• upstream changing from GPL-2+ to GPL-3+ without copyright holders p

    From Ian Jackson@21:1/5 to Florian Weimer on Mon Aug 12 00:30:01 2019
    Florian Weimer writes ("Re: upstream changing from GPL-2+ to GPL-3+ without copyright holders permission"):
    Andrej Shadura:
    then it is totally fine to choose that option, since the copyright
    holders have already given that permission you think they need to
    give.

    Yes. For more detail, see Francesco Poli's response which I think
    answers the legal question entirely correctly.

    In general, I agree. But there might be cases that are less
    clear-cut. For example, if the upgrade from GPLv2+ to GPLv3+ is used
    to gain permission to combine the work with an AGPL work, especially
    if this is done in an "open core" context.

    Florian, are you still answering the legal question ? Because I don't
    see how these kind of considerations (eg, the reaason for the licence upgrader's choice, or the business model of the original authors)
    could make much of a legal difference. So I think if you intended to
    answer the legal question then I completely disagree with you.


    If you intended to answer (given the introduction of "fine") moral
    questions instead, then of course these questions are much more
    subjective.

    Then your view is a reasonable one, but I would tend to disagree.

    Personally I have often taken GPLv2+ code and mixed it (either by
    textual copying, or by linking) with AGPLv3+ code, taking advantage of
    the GPLv2+ version upgrade and the GPLv3+'s AGPL-compatibility clause.
    I have done similar things with LGPL'd code and the GNU GPL(s).

    I dislike the open core business model; and I dislike
    software-as-a-service models where the code is in practice Free for,
    and owned by, only the service operators (usually, one dominant
    service operator) - and thus in practice proprietary for everyone
    else. I don't think there is any ethical imperative to support these
    business practices, and no obligation to avoid undermining them.

    I also think that in general, Debian should try to respect copyright
    holders' wishes, even if the project is not required to do so.
    Disregarding authors rarely leads to good outcomes.

    I think taking proprietarised GPL service code, and making significant enhancements under an AGPLv3+ licence, is a Good Thing. Even if it
    upsets the original authors or the proprietary service operators.

    Whether it is a good community-building strategy is a different
    question of course, and depends on how much interest there is in
    freeing the code in question.

    And, as individuals and companies we oftn need to make practical
    political compromises (eg in our employment, or in our business
    dealings, etc). That might mean we would avoid doing things that
    upset original authors, etc.

    But Debian as a project is committed to Free Software. As a project
    we do not restrain our contributors from annoying proprietary software companies, or more generally from fully exercising[1] the legal
    freedoms we have to deal in Free Software.

    In other words: how much weight to give to the preferences of the
    original copyright holders is a matter for the individual maintainers.

    Ian.

    [1] Of course we want to act consistently with our own values, such as
    the Social Contract, the Diversity Statement, and so on. My claim is
    that those values do not include refraining from exercising our rights
    just because it would annoy proprietary software owners.

    --
    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.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Florian Weimer@21:1/5 to All on Mon Aug 12 08:50:01 2019
    * Ian Jackson:

    In general, I agree. But there might be cases that are less
    clear-cut. For example, if the upgrade from GPLv2+ to GPLv3+ is used
    to gain permission to combine the work with an AGPL work, especially
    if this is done in an "open core" context.

    Florian, are you still answering the legal question ? Because I don't
    see how these kind of considerations (eg, the reaason for the licence upgrader's choice, or the business model of the original authors)

    I meant that someone might tru to turn the GPLv3 deriviate into some
    “open core”-like thing.

    could make much of a legal difference. So I think if you intended to
    answer the legal question then I completely disagree with you.

    My point is that authors may have chosen GPLv2+ licensing explictly
    because of its strong copyleft, and the project documentation reflects
    that. Arguably, the FSF has weakened copyleft protection in some
    areas with the GPLv3, and it is absolutely not clear what the
    resulting license conditions are if the author said both “GPLv2+” and “strong copyleft” before the publication of the GPLv3.

    I think taking proprietarised GPL service code, and making significant enhancements under an AGPLv3+ licence, is a Good Thing. Even if it
    upsets the original authors or the proprietary service operators.

    You can also do the reverse: take free GPLv2+ code, make the
    conglomerate effectively AGPLv3+, without a built-in source
    redistribution facility, and then sell AGPLv3+ exemptions for
    organizations who cannot comply with the AGPLv3+ requirements due to
    the lack of built-in source code redistribution capabilities and a
    requirement to configure the program through source code changes.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)