A lot of packages depend on hashable, so anybody wanting to use the new bytestring will have to override hashable (either via cabal.project or stack or nix). hashable has 892 direct reverse dependencies on hackage, bytestring has 5922. They are both probably depended on in almost every production codebase out there. Because of that, I think it would be best to make a revision now instead of waiting.
bytestring
release hit 0.11.x, so the bound needs to be bumped from< 0.11
to< 0.12
Maintainer was contacted on this issue: https://github.com/tibbe/hashable/pull/191 (October 11th)
A lot of packages depend on
hashable
, so anybody wanting to use the new bytestring will have to override hashable (either via cabal.project or stack or nix).hashable
has 892 direct reverse dependencies on hackage, bytestring has 5922. They are both probably depended on in almost every production codebase out there. Because of that, I think it would be best to make a revision now instead of waiting.