Open hardware

Publishing the design of a physical device — schematics, PCB layout, bill of materials, fab files, firmware — so anyone can build, repair, or modify it. The hardware analogue of open source, with one structural difference: copying still costs money. Anyone can fork a repository for free; forking a board means ordering PCBs and buying parts, so the practice depends on cheap prototype fabrication (JLCPCB, PCBWay) in a way software never depended on anything.

Two motivations, often confused

Platform building: Arduino, BeagleBoard, OLinuXino and much of awesome-iot's hardware list publish designs to make a platform adoptable. Open as a distribution strategy, with a company selling boards at the centre.

Control over the object: The Open Book publishes so that the person holding the device governs it, a stated right-to-read position against depending on proprietary reading platforms. Nobody is building a business on it; the schematic is the argument.

Substitution (added 2026-08-08): flopperziro publishes so that a commercial product's capability can be had without buying it — "a Flipper Zero clone, but cheapest." Not a platform, not a principle about ownership; a price argument. What makes it structurally different from the other two is what it requires: platform building needs a company and control-over-the-object needs a position, while substitution needs only that the product's functions decompose into parts already on sale. Where they do, the commercial advantage left is integration, firmware and industrial design.

All three are open hardware and they answer different questions, so the license alone doesn't tell you which one you're looking at.

Reuse (added 2026-09-27): suutari20, a modular macro keyboard, is none of the three above — its PCB imports key-switch and audio-jack footprints from two other published projects (ScottoKeebs and the kbd library) rather than drawing its own. That is a fourth thing a design can inherit besides a licence: a shared component commons other projects have already verified against a board house. It doesn't move the replication floor below — the result is still a custom PCB, the same rung as The Open Book — but it lowers the design cost of reaching that rung, the hardware analogue of importing a library instead of rewriting the routine.

The replication floor

The three motivations sit at different replication costs, and that turns out to track them closely. The Open Book publishes a board: replicating it needs PCB fabrication and reflow. flopperziro publishes a wiring diagram over commodity breakout modules and no custom PCB at all: replicating it needs a soldering iron. The second is the lowest rung this wiki has seen, and it buys that by giving up integration — more boards, more wires, a bigger object — rather than by better design.

Worth stating because the practice's usual barrier is money for fabrication. Remove the custom board and the barrier is skill and patience instead, which is a different filter on who can build the thing.

And the two rungs have been tested against each other. flipper-zero-diy attempted the same substitution as flopperziro by the other route — recreate the vendor's schematic, custom PCB, factory assembly — and its author abandoned it after pricing the result at roughly $220, above the retail price of the device he was cloning: "because of the form factor and factory assambling the price of creating a DIY Flipper would exeed the price of a bought one."

Same goal, same week, opposite outcomes, and the variable is form factor. That gives the substitution motive a boundary condition rather than a caveat: it holds at the breakout-board rung and inverts at the custom-PCB rung, because what the commercial product sells is integration, and integration is exactly what a build of one cannot buy cheaply.

Tooling can shrink the design step, not the floor

verodesigner (added 2026-09-27) sources a rung the spoke hadn't named: point-to-point construction on pre-drilled stripboard, sitting between flopperziro's bare breakout modules and The Open Book's custom PCB — no fabrication step, but also no ready-made modules; the builder cuts tracks and runs jumpers on generic board. The tool draws a stripboard layout from a schematic automatically, which is a different saving from suutari20's footprint reuse: reuse cuts the cost of drawing a custom PCB, VeroDesigner cuts the cost of planning a stripboard build. Neither moves what has to be fabricated or bought — the replication floor is about the physical object, and tooling only ever touches the design labor that precedes it.

Design for hand assembly

When the point is that others rebuild the thing, part count and solderability become design constraints ranking alongside cost and power. The Open Book states this directly — low part count specifically so it can be assembled by hand, with fab files prepared for the cheap prototype services. A commercial board optimizes for pick-and-place; an open one optimizes for a person with an iron, and the two produce visibly different layouts.

Licensing is messier than software

The Open Book uses per-component licensing — different terms for the hardware design, the firmware, and the documentation — which is normal here and has no clean equivalent in a single-LICENSE software repo. awesome-iot itself is CC0. Expect the license question in this field to have several answers per project.

the-open-book-ereader · flopperziro · suutari20 · verodesigner · embedded-systems · microcontroller · awesome-iot · synthesis