On the always good
https://kottke.org/ the other day, there was a link to a series of videos of professional drummers hearing songs for the first time, without the drum track, and then playing what they think would fit. Here's one of them,
https://www.youtube.com/watch?v=tbUYVcaF_l0 and while I'm not a drummer, the best part of this is how Dirk Verbeuren explains the process of listening to the song, and how to learn something never seen before and how to adapt to it. The whole series is recommended:
https://www.youtube.com/playlist?list=PLThYwnIoLwyWiF5RgHPzOzYNdqQw1-tepAnyway, got me to thinking, a while ago a friend (also a kernel developer) had the idea of a presentation for the yearly Kernel Recipes conference that would be him sending me a patch, me reviewing it, talking about how it is reviewed, and the back and forth between us on getting it into a mergable state. That process is one that someone else recently asked me "what presentations can you recommend for new kernel developers to explain how this review process works" and I didn't have any suggestions, but it really is an important thing that is not taught at all in school, or in any company that I know of.
So maybe, a series of videos, or talks, where a maintainer gets a patch series and walks through how they review it, what they look for, what they expect, and how its tested (if at all), might be interesting? Or "here's a reported bug, how do you debug it?" type of presentation to put a developer through the steps of attacking something unknown, and figuring out what the problem is, and how it could possibly be solved.
Brings me back to the days of the Plumbers conference session where we had "bring us your laptop and we will get suspend/resume to work on it" tracks that ended up being a lot of fun for everyone involved as crazy hardware/bios issues were debugged live.
Would this even be interesting? I think it might need a lot of good editing, you don't want to see me staring numly at a terminal window for a few hours while reading lots of inscrutable driver code tracing it to find a bug, that would put everyone to sleep...