After yet another conversation along the lines of “no, seriously, are we better off moving Open edX from MySQL to PostgreSQL at this point?”, I wrote up https://openedx.atlassian.net/wiki/spaces/AC/pages/3801743364/MySQL+vs+PostgreSQL to start capturing all the relevant pros and cons of such a change. If anyone else has points they want to make in favor of adding PostgreSQL support and/or sticking with just MySQL, please add them there so they can inform the inevitable future conversations on the topic and possibly even trigger action to change the status quo.
Additionally, I put that document under a new https://openedx.atlassian.net/wiki/spaces/AC/pages/3810459654/Big+Technology+Changes area for capturing and tracking the pros and cons of other serious proposals to change some significant aspect of the Open edX technology stack. If you think there’s a solid case for adding types to all our Python code, switching some things from Django to FastAPI, starting to write things in Rust, replacing Elasticsearch with Meilisearch, etc., feel free to add a page and start laying out the cost/benefit analysis.
The bar for making such big changes to the technology stack is pretty high, but the pace of change in the software ecosystem demands that we do occasionally make them, and it’s important to understand the relative value and cost of pursuing each proposed change.