For Approval: Open Source Hardware License

Brian Behlendorf brian at
Fri Jul 6 16:44:13 UTC 2007

On Thu, 5 Jul 2007, Allison Randal wrote:
> It's pretty obvious to me. On the simplest level, you want to use terms 
> describing the "work" that are relevant to hardware. Sure, you can apply a 
> software license to hardware by analogy, but it will never be clear. (e.g. 
> What do you mean by "copying the source code" of a piece of hardware? Where 
> does the distinction between hardware designs and physical hardware enter 
> into it?) We're expanding into new fields of law here, and we need to start 
> developing the tools of the craft.

Huh - while I can see "source code" in the language of many OSI licenses, 
I never took that to mean that the licenses could only be applied to 
software.  Software in source code form may have been the context in which 
these licenses are described and considered, but a GPL-licensed collection 
of source code, documentation, UML diagrams, paper napkin sketches, and 
audio recordings of the developers' favorite original jokes, are all 
still GPL-licensed.  E.g. in the Apache 2.0 license:

       "Source" form shall mean the preferred form for making modifications,
       including but not limited to software source code, documentation
       source, and configuration files.

I wouldn't want to overestimate a jury or judge, but I think anyone would 
consider "code written in hardware description languages such as Verilog, 
VHDL, or Bluespec" to fall into the above.

> As to the choice of which license to base it on, it's largely governed 
> by the intended use of the licensed hardware. If you extend the analogy 
> of the GPL to hardware, it implies that your open chip could only be 
> used within larger pieces of hardware that are also completely open. 
> Someday we'll get to that point, but at the moment, as we build up 
> momentum in open hardware, that's a huge obstacle both in convincing 
> companies to open up their hardware, and in convincing others to use the 
> open hardware.

For some companies, like Sun with OpenSparc, compelling opening of larger 
works would be part of the point.  For others, I would think it would 
simply mean that few people would consider strong-copyleft licenses for 
hardware, and would instead consider MPL, CDDL, Apache, etc.  It shouldn't 
scare them away from open source entirely, unless we did a poor job of 
educating them on what the licenses mean.

The difference between hardware and software is already one more of form 
than function.  AFAIK, anything describable in software can be burned into 
silicon or reduced to an embedded system; likewise most hardware today is 
emulated as software before being manufactured.  This even applies outside 
of chip design - consumer product design has long been a mostly digital 
process.  The biggest software company in Shanghai (I was told today, so 
without reference) is one that makes control software for cutting/shaping 
machines at steel mills.  I won't even get into DNA and synthetic biology. 
:)  In all of those, using the word "source" to an English-language 
speaker to describe the digital designs would be pretty comprehensible.

So I fall into the camp of asking for more clarification on how hardware 
really is different...

I just noticed that in the Apache 2.0 license, the only place (in addition 
to the above snippet) that the word "license" appears is in the disclaimer 
of liability.  I guess we should have made that "works distributed under 
this license" rather than "software distributed under this license" - 
because we didn't intend to take liability for other kinds of works.  :)


More information about the License-discuss mailing list