Accidental reboot broke XWindows when I was field-netbooking

The previous blog post (Soap in PKP Intercity trains is quite salty), while finished when still on the train, has been left with my inability to immediately publish it because networkmanager didn't like WPA2-PSK again. I ended up trying to grab html as qr code, then gzip of it as qr code, then finally managed to display a base64 of a gzip of it as a qr code but then i grabbed the netbook by the screen, which has the power button, and ended up pressing it which made it power off, power off instead of hibernating like i kept doing recently.

Because I was already walking around with the netbook in a rush, it supported on a bag, whilst heading / searching for the venue I was to meet with team colleagues in.

I ended up with not having ran mkinitrd (and subsequently lilo) at last kernel update, fortunately they did the right default thing without parameters.

But I had to reach that. At first I thought it was just a mouse issue and presumed my WindowMaker setup to not have had any bindings that would let me start doing things without a mouse. But keyboard was not reacting to console switching key combinations. Took me a while to realize it wasn't reacting to anything.

ACPI keypresses were not handled either. So I got stuck in a loop of force-poweringoff the stuff for each iteration of trying to do something then startx-ing and being stuck.

I kept blaming my xorg.conf stuff wherever it was to be, but it seemed I've done none. But still, I hoped for `X -configure` to help. And that's how I've noticed failed "i915" modprobes. Whilst already having thought about wanting to nixos-infect maybe, and realizing there is no wireless network device. And then I noticed the versions in /lib/modules/, the respective directory names, were not matching. And that's how I ended up realizing I need to mkinitrd and lilo. Just took a risk and let them do their default params, luckily they write their files where they should go alright.

And so I could reboot and enjoy things again. Aaand I should remap acpi power button to hibernate too later maybe. Or to suspend. With lid action already being hibernation.