9 ms·
I really respect many of Stallman's views but I still have trouble reconciling GPL'd software and sustainable commercial innovation. Doesn't the GPL level the
by flatline 15y ago
I really respect many of Stallman's views but I still have trouble reconciling GPL'd software and sustainable commercial innovation. Doesn't the GPL level the playing field to such an extent that the competition can undercut your business model with trivial ease? I would hate to put a bunch of work into something that I planned on making money from only to have perfect copycats overnight. Selling support only scales well for some software.
- jshen 15y agoI don't think your reasoning applies to most software. I'm guessing most software these days runs on the server. Think of heroku. If they open sourced their full stack they still wouldn't have to worry too much about copycats. It still takes a lot of expertise and work to run something like heroku, and they are always making progress which the copycat would have a hard time keeping up with.
- DasIch 15y agoIt works perfectly if you use GPL'd software for a web service or you don't actually sell the software but service for it.
- anon1385 15y agoIt works perfectly if you use GPL'd software for a web service I think RMS sees that as a loophole in the GPL rather than a desirable feature. Hence the existence of the AGPL. http://www.gnu.org/licenses/index_html http://www.gnu.org/licenses/index_html We recommend that people consider using the GNU AGPL for any software which will commonly be run over a network
- prodigal_erik 15y agoOr charge up front for the labor (which is inherently scarce) rather than writing on spec and charging for copies (which never needed to be scarce).
- wmf 15y agoIf the software is to have many users there tends to be a collective action problem in paying for it.
- dchest 15y agoWhat I understood (from one of his talks), Stallman doesn't care about your business model. As simple as that -- if you can't figure out how to make money while writing free software, that's your problem. The software may eventually be written by those who can figure out the business model, by those hired to write such software, or by those who don't care about making money. As for how this affects commercial innovation, I don't know. [Added:] I guess, the point is that human rights (according to RMS, proprietary software restricts freedom) are more important than innovation.
- 101001010101 15y agoSell hardware?
- polymatter 15y agoor you can simply write the software and give it away as part of your portfolio. A visible portfolio of good code will lead to more/better paid work.
- SkyMarshal 15y agoMore likely that human rights are more important than making money. GNU/Linux and its ecosystem has demonstrated that innovation is orthogonal.
- dchest 15y agoI meant "more important than any potential innovation gained from writing proprietary software". That is, if there's no more innovation in proprietary software compared to free software (like you said, with GNU/Linux) -- we're good, but if we're missing the innovation from such software, so be it, we're also good.
- pnathan 15y agoThis is my chief difficulty with the GPL as well. As a profession, we can't self-destruct with an ideology of giving everything away in a society which we have to pay rent, insurance, etc. I don't have a solution, wish I did.
- ajross 15y agoThis is a fallacy. It's equivalent to arguing that any increase in productivity (which is precisely what this is: less effort expended to develop software) is bad. But it's wrong. Over time, those freed from obsolete jobs are employed by society to do new stuff (stuff that wasn't practical before because of labor costs) and we all benefit. And even in the short term this is pretty much just plain wrong. Almost all GPL software is written by people paid to develop GPL software.
- Steko 15y ago"It's equivalent to arguing that any increase in productivity (which is precisely what this is: less effort expended to develop software) is bad." RMS is not proposing an increase in productivity, he's proposing a massive disincentive to engaging in one of the most productive, highest growth industries in the world's history. "Over time, those freed from obsolete jobs are employed by society to do new stuff ...and we all benefit." These jobs aren't obsolete. It's not like the same software will get done if all these people stop working and go do "new stuff". We just wouldn't get those advancements and would suffer greatly not benefit.
- icebraining 15y agoThe obsolete jobs are those which depend on re-inventing the wheel, because some company needs to add a new feature or fix a bug in software X and they can't, because it's proprietary. If every software was Free we'd eliminate plenty of redundant work.
- Steko 15y agoYou don't just eliminate the wasteful jobs; you eliminate the productive ones as well.
- rdouble 15y agoHis views on free software come from a very different idealized reality than what actually exists. From the GNU Manifesto: In the long run, making programs free is a step toward the postscarcity world, where nobody will have to work very hard just to make a living. People will be free to devote themselves to activities that are fun, such as programming, after spending the necessary ten hours a week on required tasks such as legislation, family counseling, robot repair and asteroid prospecting. There will be no need to be able to make a living from programming.
- maigret 15y agoWhich has some truth in it. Our world is already struggling to keep people employed despite continuous productivity gains. As developers, we know the next 20 years will probably be low risk for us, but many professions are endangered. Who will be a truck or taxi driver when the Google car will work fine? Who will need secretaries when Siri will be evolved? New technologies are just not creating that much jobs. Note that I am not disagreeing with you, I just suppose that RMS may not be completely wrong.
- jerrya 15y ago1995, Jeremy Rifkin talked about the End of Work, saying we should plan for how to structure society as it was coming. He was ignored, and now we seem to be living the results of not planning for it. http://en.wikipedia.org/wiki/The_End_of_Work http://en.wikipedia.org/wiki/The_End_of_Work The End of Work: The Decline of the Global Labor Force and the Dawn of the Post-Market Era is a non-fiction book by American economist Jeremy Rifkin, published in 1995 by Putnam Publishing Group.[1] In 1995, Rifkin contended that worldwide unemployment would increase as information technology eliminates tens of millions of jobs in the manufacturing, agricultural and service sectors. He traced the devastating impact of automation on blue-collar, retail and wholesale employees. While a small elite of corporate managers and knowledge workers reap the benefits of the high-tech world economy, the American middle class continues to shrink and the workplace becomes ever more stressful. As the market economy and public sector decline, Rifkin predicted the growth of a third sector—voluntary and community-based service organizations—that will create new jobs with government support to rebuild decaying neighborhoods and provide social services. To finance this enterprise, he advocated scaling down the military budget, enacting a value added tax on nonessential goods and services and redirecting federal and state funds to provide a "social wage" in lieu of welfare payments to third-sector workers.[1]
- Jach 15y agoIf you own the copyright over your GPL'd code, you can do a multi-license business model and sell/build proprietary stuff on top of the open core. (Or open stuff on a proprietary core, or a mixed core, etc.) When other people do the same thing, suddenly you have even more open cores out there (besides the ones done without profit motive) to build stuff on that you'd otherwise have to reinvent at a negative cost factor. Mixing open and proprietary, while perhaps not in line with RMS's views, is still a great way to make money, and you can even do it with the GPL in many circumstances. (Not always though.) In the gaming industry, I find philosophically pleasing the approach of Carmack and some indie companies who make a credible commitment to open sourcing the software (not necessarily the art assets) at a certain date in the future while making money on it at release. Edit: I also want to add that copycat fears are generally out of proportion. Copycats usually don't win, as in displace the original's dominance. (What usually happens though is the original starts copycatting others to keep a safe, steady income flow, and possibly to crush newcomers if the thing being copycatted isn't sufficiently novel. You're not going to get big by copycatting, though you might get a modest sustaining profit.)
- omouse 15y agoI think the multi-license model does work with RMS's views, it's just that you have to make sure that you're selling the same software; the GPL version cannot be limited or treated as trial software. Using a proprietary license for businesses means that you can punish them for refusing to use GPL'd software by making them pay more. When they accept the GPL, then you reward them with a lower price (but not free, you can charge for GPL software too ;-) ) hopefully encouraging more GPL'd software to be released into the world. The point is to reduce how much proprietary software there is and increase how much free software there is and multi-licensing uses a carrot and a stick based on pricing to do that.
- tptacek 15y agoOf the open source licenses, GPL is (ironically) the friendliest to commercialization because its terms make commercialized forks difficult (because it's difficult to close-source the fork). One reasonable strategy is to adopt one of the fascist GPL variants (like AGPL3) and then offer a commercial license alongside it.
- chc 15y agoBut then you can't accept any contributions, so you lose much of the benefit of the GPL, right? If there is going to be any community around the source, it will have to be a fork, and to you it will effectively be closed source since you can't relicense their code.
- rbanffy 15y agoYou can reassign the copyright of the contribution. I'd not be as willing to contribute to such a project, however.
- tptacek 15y ago"A community around the source" is not the biggest benefit of commercial open source. Before "whole features implemented by outsiders" factors in, you already have: * Drastically increased enterprise adoption because influencers w/o purchasing authority can download and do their own pilot deployments. * Free testing, with drastically decreased expectations of release quality (it's open source, what do you expect? &c). * The ability to market on the warm-fuzzy of "open source", which warm-fuzzy also includes the pragmatic issue of "what if you go out of business" (you still have the source). I think the reality of a lot of open source projects, particularly in the enterprise, is that the vast majority of the real code is written by a small affiliated core team.
- omouse 15y agoCommercialization isn't necessary for innovation. Stallman sees a world where software is free because it's abundant. There's a whole different system of economics there. Your current business model is based on our free market system where profits rule which means that giving away your software for free is stupid and unsustainable. Different economics means much less competition and much greater co-operation. Even if there are perfect copycats selling overnight, they're hitting a different market than yours. They may going after a local market or one based on a particular language