[benologist]

Yes. You can tell because in addition to their lines of business they have these massive cash sinks tinkering away on random stuff. Why does Uber need to hire people to write a tool to eliminate null pointer exceptions in Java? There really might be a sustainable business - but it won't be doing most of this stuff:

https://uber.github.io

https://lyft.github.io

[mandevil]

I have been in companies that open-sourced small bits like this. Basically, it worked like this: someone came up with a cool tool that helped them (in this particular case, likely finding NPE's in their Android app). Then others heard about it and started to use it. Once the use spread a bit, the developer who built it tried to clean up the code and document it better and make it easy for others within the company to use, and then someone said "let's open source this" because that way it looks great for the company ('see, we're giving back, we're cool- more developers will want to work here') and the original engineer (who now has a open source project on his resume, also probably gets to go to a conference and give a talk on it, which the company will gladly pay for because it makes them look cooler and attract more developers). The LoE once you've built the tool is pretty small, and 'finding things that were causing crashes in our Android app' clearly has a direct bottom line impact on Uber.

But it isn't the crown jewels of the company: their data, their scheduling algorithms, anything like that, its little side projects. (Some companies do open-source the crown jewels and make their money on support contracts, but that's a totally different business model; outside of Red Hat I'm not sure how sustainable such businesses are.)