[cc-community] Open Hardware Licence

Greg London email at greglondon.com
Fri May 8 10:48:50 EDT 2009


>> On Wed, May 6, 2009 at 14:44, Patrick Godeau wrote:
>>     http://iang.info/en/

I am not a lawyer and this is not legal advice,
but *that* is a very strange license, and I
would suggest people looking for a license
avoid it until it's hammered out and someone
uses it on a large scale with lawyers and whatnot
checking it out, and it turns out to work as
expected.



> Date: Thu, 07 May 2009 23:16:18 +0100
> From: Rob Myers <rob at robmyers.org>
>
> So if I modify some software but don't give it
> to you, your freedom to use the software
> that you actually possess is not affected.

I'm talking about hardware though, and that's
the problem. If you take source code for an
open processor design and put it under GNU-GPL,
and then some company takes that code, modifies
it, and distributes a physical chip with that
modified processor in it, then GNU-GPL does not
require that vendor to make their mods available
to the public.

Because the physical chip is not a copyright derivative.
Distributing a physical chip is not distributing
the derivative, which is what triggers copyleft in GNU-GPL.

Which means hardware has the same hole in it that
afffects open software and web-server software.
The modifications are kept private, but the public
is allowed to execute the code via a website.
Since the derivative is never actually distributed,
the copyleft aspect of GNU-GPL doesn't activate,
and the web host is not required to distribute source
code of their modifications.

But the problem is that ALL HARDWARE source code
has this loophole in it, which means that ALL HARDWARE
is effectively NOT protected by any copyleft license
that allows private derivatives.

Which means NO OPEN HARDWARE PROJECT is really
protected by any of the copyleft features if
they're licensed GNU-GPL.

Folks can argue the merits and demerits of
allowing private derivatives when it comes
to software. The only example of where that
seems to create a loophole is in webserver
software.

But for hardware, the loophole encompasses
everything the hardware community creates,
which means you might as well put it under
a Public Domain Style license of some kind.
No vendor is forced to make their modifications
public or available under a license so long
as the license allows private derivatives.

If you want open hardware to be protected
at the same level as open software, then
open hardware MUST prohibit private derivatives.
i.e. APSL.

Either that, or someone has to make a special
license geared specifically towards hardware
source code and physical manifestations of
that code, similar to the AGPLv3 has specific
language in it to deal with webhost companies
making derivatives, usign them on public websites,
but not distributing the actual derivative.

> To the best of my knowledge the FSF doesn't
> make any money customizing software themselves.

I recall way back in the trilobyte era of computers
and free software that FSF would require that people
who wanted to work with FSF on their software were
required to sign an agreement reassigning their rights
to FSF.

Maybe that is a false memory. Or maybe it worked that
way back in the day of 5.25" floppy drives but is no
longer the way they operate. I did some googling and
can't seem to find a record of it this morning.
But my google-fu has always been weak.

Greg




More information about the cc-community mailing list