Just because Oracle sells into companies that could just as well have used postgres that doesn't mean that Oracle aren't producing anything of value,..
Oracle doesn't sell software. They sell risk reduction. Someone to sue or blame.
Once you understand that, you see how brilliantly they've achieved product/market fit in a way many people here can only dream of.
Sure some people buy Oracle for risk reduction after having done a sophisticated analysis, but plenty of people buy it because that's what the nice sales person tells them they need and the sales person obviously understand this techy stuff far more than they do.
The dev team says they can just use this "postgres" but the sales person said he'd seen this situation a lot of times before and that dev team was too inexperienced with "serious software" to know they needed to use a proper database rather than a toy one that people use when learning because it's free.
(not to pick on Oracle for any particular reason; this is common in enterprise sales at all levels)
Have you ever read an eula? Have you ever heard of someone winning a lawsuit against a software vendor? Have you even heard of someone getting any kind of monetary compensation for mishaps caused by software bugs?
It's not about the lawsuits, but it is very much about risk reduction. This applies to all consultancies, not just Oracle.
If you hand-roll your own solution for (almost) free using open source components, you're the one who gets fired when it goes down.
If you buy an enterprise-level solution from Vendor XYZ, with a requisite expensive support contract, they fucked up when it goes down, and you are safe because goodness, who'd have thought an internationally renowned firm like Vendor XYZ would ever break? Besides, thanks to our expensive support contract, they're on top of it.
Microsoft, Oracle, Accenture, they all share one thing in common: their product more CYA for middle managers than anything else.
I have similar thoughts about management consulting - it's CYA for high-level execs. You rubber-stamp a risky move with a respected firm so that if the shit hits the fan the firm (and yourself) are insulated from the backlash.
Yes. Just to finish off the argument for you (I think you already get this, just forgot to explicitly state it), not only would the programmer have to assume the risk in choosing postgres over Oracle, but she would typically also not get any share of the thousands of dollars she saved the company by so doing. It's all downside and no upside, so it's no surprise that many people opt for the vendor solution.
Oracle doesn't sell software. They sell risk reduction. Someone to sue or blame.
Once you understand that, you see how brilliantly they've achieved product/market fit in a way many people here can only dream of.