In the spirit of license-problem-gfdl-invariants, lintian should check that
license blocks with short name GFDL-X.Y are actually GFDL-X.Y (in which case
they would be non-DFSG), or if they should instead have short name GFDL-NIV-X.Y
(the short name specified in DEP5[0] for GFDL with no invariant sections). In
particular, nothing with short name GFDL should be destined to main.

[0] https://www.debian.org/doc/packaging-manuals/copyright-format/1.0/#license-short-name

Hi,

How can Lintian tell the difference? Could you provide an example? :)


Regards,

Hi Chris,
emacs-ivy-doc-non-dfsg.  It has "Section: non-free/editors" and the
org file source used to generate its documentation is "License:
GFDL-1.3+", with invariant front-cover and back-cover texts.

Lintian errors with:

    E: emacs-ivy-doc-non-dfsg source: license-problem-gfdl-invariants
    debian/copyright invariant part is: with no invariant sections,
    with the front-cover texts being a gnu manual, and with the
    back-cover texts as in (a) below…

I would like to propose the following solution for Ryan's issue, which
also solves the error I'm experiencing:

Keep the existing parsing.  Definitely document that a lintian
override is not appropriate for the initial upload (such an override
feels like it should be an NEW AUTOREJECT).  Then add the following cases:

If a stanza has License: GFDL (no qualifiers), or GFDL-x.y[+|-only]
and the declared section isn't non-free/foo, then error.

If a stanza has License: GFDL-x.y[+|-only], and invariants are found,
and the section is correctly declared as non-free/foo, then warn but
do not error.  That warning is an extra hint to ftpmasters imho.

If a stanza has License: GFDL-NIV.* then warn.

I've probably left out a number of possible cases, but I'll leave
those to someone more capable.  What I'm not sure about is what to do
on subsequent uploads...  It would be nice to have the liberty to
override the warning and hide it if a maintainer is 100% certain that
upstream won't budge on their choice of license.  On the other hand, I
see the benefit of a nagging lintian output to remind the maintainer
to contact upstream about their choice of license.


Regards,
Nicholas