On Mon, Nov 21, 2016 at 3:51 AM, Stefano Zacchiroli <[email protected]> wrote:
> On Sun, Nov 20, 2016 at 03:35:07PM -0500, Martin Blais wrote: > > > If you don't want to impose your views on others, GPL2+ is really the > > > only right choice here. > > > > That wouldn't work: If someone decides to make a copy as GPLv3, I > wouldn't > > be take the changes back into my original v2+ version. Linus specifically > > addresses that issue in his response. > > But the Linux kernel doesn't have any CAA/CLA, which you seem to want > for Beancount. If you do have a CAA/CLA in place, the blocker for > including rogue changes that are made to Beancount without coordinating > with you will not be the license; it will be the lack of willingness to > sign the CAA/CLA. > > In general you will encounter either coders that value integrating their > changes back into the upstream Beancount code base (so that it is you > maintaining the changes and not them), or coders that don't care and are > fine with keeping maintaining local changes. In the former case it will > be pointless for those coders to insist their changes are GPL >= 3, > because they know you won't accept them. In the latter case you'll find > plenty of other blockers for code inclusion before hitting the license > one (in fact, very likely you won't even know those code changes exist, > because their authors will probably not distribute them in the first > place). > Based on my roughly 20 years of experience developing my own and contributing to other people's open source software, I'm convinced that what you're claiming is inaccurate. My experience is that there are a few ideologues who make a lot of noise, and a vast majority of people who are just trying to get things done and who are grateful to get things for free and contribute the occasional patch back to the project within the framework established by the author, given a reasonable license (such as GPL). Most users initially try free software without even looking much at the specific terms--most people are interested in solving some problem they have first and foremost, "git clone this and that, let's try this out", and at the end of the day, they use what works best. There are many examples of thriving OSS communities which require a CA and I've never heard anyone complain that it's putting brakes on their development. Except for a few notable instances, most OSS development really centers around a core of a few individuals in each project anyway (<10 usually, even for large projects). And even the FSF requires one for contributions, for example to Emacs. The Clojure community is very healthy too, in that way, and there are others. I still remember when I signed those projects' CAs; it was a tiny hurdle, but it made me think more seriously that I was sharing away that code, and at the end of the day, it made me proud that I was contributing something back to these projects--if anything, officializing my participation in that way only served to make me feel closer, to make me feel more of a part of those communities. There's something nice about it, even. And my itch to send back that bug fix for the change I really wanted to see integrated upstream has always _greatly_ overridden the little bits of annoyance of sending back a signed scan by email. About GPLv2: Git and Linux are examples of very successful projects which require only the "GPLv2 only". I'll keep investigating the idea of upgrading the license from "v2 only" to "v2+ or later", but I'm really not worried about estranging a few people here and there because of that. So far I'm a bit uncomfortable with some of the v3 differences. You have to interpret my reaction to this in context: I'm not interested in being part of a movement; I'm not an activist, I'm just a guy who likes to build things, things that are useful and original and which solve my own problems first, and if it's easy to do so at the same time, to spend more time to make those solutions also usable and useful for others and share some of what I've learning in doing this. I'm not _trying_ to create a "community," if it happens, it's something that just happens by itself, and I hope it's because the code works. I'm really just trying to write code to build an awesome accounting system that I can use, scratching my own/ itch, and taking pride in doing so elegantly, there's no stronger motivation. I wouldn't do this otherwise. I do like it when other people find what I've built useful and I try to help them if I have time, but in the end, success is defined primarily by being able to solve my own problems. It's completely fine with me if some small number of people don't like some small detail about the terms of distribution. I think at the EOTD we just disagree on the license issue and let's just leave it at that. -- You received this message because you are subscribed to the Google Groups "Beancount" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/beancount/CAK21%2BhPOz1TXTpLvg4jSj1gNrMDF2QtjSCu3bgw0%3DjuN%3D8P1Yg%40mail.gmail.com. For more options, visit https://groups.google.com/d/optout.
