Re: Consensus about the Academic Free License (AFL) v3.0
On 13/06/15 06:36, Walter Landry wrote: Ángel Gonzálezkeis...@gmail.com wrote: On 12/06/15 23:22, Walter Landry wrote: I would strongly disagree here. Requiring documentation of any sort in addition to the source code is a big step. This is not a minor thing. I don't think requiring that some documentation is provided with the source code, makes it unfree. Of course it does. Mandating a minimum quality before releasing things may be good software practice, but it is decidely unfree. This license prevents a certain class of modifications that, in the original author's eyes, makes things worse. But later people may disagree in good faith. For example, suppose that there is documentation, but it is out of date to the point where it is misleading. This license prevents removing that misleading documentation. Even if you write new documentation, you have to distribute the old documentation as well. Cheers, Walter Landry I very carefully talked about requiring *some* documentation, as something opposed it to all available documentation, which is what AFL requires and makes it problematic. Ironically, the most free program wrt this clause is one with no documentation at all. -- To UNSUBSCRIBE, email to debian-legal-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/557cc075.2020...@gmail.com
Re: Consensus about the Academic Free License (AFL) v3.0
On Thu, 11 Jun 2015 08:41:07 +0900 Charles Plessy wrote: Le Wed, Jun 10, 2015 at 11:48:19PM +0200, Francesco Poli a écrit : Hello debian-legal regulars, I would need to ask your consensus opinion on the non-freeness of the Academic Free License (AFL) v3.0. Hi Francesco, I think that there is a broad consensus to accept the AFL as Free license, in Debian, the OSI, Fedora, the FSF, etc. I don't see evidence of this broad consensus on debian-legal, where my own analysis went uncontested on September 2012, as I said. As far as the FSF is concerned (but please bear in mind that here we are discussing what the Debian Project, not the FSF, should do!), I see that they claim [1] the AFL v3.0 is a Free License, but they strongly recommend to avoid it for practical reasons (due to section 9). [1] https://www.gnu.org/licenses/license-list.en.html#AcademicFreeLicense [...] - point 3) is poorly worded, but assuming it is well-intented, it is Free. Section 3 is really unclear, to say the least. I don't think it will be interpreted as we would like it to be. I agree with Walter Landry that requiring all available documentation is a big step. I think it's non-free. - regarding points 5) and 9), the FSF notes that the AFL has clause similar to one of the Open Software License that requires distributors to try to obtain explicit assent to the license (http://www.gnu.org/licenses/license-list.en.html#OSLRant). This is easy to infringe, but this is not forbidden by the DFSG (which is why we tolerate advertisement clauses, which are also easy to infringe). Section 9 (especially when combined with section 5) is at least a practical problem with mirrors and with many other commonly used mechanisms for software distribution (as noted by the FSF, as well). I am convinced that it is a restriction imposed on re-distributors (thus failing DFSG#1). - The Attribution Notice sounds a bit like an invariant section, but it is also very similar to the NOTICE file from the Apache License, which is Free. As also noted by Walter Landry, there's a crucial difference w.r.t. Apache v2: the latter license requires to preserve attribution notices within NOTICE files; the AFL v3.0 requires instead to preserve *any* descriptive text identified as an Attribution Notice (even when this text includes something other than attribution notices!). I think this is non-free, unless all descriptive texts identified as Attribution Notices only contain attribution notices. Altogether, I think that #689919 should stay closed, although it would be great of course if the Subversion authors would manage to elimiate this license from their sources, because this license is not a good example to follow. Sigh! This won't happen, if the Debian Project (and other authoritative parties in the Free Software community) go on saying that the license is OK anyway... Moreover, please note that, if I understand correctly [2], the files under consideration are no longer included in the upstream Subversion source archive: they are added to the debian/ directory by the maintainer of the subversion Debian package. [2] https://bugs.debian.org/689919#25 Have a nice day, The same to you. -- http://www.inventati.org/frx/ There's not a second to spare! To the laboratory! . Francesco Poli . GnuPG key fpr == CA01 1147 9CD2 EFDF FB82 3925 3E1C 27E1 1F69 BFFE pgplzcbdw5waI.pgp Description: PGP signature
Re: Consensus about the Academic Free License (AFL) v3.0
On 13/06/15 15:45, Francesco Poli wrote: As also noted by Walter Landry, there's a crucial difference w.r.t. Apache v2: the latter license requires to preserve attribution notices within NOTICE files; the AFL v3.0 requires instead to preserve *any* descriptive text identified as an Attribution Notice (even when this text includes something other than attribution notices!). I think this is non-free, unless all descriptive texts identified as Attribution Notices only contain attribution notices. The ftpmasters do not decide whether Debian will accept particular licenses; they decide whether Debian will accept particular software. One possible outcome for this part would be the ftpmasters deciding that AFL-3.0 software is only Free if it does not have any Attribution Notices that are not, in fact, attribution notices. S -- To UNSUBSCRIBE, email to debian-legal-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/557ca492.5060...@debian.org
Re: Consensus about the Academic Free License (AFL) v3.0
On Sat, 13 Jun 2015 22:45:54 +0100 Simon McVittie wrote: On 13/06/15 15:45, Francesco Poli wrote: As also noted by Walter Landry, there's a crucial difference w.r.t. Apache v2: the latter license requires to preserve attribution notices within NOTICE files; the AFL v3.0 requires instead to preserve *any* descriptive text identified as an Attribution Notice (even when this text includes something other than attribution notices!). I think this is non-free, unless all descriptive texts identified as Attribution Notices only contain attribution notices. The ftpmasters do not decide whether Debian will accept particular licenses; they decide whether Debian will accept particular software. One possible outcome for this part would be the ftpmasters deciding that AFL-3.0 software is only Free if it does not have any Attribution Notices that are not, in fact, attribution notices. Yes, that's basically what I meant. Sorry for not being clear enough. I think this clause is OK only for works where there are no descriptive texts identified as Attribution Notices, but containing parts which are not attribution notices. I hope it's clearer now. Unfortunately, the other problematic clauses still hold... -- http://www.inventati.org/frx/ There's not a second to spare! To the laboratory! . Francesco Poli . GnuPG key fpr == CA01 1147 9CD2 EFDF FB82 3925 3E1C 27E1 1F69 BFFE pgpL8JKi6yu0u.pgp Description: PGP signature
Re: Consensus about the Academic Free License (AFL) v3.0
Charles Plessy ple...@debian.org wrote: Here are a few comments about the license. - point 3) is poorly worded, but assuming it is well-intented, it is Free. I would strongly disagree here. Requiring documentation of any sort in addition to the source code is a big step. This is not a minor thing. - The Attribution Notice sounds a bit like an invariant section, but it is also very similar to the NOTICE file from the Apache License, which is Free. It is somewhat different. The Apache license only requires you to preserve attribution notices from the NOTICE file. AFL requires preserving any descriptive text identified therein as an Attribution Notice. There is no requirement that the text actually be an attribution notice. So maybe it is OK as long as there are only attributions in the Attribution Notice. Cheers, Walter Landry -- To UNSUBSCRIBE, email to debian-legal-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/20150612.142252.589075240808081958.wlan...@caltech.edu
Re: Consensus about the Academic Free License (AFL) v3.0
Ángel González keis...@gmail.com wrote: On 12/06/15 23:22, Walter Landry wrote: Charles Plessyple...@debian.org wrote: Here are a few comments about the license. - point 3) is poorly worded, but assuming it is well-intented, it is - Free. I would strongly disagree here. Requiring documentation of any sort in addition to the source code is a big step. This is not a minor thing. I don't think requiring that some documentation is provided with the source code, makes it unfree. Of course it does. Mandating a minimum quality before releasing things may be good software practice, but it is decidely unfree. This license prevents a certain class of modifications that, in the original author's eyes, makes things worse. But later people may disagree in good faith. For example, suppose that there is documentation, but it is out of date to the point where it is misleading. This license prevents removing that misleading documentation. Even if you write new documentation, you have to distribute the old documentation as well. Cheers, Walter Landry
Re: Consensus about the Academic Free License (AFL) v3.0
On 12/06/15 23:22, Walter Landry wrote: Charles Plessyple...@debian.org wrote: Here are a few comments about the license. - point 3) is poorly worded, but assuming it is well-intented, it is Free. I would strongly disagree here. Requiring documentation of any sort in addition to the source code is a big step. This is not a minor thing. I don't think requiring that some documentation is provided with the source code, makes it unfree. However, it could be intended to mean anything from Please don't strip comments from the code or Keep the doc/ folder from the repository when producing a src tarball to Include any documentation ever written related to modifying the original work (a patch howto, an emacs manual?). If the licensor has a copy of Knuth's TAOCP (ie. it's available documentation), and it describes something on-topic for modifying the original work (eg. the work uses linked lists, described in Chapter 2) then the Licensor agrees to provide a machine-readable copy of TAOCP. ∎ -- To UNSUBSCRIBE, email to debian-legal-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/557b7d1c.8070...@gmail.com
Consensus about the Academic Free License (AFL) v3.0
Hello debian-legal regulars, I would need to ask your consensus opinion on the non-freeness of the Academic Free License (AFL) v3.0. My personal conclusion is that this license includes non-free restrictions and is also problematic with respect to Debian mirror infrastructure. My own analysis [1][2] of the AFL v3.0 was sent to debian-legal on September 2012 and received no rebuttal. [1] https://lists.debian.org/debian-legal/2012/09/msg00081.html [2] https://lists.debian.org/debian-legal/2012/09/msg00082.html I then proceeded to file the bug report against subversion, but it was closed [3] with the request to form consensus on debian-legal (which I think was already formed, since nobody objected to my analysis...). [3] https://bugs.debian.org/689919#51 Could you please explicitly express your agreement with my analysis? Thanks for any help you may provide. -- http://www.inventati.org/frx/ There's not a second to spare! To the laboratory! . Francesco Poli . GnuPG key fpr == CA01 1147 9CD2 EFDF FB82 3925 3E1C 27E1 1F69 BFFE pgpHcq2Gf5ZUB.pgp Description: PGP signature
Re: Consensus about the Academic Free License (AFL) v3.0
Le Wed, Jun 10, 2015 at 11:48:19PM +0200, Francesco Poli a écrit : Hello debian-legal regulars, I would need to ask your consensus opinion on the non-freeness of the Academic Free License (AFL) v3.0. Hi Francesco, I think that there is a broad consensus to accept the AFL as Free license, in Debian, the OSI, Fedora, the FSF, etc. Its wording is often poorly chosen, but I think that the consensus is to conclude in favor of the Free interpretation. Here are a few comments about the license. - point 3) is poorly worded, but assuming it is well-intented, it is Free. - regarding points 5) and 9), the FSF notes that the AFL has clause similar to one of the Open Software License that requires distributors to try to obtain explicit assent to the license (http://www.gnu.org/licenses/license-list.en.html#OSLRant). This is easy to infringe, but this is not forbidden by the DFSG (which is why we tolerate advertisement clauses, which are also easy to infringe). - The Attribution Notice sounds a bit like an invariant section, but it is also very similar to the NOTICE file from the Apache License, which is Free. Altogether, I think that #689919 should stay closed, although it would be great of course if the Subversion authors would manage to elimiate this license from their sources, because this license is not a good example to follow. Have a nice day, Charles -- Charles Plessy Tsurumi, Kanagawa, Japan -- To UNSUBSCRIBE, email to debian-legal-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/20150610234107.gd15...@falafel.plessy.net