Conversation
|
I can't see anything risky in changelog: https://www.openssl.org/news/cl31.txt |
|
According to https://www.openssl.org/policies/releasestrat.html: So maybe we actually want to stick with 3.0.x? |
|
I don't think we should stick with it. Maybe we should keep it around, but I'm not sure yet who its consumers might be. Let's wait until someone complains? A version we track is something we have to clean up at some point, so the fewer things we track the better. |
|
@ofborg eval |
|
Well, My argument for sticking with 3.0.x boils down to "it's an LTS release". Apparently that gives us 1.5 more years to switch to the next minor (or mayor) release. |
Agreed, there shouldn't be a need for both 3.0 and 3.1.
The failure seems to be with |
|
Personally, I intend to keep OpenSSL on LTS releases, because even updating it to those is more than enough effort (see e.g. #150093 and #210452). If anyone sees real value in being on 3.1 (or any other non-LTS release) instead of 3.0 (or a future LTS release) and is willing to put in the work for this to actually work in nixpkgs… well, they can go ahead and do that, but tbh I'll probably stop trying to work on OpenSSL in nixpkgs, because those efforts will probably just end up like what lead to #215109. |
Description of changes
https://www.openssl.org/news/openssl-3.1-notes.html
Things done
sandbox = trueset innix.conf? (See Nix manual)nix-shell -p nixpkgs-review --run "nixpkgs-review rev HEAD". Note: all changes have to be committed, also see nixpkgs-review usage./result/bin/)