Nice write up. I already was vaguely aware of the amount of rigor around avionics development which is why I was surprised to see how fast this update was rolled out--that with all that rigor they still got the change out quickly. Hopefully someone thought about the relative risks of a quick update introducing an actual bug vs. a repeat of a (theoretical bit flip).

My typical Emacs session has 3-4 Shell Mode buffers, each typically with tens of thousands of lines of input and output (often spanning weeks or months of work). I watch co-workers running in xterms (or equivalent), running commands repeatedly to see their output or using "more" (or "less"), or not catching or remembering some error message or whatever and I just cringe internally. I want to scream "It's 2016--you can have a searchable record of more than the last screenful of output!". But I don't :-) I explicitly try to extend my Emacs skills periodically and it's had great pay-off. Earlier this year a co-worker introduced me to Magit (a Git UI that runs inside Emacs). It took a little while to get the hang of it, but it's been a life-changer. Check it out. -- Happy Emacs user since 1980.

The "suspend" command is like doing ^Z to shell you're in so you can pop up to the shell that invoked it. E.g.,

$ su smith
$ ... do some commands as smith ...
$ suspend
$ ... do some commands as yourself ...
$ fg
$ ... do some commands as smith again ...