Re: [all candidates] Return to the desert island (cont.)

[Michael Gilbert <mgilbert@debian.org>]

On Tue, Mar 19, 2013 at 7:46 PM, Steve Langasek wrote:
> On Tue, Mar 19, 2013 at 11:39:29PM +0100, Jérémy Bobbio wrote:
>> 3. One test I've been taught to use to reason about free software is the
>>    Desert Island test [2] which starts by:
>
>>      Imagine a castaway on a desert island with a solar-powered
>>      computer.
>
>>   Obviously, software that are only frontends to unreproducible “cloud”
>>   services do not pass the desert island test.
>
> This is a mischaracterization of the "Desert Island test" as it was
> formulated on debian-legal.  The Desert Island test is about whether a user
> can *comply with the license* of the software on a desert island when they
> have no contact with the outside world.  That the software may not be
> *useful* to them on a desert island is a separate question, and applies to
> many sorts of software, not just those used for connecting to particular
> services over the Internet.

Then again, this is a misinterpretation of the fundamental question
Jeremy has attempted to pose.  Although admittedly, bullet 3 does
unfortunately lead to that kind of reasoning since it frames the
question in a freeness context, which I agree is misleading.

So, what then is the fundamental question Jeremy meant to pose?  It is
this: given the distinction between main and contrib as stated in
policy 2.2.1 and 2.2.2 respectively:

    The main archive area comprises the Debian distribution. Only the
    packages in this area are considered part of the distribution. **None
    of the packages in the main archive area require software outside of
    that area to function.**

    [...]

    The contrib archive area contains supplemental packages intended to
     work with the Debian distribution, but **which require software
     outside of the distribution to either build or function**.

are the packages currently in main that "require software outside of
the distribution to either build or function" ultimately violating
policy?  And if so, should they be moved to contrib?

I think Jeremy's suggestion of a Desert Island "Use" test seems quite
appropriate to decide that question.  And again, just to be clear, the
Desert Island "Freeness" test has no relevance in answering the real
question posed here.

Note that I've looked into this problem in the past, and such cases
where this question is relevant in main are fortunately quite rare.
Nevertheless, it's an interesting subtle aspect that isn't much
thought about (at least not in the right way).

Best wishes,
Mike