"He finally concluded with some requests to the postgres developer community"

... You're one of the most well funded companies in the world, you shouldn't be asking for features to open aource devs, but you should be opening PRs

That assumes that every problem can be solved with money, and that the money (or the PRs the money funds) will be welcomed by the community. By talking about what OpenAI needs Bohan is hopefully starting a conversation with the community in order to engage with everyone respectfully and to work with them.

Railroading an open source project with money or dev time in order to force it to go in the direction you want is not the right way. Those things should be available if the community asks, but they shouldn't be the opening offer.

Going to PGconf.dev, explaining what you are doing to the core devs, the problems you faced, and what you would love to have as feature in Postgres, is exactly the right way to go about improving Postgres.

Blindly opening "PRs" (Postgres doesn't work in this way, it's diffs on a mailing list) would not get you anywhere very fast. You need buy in and consensus from the wider development team.

Just like with any other OSS project, you have to convince the upstream developers that your PR is worth merging.

You don't do that by throwing PR's over the wall and then moving on. You do that by being part of the community.

That said sometimes you just don't have the resources to engage another community at the moment, so you push the PR over the wall anyway, assume it won't ever land and act accordingly.

The smaller and less impactful the change, the bigger chance of it landing. I'm always clear with my PR's that I push over the wall though: I probably won't be around to maintain this, feel free to not merge, etc. I also try to thank them for their service to the community and share how their code made my life easier.

It does not matter if you open a PR if the project is not open to that solution. You can't just add any random thing to Postgres as an outsider. You first have to convince the people in the project that X is a good thing to add and only after that does it make sense to actually implement it (and even then unlikely that OpenAI has a developer on staff with enough Postgres experience to actually write it. More likely they will then just sponsor a dev from the project).

> ... You're one of the most well funded companies in the world, you shouldn't be asking for features to open aource devs, but you should be opening PRs

PG is really difficult to contribute to because it's such a fast moving target. You get your patches into one commitfest and then you don't get them accepted in time, now you're into the next commitfest, and now you have to rebase across 1,000 commits, lather, rinse, repeat.

Contributing to PG is nearly a full-time job!

I bet it's much easier to find an existing committer to PG and pay them as consultants to do the work you need.

And as siblings point out, you have to figure out what the upstream might be willing to accept, and they might have to tell you that. This requires a conversation. Presenting to them is a way to start that conversation.

that's not how that works.

Money doesn't mean "I built whatever I want and Postgres will evolve into whatever I want by pushing my code". They still need to align and plan. Sure, they'll build the things and contribute, but they don't own it and still need to accomodate the wishes of the project