Since I admitted myself that I feel old, I guess it’s my solemn duty to yell at clouds time from time.
I consider satire to be the most realistic depiction of the world (unless it’s a pasquinade) and as Shakespeare put a phrase in a mouth of one of the King Lear characters, jesters do oft prove prophets.
For example, I still like to re-read various satirical pieces by Ilf and Petrov (probably the best Soviet satirists in the first half of XXth century) and one of those pieces gave a title to this post because of its similarity to what I want to talk about.
That short story mentions a game “loading a ship” (here’s the title) where a group of bored people tries to “load a ship” by calling things starting with the same pre-defined letter—lamps, Lilliputians, locomotives, liquors etc etc—until people can’t recall any more things starting with that letter or somebody suggest a name and people start to argue if that should be allowed. The story itself is about one man who metaphorically loaded a ship by inventing new and new social activities until at some point it was discovered that nobody remembers what are his direct duties should be and he was fired.
And here we transition to the Linux Foundation. Previously I thought it should promote Linux adoption, make some Linux-related standards and pay money to main Linux developers and maintainers so they can work on it full-time. But news in last couple of months showed me I was wrong.
First there was a piece of news about AgStack (Linux for Agriculture). Then there was a piece of news about Linux Foundation organising some unrelated initiative for Microsoft (while not participating in it itself). And finally there’s a piece of news about Open Voice Network, an initiative for ethical standards of voice assistants.
This made me wonder not just why I encounter such news but also what does the foundation really do. The answer did not make me more optimistic.
So it seems that Linux Foundation is transitioning from a foundation that does what I expected it to do to an organisation that does IaaS (initiative-as-a-service): you pay them and they prepare all required infrastructure to have it all running, just invite some organisations to participate. Beside that what are those member companies are paying for? I don’t know the official explanation but to me it looks like three major categories to put it very bluntly and impolitely: bribes, extortion money and indulgences. Bribes is what you pay to affect the politics of the foundation in a way favourable to you (maybe it’s called lobbying, maybe such things do not happen at all—give me facts and I’ll amend this post). Extortion money is what you have to pay to participate in various standardising activity (aka membership fees, no different from any other standardising activity). Indulgences are money you pay to avert attention from your GPL violations regarding the kernel sources (just search for GPL violators and Linux Foundation, you’ll find not just Chinese but American companies mentioned there; at least in one case the known GPL violator hasn’t published its modified kernel sources after becoming a member).
I could not find any reports about the foundation beside their 2020 annual report so I can only speculate how it has changed in the past regarding income, spending and stuff. Yet I can foresee three scenarios on how it may develop in the future:
- Shifting to a different activity. Even now actual Linux-related things seem to be less than a half of what the foundation does nowadays, so maybe in the future it will realise that Linux is a legacy they don’t care any longer and maybe even rename themselves to reflect their new main occupation;
- Fossilising. Linux Foundation may exist in the future but both its activity and being a member will become a tradition that nobody understands but they’ll keep doing it because it’s a tradition;
- Withering. While Linux is relatively popular kernel, nobody can guarantee it will remain popular in the future. Linus himself is not immortal and his successor may be not as skilled, we have IBM trying to replace Linux kernel with systemd while keeping the name—or maybe some other reason will hurt Linux popularity. Or some other kernel will replace it in its major niches (I can easily imagine Android being rebased on Fuchsia; the same may happen to servers as well). In result Linux and Linux-related things will become not interesting to most companies and they’ll drop their financial support.
You may say that there might be a scenario where Linux Foundation will concentrate on Linux-only stuff but I’m sceptical. AgStack is a clear signal they run out of ideas where to promote Linux but in accordance with Parkinson’s law they’ll still keep growing by inventing new stuff as long as there’s enough income to employ more people, organise more conferences and such.
As usual, I’ll be happy to be proved wrong.
1 and 3 are already happening
1. https://github.com/servo/servo are hosted by Linux foundation. Don’t forget their women outreach program
3. Linux will be on top for a while, and even though there will be differing visions for what Linux will look like, I think it will stay relatively stable in its development progress. You personally know first hand how open source forks and disagreements work….
I mean research into the BSDs are always ongoing. Once bhyve and the BSD hypervisor and virtualizer stack is mostly feature complete we will see alot more BSD running servers since many Linux servers are just containers for other operating systems. I do think there is more research to be done in systems research (Redox OS, MinuetOS, TempleOS, etc), but applying that research to existing OSes are easier than making a whole new one.
But again, the Linux Foundation has been a joke for a while. They produced their 2020 annual report on a MAC, even though there is free software that could have made their doucment. (Details here: https://www.youtube.com/watch?v=a-2dYfYvJGk)
Well, outcome 1 is the most logical in the current situation. And while writing the post I looked into their 2020 annual report (the only annual report I could find) and it tried hard to reinforce that opinion.
As for Linux development, it’s no secret it has been mostly done by corporate engineers nowadays. And some people claim that Linus has been convinced to change his behaviour recently, so maybe Linux kernel development will turn into faceless corporate product (and stable too). But as long as they don’t break the stability of kernel interfaces it should be fine.
Different OSes look promising, I’d probably switch to RedoxOS if it becomes usable. Adding new features to an already existing OS might be easier but you risk ending up with something monstrous like C++ or Windows 10 (with its eclectic GUI and WSL2).
As for the last point, that’s what you get with an organisation run by professional bureaucrats. IIRC you had similar stories about RedHat and where is it now?