On 6 May 2012 18:07, Peter Stuge <pe...@stuge.se> wrote: > Your proposed commit message says clearly what the commit adds to > libusbx.git, but the difference between my work and yours is not > clear while I am named as author with no mention of you, and that > isn't really accurate.
Happens all the time on every FLOSS project I know of. Someone does the bulk of the work, and somebody else adds minor bits and pieces onto it, and attribution is left to the person who did the original work. >> I don't see much point. > > The point is to assist anyone who is interested in code origin. Unless you are buidling a project for git historians rather than users, attribution is what matters. Also, what good would anybody get from this specific commit? You didn't discuss anything about timestamped logging on libusb-devel, so whether it was picked from git, or from a patch you would have submitted directly to us without mention of an origin, not much is lost from not referencing the commit. > Your efforts to reference the origin of work in the past (be it bug > reports or code or design advice) lead me to believe that we agree > that this is a good idea. A commit in libusb.git is not an origin. It doesn't provide any other relevant information than "yup, there was a commit there". A discussion in libusb-devel, which is something that didn't happen there, is one, and is something that I'd have referenced if I thought it was relevant. >> May I remind you that you have no authority whatsoever over what >> happens in libusbx. > > Sorry, but that's not true. I own my work which is part of libusb and > libusbx. Indeed, which through the license you entitle anybody to produce derivative work of, as long as the license is respected (and the license doesn't mention much about attribution or ownesrhip). As soon as you publish your code under LGPL, you lose all autority with regards to how that code should be used in derivative work such as libusbx, insofar as the reuse is carried out under the same license, which it is. Just like I lose all authority under any commits I produced for libusbx and which you may decide to reuse in libusb. If it wasn't the case, the production of derivative works would be very much impaired. > The license terms apply to that work, and both libusb and > libusbx follow those terms of course, but I remain the author of my > work with all that it entails including full ownership of my work Then please define full owership. Especially, can you define what extra this entitles you to, with regards to reuse of your code under a project that uses the same license? > while you are restricted to the authority over my work which you are > granted by the license terms. And this is exactly what I used. I reused your code as I saw fit, without modifying the license, and under the terms granted by said license. Please re-read the license and let me know which parts implies that, should I integrate it in a separate project and you have an objection, I am forced to honor your objection. Code published under LGPL can be reused in any other LPGP project in any which way the second project sees fit. > If you want to make it so that I really have no authority whatsoever > in libusbx then I think the only way is to remove every line of code > which I have authored, but note that I am absolutely not asking you > to do so. Please re-read the license. Or contact the FSF - they're quite friendly and will be happy to enlighten you on the subject of having your say on how (L)GLP code can be reused in any derivative that also follow the same license. > Who owns the code does matter every now and then; a project comes to > mind where this week someone wanted to change the license of a few > files in order to add them to yet another project. Not exactly accurate. As long as the derivative project uses the same license, who owns the code doesn't matter. This is one of the tenet of the (L)GPL: anybody is FREE to reuse someone else's code and produce a derivative work, without the original author having any say about it, as long as the license is respected. The ONLY place where ownership matters is when a derivative project wants to change the license, because of course, this can potentially break the terms of the license the original author of the code agreed to, in which case, the rights and duty that the original author granted may be modified in a way that the original author disaproves (and is rightfully entitled to disaprove), so they should have a right to make their voice heard. But that's the only case where ownership matters. > I think it's still a good idea to try to be consistently > clear about authorship, even for smaller changes such as adding the > thread id. The addon of the thread ID is minimal, so I don't care about attribution here. What I care is getting that change to our users ASAP. > You ought to have your name on the addition since you are the author > and since it isn't a trivial change. It's fairly trivial in terms of code. > I shouldn't be able to claim > authorship on code you authored, but if the patch is committed as > proposed then actually I can. And this was exactly my intention. The bulk of the patch is yours, and I don't really see the need to split it when the addon is minimal. The only issue I see is if someone contacts you about the code addon, but I don't anticipate it to be a problem (and you can of course reply to contact the libusbx maintainers about it -- as a matter of fact, as soon as someone reuses your code in a project you don't want to handle, you're free to tell them to contact the original project maintainers, as it is of course your right not to have to support the reuse of code that you didn't explicitly agree to). > If you don't care about that then of course I will be happy that you > assign your copyright to me, and then you can ignore my request, but > please confirm if that is indeed the case! It is, and has always been the case. If I commit something that contains some of my code with an attribution that isn't mine, then I don't have an issue with that other person claiming the copyright of the whole sectio, as the only impact this can ever have with regards to ensuring the freedoms enforced by the LGPL are respected, is if someone wants to create a derivative and switch license. But if that it the case, then I think the author of the bulk of the code should have the final say over whether they accept the license change. In other words, if it suits you to consider the the thread ID addon should be attributed to you, please consider it so. Or if, on the contrary, you consider that having your code reused and modified without your approval is not something you want to support, feel free to direct questions about it to libusbx-devel or ignore them altogether. Regards, /Pete ------------------------------------------------------------------------------ Live Security Virtual Conference Exclusive live event will cover all the ways today's security and threat landscape has changed and how IT managers can respond. Discussions will include endpoint security, mobile security and the latest in malware threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ _______________________________________________ libusbx-devel mailing list libusbx-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/libusbx-devel
