Update CHANGES #162
Update CHANGES #162
Conversation
|
Whoops. I think it's possible to yank a version from pypi, but I don't know how it's done. That might be the best thing, and then we can cut 2.6.0? Want to poke around and see how easy it is to yank a version? |
|
Interesting, this blog post suggest it's pretty easy: https://snarky.ca/what-to-do-when-you-botch-a-release-on-pypi/ The "yanked" release is not removed from pypi, but it is suppressed so |
|
Yeah, that is easy. I'll publish a new version and track the bad one on Monday. |
|
Matt, I was about to yank the bad release, but I got scared that we now rely on it in CL, so maybe before we yank it, we need to release 2.6.0. Do you want to tune up this PR for a 2.6.0 release? Then I can merge it, cut the release, get it on Pypi, and then yank 2.5.5 (once 2.6.0 is out)? |
|
Yes, sounds good, I will do that tomorrow morning! |
mattdahl
force-pushed
the
update-changeme
branch
from
0bacc8b to
18aaa64
Compare
|
Okay, I moved all the previous changes (and the ones newly memorialized here) into a new Let me know if you need anything else! |
|
OK, this is merged. @ERosendo, can you cut version 2.6.0 of eyecite, and then use Bill's new Github Action to upgrade CL's dependencies to use it? |
|
Thank you Matt! |
I just noticed that @flooie did a couple of new releases a few weeks ago (dd9472c and 2cbe140). Unfortunately, due to my own negligence, I never updated the CHANGES file to reflect my hashing changes in #155.
This PR memorializes those changes in the CHANGES file (and also adds a line about the changed Python support) by backdating them into version
2.5.4. This might be fine, but it also puts us in the slightly awkward position of having pushed a "patch" version that actually contains some more fundamental changes. Ideally, I think it would have been best to have released version2.6.0. (To be clear, this is my own fault because I never put the changes in the CHANGES.) However, at this point, I'm not sure the best way to proceed? One option is to just merge this PR as-is; the other option would be to also push a minor version update for clarity, though the changes are already baked-in to the previous patch version?