Revert "Multi-interval set operators"

[rofinn]

Reverts #179

[morris25]

Relevant comment for reversion #179 (comment)

[codecov]

Codecov Report

Merging #191 (7f7fd0e) into master (7a7ec36) will decrease coverage by 2.90%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##           master     #191      +/-   ##
==========================================
- Coverage   84.63%   81.73%   -2.91%     
==========================================
  Files          12       11       -1     
  Lines         794      624     -170     
==========================================
- Hits          672      510     -162     
+ Misses        122      114       -8     

Impacted Files	Coverage Δ
src/Intervals.jl	100.00% <ø> (ø)
src/endpoint.jl	98.11% <ø> (ø)
src/interval.jl	96.36% <100.00%> (+0.34%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 7a7ec36...7f7fd0e. Read the comment docs.

[omus]

FYI the proper way to have handled this would have been to yank version 1.7.0 from the General registry instead of making a 1.7.1 release

[rofinn]

That's what I thought, but @oxinabox said that yanking releases is typically only done for security or compat issues. Honestly, yanking releases appears to be an undocumented black art that I'm happy to avoid :)

[omus]

That's what I thought, but @oxinabox said that yanking releases is typically only done for security or compat issues. Honestly, yanking releases appears to be an undocumented black art that I'm happy to avoid :)

Definitely could be documented better. As the 1.7.0 release is still available it could still be used and break our semver requirements. I can yank 1.7.0 if that sounds good to you. In the future I suggest using yanking for this kind of rollback.

[oxinabox]

As the 1.7.0 release is still available it could still be used and break our semver requirements.

No more or less than any other release that is not the final patch release.
It's fine to yank it, but that exact same logic can be used to yank every patch release but the last.
Which, I have argued, should actually be the behavior of Pkg by default.
Namely that, post v1.0.0, only the last patch release should in installable unless some override flag is passed.