What amazes me is that they (and a number of other projects) do it without any traditional “management.” Sure, there might be a release manager, but it’s more like a coordinator/comptroller function. It’s just a bunch of (mostly)”let’s act like adults” sort of people, and without all of the bureaucracy that usually accompanies some of even the smallest software efforts, it just keeps going.
It's inspirational, to say the least. Linux is a testament that a truly distributed team can not only work efficiently, but deliver a successful product.
I'm inclined to think that it's partly because they're not chasing growth metrics, have to report to stakeholders, etc. Most companies drown in the overhead related to such processes.
I don’t disagree but there are many corporate contributors who are getting paid based on the success of the optimizations and features they upstream to the kernel. At FAANG it’s the same metrics-based success you refer to.
Sure, but if any one of those corporate contributors could merge any change they wanted into the kernel they could likely get way more "metrics" gain, but they can't, because that's not the goal of the kernel team. The gains of one corporation are secondary to the gains of everyone.
Those corporations work together to contribute to the same open source code base and at the same time they are trying to (commercially) kill each other. Quite a wonder.
Good point. But there’s not one such paradigm ruling the whole process of developing Linux. I guess we actually could say there are competing paradigms.
What amazes me is that they (and a number of other projects) do it without any traditional “management.” Sure, there might be a release manager, but it’s more like a coordinator/comptroller function. It’s just a bunch of (mostly)”let’s act like adults” sort of people, and without all of the bureaucracy that usually accompanies some of even the smallest software efforts, it just keeps going.