Where do evolution proposals discuss back-deployment strategy?

[vanvoorden (Rick van Voorden)]

I'm putting together a new proposal that might introduce a backDeployed API. What I'm not completely clear on is to what extent marking an API as backDeployed is discussed during a "proposal" review as opposed to a "diff implementation" review.

One potential example in favor of putting backDeployed in the proposal itself is SE-0388 which does explicitly mention a new backDeployed API.

AFAIK there still is no explicit guidance on where a backDeployed API discussion belongs. Some comments from the review thread on this topic:

@xwu

Since platform owners and not the Swift Evolution process control what is back-deployed, I would be very hesitant to compromise on the proposed design of a feature on the basis of what can be back-deployed. It would set a very poor precedent to contort the design of permanent additions to the standard library on the basis of transitional bring-up matters that are not even within the full control of the Swift project.

And this:

@Ben_Cohen

Back deployment isn't something the language group generally comments on – that is a decision for the platform owner not evolution. But since this doesn't declare any new types, it can probably make use of the currently-in-review @backDeploy feature.

And here is another comment on this topic:

@Douglas_Gregor

The Core Team discussed the issue, and their view aligns closely with the above: backward deployment is a vendor-specific issue, and not part of the evolution process. However, calling out the runtime impact of a proposal can help us understand concerns around back-deployment and also the overall impact of a feature on the language and runtime, which is valuable.

Which all generally seems to suggest to me that language evolution proposals do not need to take an explicit opinion on backDeployed. Is that correct? A discussion about backDeployed would then take place in the diff implementation reviews on swift repo through the RFC diffs. Is that correct?

[Jumhyn (Frederick Kellison-Linn)]

Yes, that’s correct—back deployment concerns are mostly out of scope for evolution. It can sometimes be helpful to note when specific design choices actively frustrate any sort of back deployment, but it is generally more of an implementation question.