Why the major version bump for v1.13?
sjakobi opened this issue · comments
The changelog doesn't indicate any breaking changes. Why didn't you go with 1.12.1 for the version number?
I probably could have. Is it causing you problems?
It's not exactly a problem, it's just a bit of work updating library bounds. On https://packdeps.haskellers.com/reverse/time I currently count 32 packages with bounds <1.13
where maintainers would need to do the same work.
This would be avoidable if you'd re-release v1.13 as v1.12.1 and deprecate v1.13.
I missed something anyway in 1.13 so I may do this.
I've deprecated 1.13. 1.12.1 will be up this weekend.
I made a revision to blacklist it completely, https://hackage.haskell.org/package/time-1.13/revisions/
So that if someone relies on time-1.12.1
features bounds like time >=1.12.2 && <1.15
won't cause problems where 1.12.2
adds something which isn't in time-1.13
.
I.e. now it looks like (if you squint) that time-1.13
never happened, there is no way to have it in the install plan.
Thanks a lot, @AshleyYakeley! :)