The Current Challenges With Using Linux On Airplanes

I'm not really going to address the speaker directly since after reading NSF forums for a few years, I'm convinced aerospace engineers can devolve any innocent or academic discussion into 4chan levels at rates exceeding the speed of light. Of note: the speaker doesn't speak to anything specific that is being worked on to address issues, and only addresses "linux" as a whole, which is about as useful as addressing SVR4 as a whole.

I will address the blog writer as not being particularly diligent in filling that gap, though. Here's a few links of what's going on in that realm since there's people here of all walks and ages:

Wind River RTLinux

Zephyr Project

OSADL SIL2LinuxMP

Xenomai

I will admit reading the reasons against Linux made me a bit butthurt given how much it has accomplished and the ubiquity of it running on servers that host our services.

However, a (real time operating system) RTOS is distinct from a time-sharing operating system, such as Unix, which manages the sharing of system resources with a scheduler, data buffers, or fixed task prioritization in a multitasking or multiprogramming environment. And a time-sharing kernel is likely not suitable for the demands of real-time feedback that airplanes, especially fighter planes, are under. Admittingly I don't work on the kernels so I might be out of my league talking about it if there are ways to optimize them for RTOS applications. And I also don't work on airplane but hobby learning about them.

In an airplane, the pilots are voting members. If they they are nose diving and tell the airplane to pull up, the airplane will calculate how much of the full range of motion to allow the flaps to "bite" the air so that the plane doesn't desintigrate. This takes into account speed, altitude, air density, load on wings, load on frame, G forces pilots would experience, etc etc.

It should be added that they likely talk about the controller of the airplanes. Afaik there already is Linux on airplanes, but to handle e.g. the entertainment system. Which seems like a better fit.

Having a non-Linux main controller OS with specific design and a lot smaller feature set than Linux flying the plane seems like a good idea. But maybe this could also be open sourced and standardized?