Oleg Grenrus

I think it's not worthy copying `Wrapped`. Firstly, it's not just newtype unwrapper, there is ```haskell instance Wrapped (Seq a) where type Unwrapped (Seq a) = [a] ``` for example....

@nh2 > In the future, we may want to create such places so that one can make changes to syscall implementation behaviour without a proposal. - step 1: create a...

> However, `FD` is indeed the only `IODevice` instance, so we don't get much info on what other devices could exist, to determine where seekability should live. Not true. On...

> Non-experimental packages are actively discouraged to use `ghc-internal` Yes, so the needed functionality, once identified, can be re-exported from `ghc-experimental`, so `base` can stay away from GHC specifics.

> We just need to get this right with regard to — say — excluding autogenerated modules You don't need to exclude them. Autogenerated modules will also have files on...

@Bodigrim based on https://github.com/haskell/hackage-server/issues/1351 it's not enough to just use the newer library version.

Yes, but we are careful not to change error messages in minor releases, as something (usually tests) may break somewhere. But alone spelling errors don't justify a major release, so...

https://github.com/well-typed/optics/pull/312#issuecomment-612912594 I also explained this kind of out-sync situation in https://github.com/well-typed/optics/pull/312#issuecomment-612919951 but neither you nor @adamgundry have responded to that, so I'm not sure what we agreed to do, if...

> import stanza retention in GenericPackageDescription ( Please don't. A lot of code wants an elaborated (= stripped down of syntactic convenience) representation of package description, and GPD serves that...

This PR does add ```haskell , exactComments :: ExactComments Position ``` field to GPD. I don't see a point of having comments in GPD. (As noted in tests, it "breaks"...