A criticism of the "grow at all costs" mindset: they aren't even executing on that efficiently. Per their GitHub account, they have 162 open source libraries. That's 162 open source solutions they could have found elsewhere and moved on to more pressing business needs.

Perhaps the thing didn’t exist and they had to create it? Perhaps the open source thing didn’t fit their needs?

Perhaps not perfect substitutions, but business acumen is seeing a solution that's 90% of what you're looking for is more than adequate. It seems unlikely the organization identified 162 truly novel problems. It appears some needless reinventing of the wheel may have occurred. It's a common problem that certainly isn't exclusive to Uber.

https://en.wikipedia.org/wiki/Not_invented_here

Licensing can also be an issue sometimes.

Tell me about it. Oracle wants a pound of flesh for a few licenses, and Windows, VMware, and others can be onerous (though I'll concede Windows is getting better)

How many of those libraries are even used by Uber? While I think open source is intrinsically valuable, it feels like a lot of work out there is innovation theater by those with extra VC cash.

There’s very little downside. Spend some money from wealthy people and institutions, get some engineers paid, get some code written. The only losers are the investors who, per their job description, take the risk. Perhaps simplified but I like companies spending money on this stuff.

Risk is part of investment of course, but isn't the assumption the money will be spent in ways that focuses on return? (I know the answer being a developer, but taking the broader perspective)

So they shouldn't have open sourced?

I think the point is that they shouldn't have written the code in the first place. Although if it is innovation theater, then yeah, maybe they shouldn't release it, either. A real open-source project is a long-term commitment to a community, not a PR checklist item.

Yeah, I have to wonder how much of that is Promotion-Driven Development. A few contributions to an existing package is way less impressive than writing a whole new thing from scratch. In case others are curious, here's the list:

https://github.com/uber

I'm sure some of those were necessary. But I have a hard time believing all of them were.

I once saw some article about them writing their own rich chat client for their internal tooling because their existing solution wouldn't "uber scale" - whatever that meant.

That reminds me of the delightful "MongoDB is web scale" spoof video from years ago: https://www.youtube.com/watch?v=b2F-DItXtZs

Clearly Uber scale is one level up!