2

Can the author of software make it proprietary after making it gpl?

1 Answers1

5

Yes, the author (or whoever holds the copyright) can do as they please, changing the licence at will. But anyone who already received a copy under a free licence continues to enjoy the rights that accompanied it; the author cannot take those away merely by ceasing to distribute further copies under the same terms.

Case in point is SSH: It was open source up to version 1, version 2 (clearly a development on version 1) is closed. OpenSSH took version 1 (still open source) and created an extension handling the new protocol, released as open source.

MadHatter
  • 48,547
  • 4
  • 122
  • 166
vonbrand
  • 5,267
  • 1
  • 13
  • 30
  • 2
    "what was distributed under an open source license stays forever open" I'm not sure that's entirely true. It's perfectly possible for me to take code to which I own the rights, and which I was distributing under (say) GPLv3, and instead start distributing it under a proprietary licence. People who have already got it from me are completely entitled to continue to use it under the terms of GPLv3, and distributing it themselves (under GPLv3) should they wish to. But I am not obliged to continue doing so should I choose not to. However, that's a fine point, and mostly I agree with you. – MadHatter Jul 18 '19 at 09:00
  • @MadHatter, a license gives the recipient certain rights (in the open source license cases, it includes the perpetual right yo modify and distribute). The author can't just take that away. – vonbrand Aug 04 '19 at 14:58
  • The author (well, rightsholder) can't take that away from people to whom (s)he has already given it, I agree - but (s)he is completely free to stop giving those rights to future recipients of the code. – MadHatter Aug 04 '19 at 15:06
  • @MadHatter, if someone got the right to redistribute, they can still do so. If I decide not to distribute directly anymore makes no difference. – vonbrand Aug 04 '19 at 15:09
  • 1
    I completely agree there's no practical difference, assuming you (wanting a copy of the code under the original licence) can find an earlier recipient thereof, and they agree to give you a copy. It's a fine point I'm making, as I think I've said already - but to say the rightsholder can "only close up later versions" is, at least, ambiguous. Some people think that once you've distributed foo-1.1.1a under GPL you are obliged always to do so, and that's simply not so. – MadHatter Aug 04 '19 at 15:16
  • Is this also true of the MIT license? – n8bar Oct 13 '21 at 22:04
  • @n8bar, this applies to any license whatsoever. – vonbrand Oct 13 '21 at 22:27
  • @MadHatter, we do agree. But perhaps my wording isn't ideal. Care to improve it? – vonbrand Oct 13 '21 at 22:28
  • 1
    @vonbrand well, since you asked... hopefully the above is more precise, though I don't think it reads anything like as well as what you originally wrote. – MadHatter Oct 14 '21 at 06:18
  • I find the whole thing weird since certain extremist forms of licenses infect everything they touch on purpose. It's weird that the future author (time evolved author) can simply cancel that extremist mandate.

    Are there any licenses that revoke ALL RIGHTS OF ALL HUMANS FOR ETERNITY to alter the license in the sense of existence of The Law at least?

     – mathtick Mar 01 '22 at 23:01
  • And another question is can an author unrestrict the license and thereby free the work from restrictive OS licensing? What if two versions are published simultaneously or the timestamps are not clear? Isn't it a risk? – mathtick Mar 01 '22 at 23:02