James Blackwell

From jblack Thu Oct 19 00:10:20 2006
Date: Thu, 19 Oct 2006 00:10:20 -0400
To: [Someone]

On Mon, Oct 16, 2006 at 11:04:16AM -0700, [Someone] wrote:

> James, since you are in the GPL v3 discussion committee, I would like to
> ask you what is the current situation with respect to the recent GPL v3
> FUD from the kernel developers. As a GNU supporter I am somewhat worried
> about the implicit threats posted by that letter from the kernel people.

I haven't been very involved lately but I'll answer what I can.

> I heard, in the same room, from the mouth of a key kernel developer,
> that, if the GPL v3 process continues as planned, "another organization"
> will come forward to fork all important projects under the FSF's control
> (namely, gcc, glibc, binutils, etc.) to continue under the GPL v2 such
> that the FSF will become irrelevant. I am not sure how serious is the
> threat but I believe the kernel people seem confident this is what will
> happen. So I am curious what do people in the GPL v3 committees think
> about it? Will the GPL v3 process/time schedule need to be changed
> (maybe delayed) in response to this threat?

I haven't talked with the committee in some time. I wouldn't be privy to
how they think internally. There is certainly a disconnect between the
goals of many of the linux kernel developers and the free software
foundation.

One of the two key disputes between these two groups is what I'd call
"software locked devices". This sort of devices that concerns the free
software foundation all share a certain set of characteristics:

* These device uses free software. Typical software include the linux
kernel, busybox, firewalling tools, filesystem management tools, etc.

* The maker of the device typically releases their modifications to free
software that is contained in the device. Good examples of this is
Tivo and Linksys releasing their modifications to the linux kernel.

* The device hardware includes features (misfeatures to organizations
such as the FSF) that will only run binaries that have been digitally
signed by the maker of the device.

This software locking presents an opportunity for companies to comply with
the GPLv2 license while barring part of the original intent of the gpl.
The primary intent of the gpl is to allow for the study, deriving and
modification of code. Software locked devices, while allowing for
the study and deriving of code, effectively succeed in blocking
modification.

The FSF sees this as an unintended loophole in the gplv2 that, if not
closed up, will eventually result in the widespread locking down of
consumer oriented hardware that returns users to the dependant state that
they were in prior to the advent of free software. Its also a slap in a
face; manufactures first use free software written by a set of developers
and then turn right around and prevent them from fixing their own software
once its been uploaded to the hardware.

The group of outspoken kernel developers aren't concerned about companies
that protect unauthorized modification of the software that goes with the
device. Their public stance seems to be that software locked devices will
ultimately fail due to two reasons:

* Consumers will prefer more flexible devices and will perfer devices
that are not locked over those that are. Unlocked devices will get
better revenue than locked devices and the locked devices will thusly
end up less well funded over time.

* Companies that rely upon locked software will effectively exile themselves
from the free software community. Companies with unlocked software devices
will be listened to better than locked software devices and are more
likely to get their modifications merged into mainline.

That's the public view of the kernel developer side. These arguments from
the kernel team don't convince me. Digital video recorders such as Tivo
and DirecTV's, both of sign software and prevent end users from replacing
the free software devices, have not lost out to true free software DVRs
such as mythtv. My suspicion is that many of the kernel developers make a
living working for companies that in part want locked down hardware....

So now we get to your question... whether some group that sees things more
like the kernel developers than the FSF developers can write something to
put the FSF out of the license business.

No, I don't think so. Not for several reasons.

* The first is that there are already a variety of licenses available that
are both more and less free and active than the gplv3 draft. People that
don't like the gplv2 already switched to something else (typically
freebsd). Those that don't like the gplv3 draft can switch after it
comes out.

* The second is that the gplv3 changes only show an effect when you start
talking about proprietary hardware preventing behaviours intended by
free software -- tinkering. Well, that and patent protection.

* Third, there will be a large library of software out there that is
automatically licensable by anyone as gplv3. Software licensed as gplv3
may or may not be compatible with the gplv2 according to which
license extensions are given to the software.

* Fourth -- well, I really hate to say this, far too many people out there
that really couldn't care less. A small fraction of software users could
care less about the distiction of free vs. proprietary. A small subset
of those are going to care one way or another. If they like the gplv2,
then they'll almost certainly like the gplv3 as well -- provided they
don't work at companies that like to contribute code and then sue for
patent infringement or want to prevent users from replacing the software
on their devices.

* Fifth: The FSF in the world of free software licensing is a behemoth.
There are enough people that will take the gplv3 on gospel based on its
previous history that its seeing something that the general public does
not.

* Sixth: I'm willing to stake my reputation on the belief that within
twelve months I think that the people that continue to decry the gplv3
will be vastly outnumbered by the people that support it.

* Lastly: The FSF isn't known for backing down from detractors (if it did,
it wouldn't be here today!). Its been known to outweigh controversial
decisions. As time passed, the general view moved from "their insane" to
a more prescient one.

Regards,
James