Circadian Results — An Honest Assessment

Results have been mixed — and the reason is worth understanding.
Due to disrupted sleep patterns, my best sleep window currently falls between 8am and 10am.
Setting a sunrise alarm for 9am interrupts the only quality sleep | get. Any sleep is good sleep

right now. The alarm function has been set aside for the time being.

As a sunset lamp and low intensity therapeutic light source — it excels. Deep red. Near infrared.
Ghost mode. A controlled wind-down signal the nervous system actually understands.

The tool is right. The timing is wrong. That is not a design failure — that is an injury.

When normal sleep patterns return, the sunrise alarm will be there waiting.

The Good Machine

| believe in the good machine.

Not the most expensive. Not the most complex. The most considered.

My old Chevy Spark was a good machine. Simple, affordable — but the engineering decisions
were honest. Multi-port injection over direct injection because valve life matters more than
marginal fuel economy gains. Cable controlled heater because cables don't fail like motors do. A
real emergency brake because a real emergency brake works when nothing else does. No auto
start-stop. No unnecessary complexity.

CozZie Glow is a good machine. A Pi Zero. An Arduino Uno. Salvaged relays from a train
bearing detector. Automotive bulbs. A $7 build that outperforms $400 commercial products on
the single metric that matters — biological accuracy of the light spectrum.

No cloud subscription. No proprietary app. No planned obsolescence. No vendor lock-in.

Just good engineering decisions made by someone who understood the problem.

The good machine doesn't cost the most. It wastes the least.