Closed hernot closed 2 years ago
@1313e and @hernot -- where are we on this? Ready to pull the trigger?
Yeah, I am fine with this merging.
@hernot double checking you're done with the dill cleanout and ready for the big merge?
yes I'm done
1) hickle 5. Files are dumped without dill 2) legacy loading of hickle <= 4.x files still needs dill. therefore dill is still in the requirements for legacy loading but indicated that only need for legacy loading in README.md can be removed any time from requrements.txt but must stay in requirements_test.txt as long as there is the need for legacy loading hickle <= 4.x Files
Ok here we go as agreed discussing pr #138 here the fully assembled hickle-5-RC branch. In case you prefer to directly review within my repo feel free to close this pr again. I'm fine with whatever you prefer. Not sure if all commits reported below are really visible from within this branch or whether the log below is an exceprt of the git reflog.
I also expcet that appveyor and travis to complain a bit due to the tox related changes would be surprised if things would work on the first try. Anyway i cross fingers
Reason for python 3.5 failure known (cant test here any more lacking Python3.5). Fixing if support of Python 3.5 should be kept beyond hickle 4. Is easy just one change in one line.
Fixing fail in Pyhton 3.5 is one change in one line will do if support for Python3.5, which i have no means any more to test locally here) shall be supported beyond hickle 4. No problem at all.
With the astropy warnings i would need your help as we do not use Astropy here so i have not any clue how to fix.