ensure that key questions have reasonable solutions and have no major flaws and no significant lack of consensus.

Overall with this proposal, if we had community members working closely together, I could imagine staging the development of a CFP where an initial version just describes the summary and goals to establish high level consensus on the core idea, then subsequent PRs could develop the idea, each of which could undergo some degree of review to merge, then as key ideas/questions/impacts are introduced, even those could be separately committed, and only once the CFP gets to a point where there is rough consensus and committers believe it has been sufficiently explored, then a subsequent PR could move it to "Implementable". While I don't know for sure whether contributors are looking for this degree of incremental progress on CFPs given that CFPs are often written proposed by a single individual, I would be open to exploring mechanisms like this for more incremental progress.