Open x – a bit more
FYI – I’m approaching this from a hardware designer’s point of view.
Rev0 (first pass of HW) is often tricky, even if we’re putting together known schematic blocks from other designs. Sometimes we learn tricky power supply issues, other times we learn that JTAG/ETM or full speed memory writes won’t work because of signal reflections.
Often these Rev0 boards are useful and available 1-3 months before Rev1 (1st redesign of HW) comes out. As a HW guy I would simply toss the Rev0’s away, but the SW people I work with can start programming and preparing for the real deal on these boards. (I wouldn’t suggest these boards be sold – they would be for the core design group)
Once Rev1 is out it is typically a board that the device driver, bootloader and HW engineers can start testing. Once communication is proven to most of the devices, I think it is a good time to distribute (or sell) these boards to interested parties who want to work on the new technology as early as possible.
The above is my definition of proto-hardware.
My definition of Open Hardware would be defined as the system with none of my special algorithms implemented.
For instance on a Open Hardware Phone it would be able to make calls, the display would work but only a rudimentary UI would be provided, and the driver to the TI WiFi/BT chip would work, but the ANT+ drivers may not be provided yet.
This would give the ANT+ driver developer a chance to play with the TI chip to understand the nuances of bring-up.
For an Open Product Phone – imagine getting a Droid with all the available source code so you could modify it in any manner you wanted – apps, keyboard, UI – anything they did you can simply tweak.
Hopefully this clarifies it a bit more and lets us continue the conversation and help define OHW