A few weeks ago the initial full release was made of the OSHW specification. The Open Source Hardware spec is the first attempt to start an open source revolution for hardware. At first glance it isn’t terribly interesting. It has a lot of license and definitions about what must be included to be OSHW compliant. And if you are like me, were expecting something more technical and meaty. You are hoping for pin outs or a platform specification, not the mention of documentation that must be written. After reading reading it, I moved quickly onto other things not thinking about the possibilities.
This week I revisited it, thought about why this is important and how it could affect professional hardware people (EEs, MEs, and FWEs). Honestly, I struggled because I wanted the reasoning to have technical merit. I read a few blog posts that pointed out the current issues with trading designs and what that might mean for hobbyists where tools are not standardized and tool costs matter heavily for design adoption and permeation through the community. I had not considered this problem. Not to mention the uncouth ways that vendors and manufacturers can make it hard to share information later on. It wasn’t until late last night I thought about what the impact could be.
I sense two possibly positive impacts in my work. In my mind, parts vendors are a pain. They want you to sign NDAs for the most mundane of parts. They obfuscate whenever possible for no good reason until you tell them you are looking at other vendors to replace them. I would love it if I could know beforehand if a vendor was held to the standard of providing reasonable documentation as well as let me share my design in the future without a legal gymnastics routine and costly per chip royalties paid out. OSHW provides for that! A vendor with the little logo such as the proposed below:
With a consistent way of allowing for release, distribution, and reuse, the hardware hobbyist community can flourish in a similar manner that the open source software community did. At first, Open Source Software not taken very seriously by the professional community, but in the last 10 years has become a force to be reckoned with. How many times have I had to place a Linux distribution on a microprocessor that was not x86? Too many to count, but each time it gets a bit easier because the OSSW community is updating and refining it to make it easier. This movement was enough to get WindRiver’s VxWorks to start thinking about Linux and restrategizing due to developers taking their money elsewhere. Perhaps it will be enough to get large vendors to provide more client-favorable sales strategies or at least take their documentation more seriously so more people will be willing to use parts from them.
My second vision of how this could upset the current status quo is based on how it could reshape rapid prototyping and potentially change the web-driven startup landscape to something more diverse. Right now, prototyping hardware falls into two camps:
- cheap and clunky, or
- expensive, over-kill, and difficult to obtain.
OSHW could allow more iterative design among the community which bring a lot of choice; choices that bring down cost and clunkiness in prototyping. I could try an idea out faster and closer to my desired specs and form-factor without signing a bunch of NDA’s before I even start to dig in. Sadly, removing this road block saves me days, not just hours. Think about how productive you could be without this. Then roll the fact that in that mechanical drawings, layout, and so forth will also be opened and available. The only thing that could be better is if the design was already FCC certified. How fast could you now deploy? And at what cost savings? We are now talking about development iterations closer to software and web projects. Hardware now has a chance to wrap its head around Agile methodology. And faster development times, mean the possibility of getting to market faster and breaking even sooner. Could this mean VC’s eyes will start looking at hardware again as a good place to invest? One can only hope.