First Boot On Real Metal
The pitch
Section titled “The pitch”A Raspberry Pi OS image landed in hardware/ on April 19 (2026-04-13-raspios-trixie-arm64-lite.img.xz, 562MB). That suggests flashing a Pi Zero 2 W and booting the prototype is days-to-weeks away — a real hardware event. When it happens, it’s a dispatch. Arguably the dispatch of Q2. A screen the size of our hand lights up amber, and the same framebuffer the desktop emulator has been drawing for weeks becomes a physical object.
Stage: event-gated research. We can pre-load everything except the thing that has to actually happen.
Why this is research, not outline
Section titled “Why this is research, not outline”We don’t know when the event happens. Could be this weekend. Could be three weeks from now if a component is wrong. Could have a false start that’s itself a story. We can’t outline a narrative until we know what actually went down.
What research can do now: stage the pre-shoot checklist so that when the boot happens, we capture everything we need for a great issue, and nothing is lost because Josh was too excited to document it.
Pre-boot capture checklist
Section titled “Pre-boot capture checklist”Write this down now so it’s ready:
- Photograph the unboxing. The Pi Zero 2 W in its clamshell, the Elecrow 7” panel in bubble wrap, the SD card before it goes in. These are the “before” shots.
- Photograph the workbench. Wide shot of the build station. Whatever iron / stand / lights / wires are nearby. Atmosphere, not clarity.
- Record the flash. Quick screen capture of
ddor Raspberry Pi Imager running. This isn’t for the reader; it’s for us, to prove when this happened. - Video the first boot. Phone on a tripod or leaning against a coffee mug. Angle on the Elecrow screen. Let the camera roll through the boot sequence, even if it takes 30 seconds and nothing happens. Especially if it takes 30 seconds and nothing happens.
- Photograph the first on-screen artifact. The moment the amber appears. Even if it’s a kernel panic, it’s a story.
- Photograph a cartridge on-screen. If we can get all the way to
./kn86emu ice_breaker.kn86running on the Pi, that’s the money shot. - Write the 200-word raw diary entry within 24 hours. Not publication-quality — notes. What worked, what surprised, what broke, what smell the solder had. We’ll mine this for the actual dispatch.
Probable structures (pick one when event lands)
Section titled “Probable structures (pick one when event lands)”If everything works on first try
Section titled “If everything works on first try”Short issue. 700w. Three paragraphs of narration + three photos + a closing note on what’s next (the button matrix, the YM2149 wiring, the case). Tone: quiet awe. Don’t over-claim.
If it works on the second try
Section titled “If it works on the second try”Medium issue. 1000w. Three-act structure: expectation → failure → fix. The failure is the interesting part. Cite the specific thing that broke (a dtoverlay missing, a display driver module not loaded, a resolution mismatch) because that’s the texture readers come for.
If it fails and we’re still debugging on May 15
Section titled “If it fails and we’re still debugging on May 15”Still a dispatch, just a different one. 900w. “What we tried, why we thought it would work, what we’re trying next.” Honesty is the voice. The dispatch goes out even if the device didn’t boot. That’s part of the contract with subscribers.
If the hardware doesn’t arrive yet
Section titled “If the hardware doesn’t arrive yet”Skip T04. Move T05 forward. Update the pipeline. Don’t send filler.
Source material to draw from
Section titled “Source material to draw from”| Source | Path | Use |
|---|---|---|
| Modern Build Spec | docs/hardware/KN-86-Modern-Build-Specification.md | BOM, assembly notes — use as background for what’s in the prototype |
| Sourcing Guide | docs/hardware/KN-86-Sourcing-Guide.md | If specific components deserve a callout (Alps SKFL switches, Elecrow panel) |
| Prototype Architecture | docs/software/runtime/prototype-architecture.md | For the “what this Pi is pretending to be” paragraph |
| Raspberry Pi image | hardware/2026-04-13-raspios-trixie-arm64-lite.img.xz | The artifact at the center of the issue |
Open questions
Section titled “Open questions”- Does the Pi Zero 2 W boot story belong in one issue or two? (First issue: “it boots.” Second issue, maybe T06: “now it plays a cartridge.”)
- How much BOM detail goes in the dispatch vs. a link to the Modern Build Spec? Probably one specific component mentioned by name per issue, rotating.
- Do we publish a hardware photograph at full resolution on the blog and a reduced version in the email? (Yes, probably. Set this as a default pattern.)
Decision to defer
Section titled “Decision to defer”When this ships depends on Josh’s schedule, not ours. The pipeline’s job is to be ready when the event happens, not to force the event to happen on a Friday.
What moves this to OUTLINE
Section titled “What moves this to OUTLINE”- Boot event occurs (success, partial success, or educational failure).
- Josh writes the raw diary entry within 24 hours of the event.
- Photographs selected (3–5 for email; more for blog).
- Pick one of the four probable structures above and adapt.
Target: outline within 48 hours of boot; ship by the following Friday.