[Dnsmasq-discuss] GPL v3

Simon Kelley simon at thekelleys.org.uk
Wed Sep 12 09:59:19 BST 2007

Paul Chambers wrote:
> (putting on my professional hat for a minute)
> If you value the inclusion of dnsmasq in commercial products, as part of
> 'prising them open', then don't force GPL v3 as the only choice. The v3
> license is seen as a step too far by many in the industry, upsetting the
> equilibrium between community and commercial use. In my personal
> opinion, if the v3 license comes into widespread use, it'll do great
> harm to the growth of Linux usage in embedded devices. In my
> professional life, we've has already forbidden GPL v3 code in our
> product codebase, and we're requiring the same of our vendors.

In a sense, it's a redundant question, since it's too late to change
GPLv3, but what specific parts of GPLv3 cause you to implement this ban?
> I just think that GPL v3 actually *lessens* the impact that open source
> can have outside the desktop PC world, and that makes me sad. I want to
> drive more open source into products, but this makes it harder to do so.
> If 'anti-TiVoization' clauses mean TiVo and others are forced to go with
> non-open source solutions (or inferior ones with more commerce-friendly
> licenses), seems like everyone loses.

I guess this is a judgment about hackability. Is a $(TIVO) which uses
open source but can't have it's software changed by anyone but the
manufacturer preferable to a $(TIVO) which uses closed source code? As a
user, I can't see how Open Source helps me unless I can change the
software in my box, or pay someone else to do the same. The fact that a
common reason to lock-down a firmware image is to secure evil DRM just
adds insult to injury.

It's certainly the case that the GPL has been instrumental in opening
development for the *WRT* boxes; I've declined requests to relicense
dnsmasq under the BSD license for exactly this reason.



More information about the Dnsmasq-discuss mailing list