My decision making process is not coherent.
My decision making process is not coherent.
I have no justification for this largess. I just wanted one or two spares to mess with because I have five in production and sometimes I need to FA to FO. I could have spent $40 on one machine but I chose to spend $160 on 16 machines.
I haven't tested any yet but they all look to be in good physical shape and all appear to have their (incredibly rare) wifi cards. If they all work I can plausibly part out half the wifi cards for what I paid for the lot.
@webology I have a draft blog post somewhere. Basically I use them in homeprod for talking to zwave and zigbee and serial ports of a few things around the house. One of them is also a local gps-driven stratum0 NTP server.
I'm probably going to mess around with HA voice remote. Maybe k8s. Maybe a family dashboard kind of thing in the kitchen or mudroom.
Sigh. This was probably actually a mistake. I have had one up on my desk for a couple weeks now running fine but it seems to have turned itself off and forgotten its power restore option, which means either the CMOS battery is dead or something is horribly wrong. I can't turn it back on without taking it apart because the power button is busted.
Turns out I never actually replaced the battery in that machine. Ha ha.
I went through all 16 to replace battery, unlock the bios, change to boot on power restore, and run the self check. A bunch have broken power switches, which I knew. Some of them even have broken power LEDs.
Only one has a bad eMMC chip, which is great.
Next subproject is to get a headless immutable USB booting Alpine running.
Also I want to try writing a kernel driver <_<