Re: Can a license be "DFSG approved"?

[Ian Jackson <ijackson@chiark.greenend.org.uk>]

Paul Wise writes ("Re: Can a license be "DFSG approved"?"):
> On Thu, Oct 26, 2017 at 9:53 PM, David Seaward wrote:
> > I was also reminded of the comment below, and wondered in retrospect if
> > a flagging licenses as "DFSG approved" actually makes sense?
> 
> The DFSG is not a body that can approve things. You could say "DFSG
> compatible", or "foo version X under license Y was approved by the
> Debian ftp-masters".

I think a better and more verifiable thing to say would be:

"Software wholly under this licence has been accepted as Free Software
by Debian".

The evidence required for this assertion is, of course, a pointer to
the Debian archive.  A possible counter-evidence might be a bug
report.  Depending on the response of various parties, bug severity,
etc. the status of the software might then be "disputed" or
"rejected".

A difficulty is that a licence's status according to Debian might
change over time.  Does this attempt at taxonomy handle that ?


Note the important difference between the following two kinds of
reasoning:

 A. Package X is in Debian ([url to archive]) and the version in
    Debian is solely licenced under licence Y.  There are no bugs in
    the Debian BTS against X claiming it is non-free.  Consequently,
    licence Y seems to be sufficient in Debian, if there are no other
    problems, and other packages under the same licence are likely to
    be accepted (assuming they pass copyright review etc.)

This is sound reasoning.

Verssus:

 X. Package X was discussed somehow by Debian.  { ftpmaster rejected
    it | there is an RC bug open against the version in the archive
    | it is scheduled for removal }.  The version Debian considers
    questionable is licenced solely under licence Y.
    Therefore XXX Debian does not like licence Y XXX.

The conclusion surrounded by XXX does not follow from the premises.
Quite possibly there are other problems with the specific package -
missing source code; some files whose authors do not actually appear
to have been licenced them under Y; or whatever.

To decide that the problem is with the licence would require more
interpretative work of Debian's public stance and discussions.  This
is difficult because Debian does not always provide clear reasons for
rejection.  It is also difficult because there are many licences and
packages that are so obviously non-free that no-one ever discusses
them - and bringing them up on debian-legal, or asking ftpmaster about
them, would be a waste of time.


I do wonder if this approval tag might set up a bad incentive, for
people who want to promote licences to try to get a package or two
with that licence into Debian.

Ian.

-- 
Ian Jackson <ijackson@chiark.greenend.org.uk>   These opinions are my own.

If I emailed you from an address @fyvzl.net or @evade.org.uk, that is
a private address which bypasses my fierce spamfilter.