Re: [gNewSense-users] KFV: how to handle COPYING?

[Bake Timmons]

>> Thus, summary scripts would be OK, since COPYING would appear as any
>> other non-free item--except having a comment "waiver" (footnote).
>> This would be less ambitious than what Sam suggests, since scripts
>> would still be "ignorant" about non-free exceptions.  However, there
>> is nothing preventing footnotes from being analyzed to provide such
>> "knowledge" in the future.  Indeed, there might be additional bits of
>> knowledge in footnotes that contribute to automated uses.
>
> So to make sure i understand correctly: It would be marked /non-free/
> with a note about why we include it, rather then /free/ with a note
> about why we consider it free (its a licence).

There are certainly different motivations to justify either way.
Thinking about it again, "free" could indeed be made acceptable with
sufficient explanation, which would matter even more than in the
non-free case given the controversy that has flared up before (e.g.,
in Debian).

Perhaps some people might feel "free" is deceptive, but that is why a
*good* footnote matters--even more in the "free" case.  It matters not
to just avoid deception but to avoid the *perception* of deception.
Again, it should link to a separate page, in part to demonstrate how
much we care about careful explanation and encouraging readers to
think for themselves.  The page can be quite brief, but should at
least be very, very clear with canonical references, etc.  In addition
to the footnote and web page, we should also look for other areas to
clarify, such as the pages introducing (K/P)FV.

A plain, no-nonsense way of presenting it is:

  For the main purpose of Freedom Verification--which is to identify,
  document, and remove non-free sofware that has no place in a free
  software distribution--COPYING is indeed completely "free".

However, we need to be more thorough and clear.  We should generalize
a bit more than we have done, since it is not just COPYING that is
"non-free" in a strict sense.  By *that* measure thousands of files in
the kernel source are non-free since they contain invariant parts,
namely, text or references to licenses of many kinds.  Thus, a better
explanation would be careful not to inadvertently put COPYING in the
doghouse, as if it were any different in this sense than other license
text.

BTW, regarding non-free documentation, while Ubuntu may already filter
it out, should we not be more explicit in rejecting it?  (Maybe I am
not acquainted enough with the gNS web sites.)