The main problem is finding that useful subset that is A) useful to everyone and B) at a level that most find the distribution of information useful. Personall

The main problem is finding that useful subset that is A) useful to everyone and B) at a level that most find the distribution of information useful. Personally I find there are two ways to handle it. The first is to focus on work flows. As in how to do something a person on your team does regularly really efficiently. The second is to take a common tool and go in depth on it. Yes you end up talking about ls switches for 20 min but it should also focus on how this tool integrates in to the larger unix system. The last thing to pay attention to is documenting your sessions. I use find about once a year, I never remember how to use it and spend a good 20 min looking at the man page. If I had in depth training on grep and find then I would be annoyed that I was back at the man page the next time I needed it because I didn't have a useful reference.