[cc-licenses] Subject: Re: Version 3.0 - List Discussion Responses

Terry Hancock hancock at anansispaceworks.com
Mon Oct 2 12:56:20 EDT 2006

I haven't read the other replies yet, but I think it matters that I 
respond to a couple of particulars (even if this is repeating others' 

drew Roberts wrote:
>  On Saturday 30 September 2006 02:15 am, Terry Hancock wrote:
>  So, I have not kept the fine details in my head lately but I will
>  state it again another way to try and help clarify and wait for
>  corrections.
>  The GPL-3 addresses DRM from two points.

Precisely: in two separate parts of the license.

>  1. GPL-3 code cannot be incorporated into a DRM system that is put
>  forth by its maker as being "an effective TPM" as per the DMCA.

This is what is usually called the "anti-TPM" part. AFAIK, it is not 
really very controversial (there were wording problems in draft 1 which 
have, IMHO, been fixed, and the new wording is pretty solid). In fact, I 
heartily applaud the new wording in this part.

>  2. The GPL-3 also contains language that would prevent a TPM/DRM
>  system containing (made from) no GPL-3 code whatsoever from being
>  used to "protect" GPL-3 code unless keys were provided such that
>  those getting the platofrm and code can mod the code and sign and run
>  the modded code on the platform.

Yes. This is addressed much earlier under the definition of the term 
"Corresponding Source". It is the language that Linus Torvalds and other 
Linux developers objected to; it is the part that fights "TiVo-ization"; 
and it's the part that I'm talking about when I say that the GPLv3 would 
not allow a work licensed under it to be "ported" to a DRM-only platform 
(assuming that Greg London's "DRM Dave" example is correct).

It *might* even be construed to violate Eben Moglen's claim that a 
ROM-only embedded device could run GPLv3 code (I think this is really a 
wording problem, and I'm hoping it'll get fixed).

> > The GPLv3 is a big, complicated can of worms about to be opened,
> > and to dismiss so casually the legitimate concerns of people who
> > have a high stake in the use of the GPL license and 15 years of
> > experience with the older version, is naive, IMHO.
>  I am not so sure the kernel guys have such a high stake in the GPL3.
>  Some seem to think that they can't put the kernel under the GPL3 even
>  if they wanted to. At least without (major?) re-writes due to code
>  contributed by people who will not go along or who can't be found, or
>  who are dead...

No, it's the GPL *v2* that they have a high stake in. The problem is 
that GPLv3 could potential de-value the GPLv2, by introducing a large 
body of GPLv2-incompatible code. The Linux hackers therefore fear 
fragmentation (forking) of the free O/S world into GPLv3 and GPLv2 
camps, much the same as the BSD/GPL fork.

I'm not sure I agree with the Linux developers' opinion, but I don't 
think they are "mistaken" about any of the legal language in question. 
IMHO, the difference is a real difference of intent, not merely of 
misunderstood semantics.


Terry Hancock (hancock at AnansiSpaceworks.com)
Anansi Spaceworks http://www.AnansiSpaceworks.com

More information about the cc-licenses mailing list