[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Dissident test (was re: CDDL)

On Wed, Sep 14, 2005 at 09:29:52AM +0000, MJ Ray wrote:
> Marco d'Itri <md@Linux.IT> wrote:
> > mjr@phonecoop.coop wrote:
> > >It seems grounded in the DFSGs "1. Free redistribution" and
> > >"5. No discrimination against persons or groups".
> > It may seem so if you are willing to stretch enough the meaning of the
> > DFSG in ways it was never supposed to be. I'm not.
> If non-discrimination doesn't cover groups persecuted by
> governments, who does it cover for you?

I think the point here is that a licence doesn't discriminate against such
groups, it only forbids anonymous changes from being distributed.

Apart from such anonymous changes using a pseudonym or something such, which
would solve this for all considered, i feel that the main point is that if the
DFSG is really about not forbidding anonymous participant to
modify/distribute/whatever their code, it should then be said explicitly,
instead of abusing DFSG #5 to include all sort of strange things.

So, if you really believe we should allow for anonymity, please go ahead, and
propose a a DFSG amendment, i will even second you on this.

> In April 2005, Shi Tao was imprisoned for 10 years for "providing
> state secrets to foreign entities", partly because the local
> police traced his email address back to him (source Reporters
> Sans Frontiers). In that case, it was news of a censorship order,
> but why not news of a security vulnerability that state agents
> are exploiting? Anonymity has benefits for freedom.

Sure, but irrelevant to the DFSG #5 as drafter. The DFSG clauses are meant to
be clear and understandable, and not to resort to chancy interpretation not
everyone agrees with.

Also, i belive he got caught because he was careless in his email handling,
not because he distributed a modified version of some free software with his
email address in it. Also notice that nothing in the CDDL clause force you to
include the email address or even your real name for code attribution, i think
a pseudonym would be widely accepted in such cases as the above.


Sven Luther

Reply to: