## Summary For two independent MRs (!15770, https://gitlab.haskell.org/ghc/ghc/-/merge_requests/15740 ) in the 9.12.4 release candidate cycle the question surfaced if we could include them into the final release. Concretely this raises the questions: * Should we be willing to do another rc cycle (delaying releases) for fixes to non tier 1 platforms? Does the severity of breakage on those platforms matter? * Should we be willing to include "low risk" MRs without doing an additional release candidate cycle? * Or should we simply push back such changes to the next minor release. Should any of this depend on how "urgent" the next minor release is? At this moment in time I'm leaning towards following a general policy of: * Do not delay a minor releases for non-T1 platforms *unless* it's expected to be the last minor release for a series. * Do not add new code change commit without doing a new rc even in general. However I do wonder if it's worth making an exception for !15770 as it doesn't impact non-wasm platforms *at all*. ## Environment * GHC version used: Optional: * Operating System: * System Architecture: