)]}'
{"/PATCHSET_LEVEL":[{"author":{"_account_id":11816,"name":"mike bayer","display_name":"Mike Bayer","email":"mike_mp@zzzcomputing.com","username":"zzzeek","status":"Red Hat"},"change_message_id":"d15cb3025b9ce55c10a7dd857de810f33476dc93","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"b7fa4c38_57e2feb2","updated":"2026-01-05 13:08:10.000000000","message":"I guess they recommend against psycopg2-binary because your production systems would have some specific version of libpq you\u0027d want to build against, or something, but this is a very vague guideline on their site which is lacking a clear rationale.\n\nthat said, yes, psycopg3 (which is called \"psycopg\") is where the development is going on.\n\n\nbigger picture though remind me if oslo.db is actually used in practice with a non-MySQL database?   I\u0027m sure there were \"drop postgresql\" conversations that ended with \"no, because....\" somewhere.","commit_id":"181f4d34516d72a01b21225eb8baa937eba7763c"},{"author":{"_account_id":15334,"name":"Stephen Finucane","display_name":"stephenfin","email":"stephenfin@redhat.com","username":"sfinucan"},"change_message_id":"3f5f5e910c5dd1e8e303a326a074fac6e34d6b4c","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"8d4f4e46_8f02e3b5","updated":"2026-01-05 10:42:52.000000000","message":"This probably warrants a discussion on the mailing list since it\u0027s used in a few places. I was under the impression that the `-binary` package was a binary wheel (since there are no binary wheels published for the `psycopg2` package are [Windows-only](https://pypi.org/project/psycopg2/#files)) but it seems I was mistaken 😕\n\nI also did not know that [psycopg 3.x is a thing now](https://pypi.org/project/psycopg/). We may wish to pivot to that altogether.","commit_id":"181f4d34516d72a01b21225eb8baa937eba7763c"},{"author":{"_account_id":9816,"name":"Takashi Kajinami","email":"kajinamit@oss.nttdata.com","username":"kajinamit"},"change_message_id":"23b60b4f66f203aba5d1fb853b4cdaac75687075","unresolved":true,"context_lines":[],"source_content_type":"","patch_set":1,"id":"d12f2236_71ab2465","in_reply_to":"b7fa4c38_57e2feb2","updated":"2026-07-12 06:47:37.000000000","message":"I sent https://lists.openstack.org/archives/list/openstack-discuss@lists.openstack.org/thread/IUVWTGY5IYWT5PNRSG2MPDDX56O6F452/ to ml to gather input about PostgreSQL support. I don\u0027t expect many users are using it nowadays due to no PostgreSQL support by neutron, though.\n\nI found the current status a bit wired because we use -binary as a runtime dependency while we use non-binary for test requirements. Although the context behind that recommendation by psycopg2 community is unclear, the current set up totally goes to the opposite direction.","commit_id":"181f4d34516d72a01b21225eb8baa937eba7763c"},{"author":{"_account_id":11816,"name":"mike bayer","display_name":"Mike Bayer","email":"mike_mp@zzzcomputing.com","username":"zzzeek","status":"Red Hat"},"change_message_id":"2337a01af3f97cca4984d7ec7d8e523c572dbfc0","unresolved":true,"context_lines":[],"source_content_type":"","patch_set":1,"id":"919cb13c_83eeb1aa","in_reply_to":"d12f2236_71ab2465","updated":"2026-07-13 18:35:24.000000000","message":"how do we go about removing it such that anyone that actually cares would wake up and notice at an appropriate time?  \n\nI would say either remove it or upgrade to psycopg3.   the transition from 2 to 3 (ha) is not terribly non-straightforward but the two drivers definitely do things differently and there would likely be hiccups.","commit_id":"181f4d34516d72a01b21225eb8baa937eba7763c"}]}