Some next steps for Codex, ordered by priority:
- db/yugabyte - working out the kinks of switching from cockroachdb to yugabyte
- long polling - adding a long poll endpoint to gungnir
- connection grouping - use the connection information to group events so that we can provide a more accurate status endpoint in gungnir (requires XMiDT work: https://github.com/xmidt-org/talaria/issues/111)
Does this sound like a good plan? Is there anything else we may want to do for Codex in the near future?