There are clearly good reasons for making PRs against master, so I withdraw my suggestion of a using a branch.
I like 1, 2 and 4 of Dave's list (https://github.com/cf-convention/cf-conventions/issues/130#issuecomment-412947016) but don't see the need for point 3: "branches are used as short and/or long term feature development space.". This, I think, is not relevant to creators of PRs so we don't need to mention it in this document. The same seems true for the management of near release time - we don't need this in CONTRIBUTING.md because it won't affect contributors, who would carry on making PRs to master, regardless of the release mechanics being employed at the time. -- You are receiving this because you commented. Reply to this email directly or view it on GitHub: https://github.com/cf-convention/cf-conventions/issues/130#issuecomment-413121295
