Re: A possible GFDL compromise
On Mon, 25 Aug 2003, Florian Weimer wrote:
>> Nowadays we have to struggle constantly against the tendency to bury
>> the free software movement and pretend that we advocate "open source".
>> So I don't think we can conclude that such precautions are no longer
>It's true that many have gladly taken GNU software while ignoring
>the GNU philosophy (or actively working against it). But I doubt
>that invariant sections alone can ensure that the message will be
IMHO, it is not about "be heard". It is about "not to be
>For example, I might want to distribute the GNU Emacs manual without
>the GNU Manifesto. I could achieve something which is very close,
>even though the Manifesto is an invariant section: I just patch the
>Info viewers not to display the Manifesto. As far as I can see, I'm
>still allowed to distribute the modified Info viewer under the GPL,
>and the (unmodified) manual under the GFDL.
It is not the case. Users still have the copies of
Manifesto, and everyone can find and remove the maleficient code.
>However, if someone did something similar, I'd expect quite a lot
>of additional publicity for the GNU Manifesto. Furthermore, the
>publicity wouldn't depend much on the legality of the removal or
>suppression. Journalists who are interested in free software
>philosphy and its battles would report it nevertheless, and those who
>are after awkward legal problems have such a limited audience that
>their silence wouldn't matter that much.
Your position is excessively depends on existense of large
amounts of independnent honest journalists, independent honest
distributors of free software|manuals, independent honest
web-hosters, reasonable laws etc.
Often, it is not the case even now. Very likely, it will not
be the case at all in foreseeable future. System of free
documentation, like the system of free software, should be robust,
and should not depends of the honesty and willingness of someone
outside this system.
Btw, in the FAQ of this list mentioned several noteworthy
tests for freenes, namely "Dissident test" and "Evil Corporation
Test". Why not apply they to this situation?
Consider the cases
* Evil Co takes a free manual, removes all mentions
about free software and published it as
documentation of (compatible in some aspect) its own
* Evil Co takes control over writer's site and
replaces all copies of manual there by "stripped"
version. Majority of mirror hoster notices errata
only after "upgrading" the manual.