What did we do with all this free software? Used it to build things like facebook, google and various SaaS sites that not only tie our data into proprietary clouds but due to the nature of the GPL etc don't really have to share their code anyway.
What they have really achieved is to turn everything they touch into a commodity and moved the "value" of software elsewhere.
We seem to be moving to a world where most of our devices and the servers powering our apps will be running some form of Linux or BSD under the hood but we are actually more restricted than ever.
PS: AGPL is the strictest GPL there is, like, LGPL is a looser version of GPL. It pretty much means, you give away almost all the stack. I don't understand the whole license, but I do know it's the strictest.
However I have a feeling that there is a sufficient body of GPL/LGPL code out there that people will just modify that.
Since a lot of OSS code is contributed by companies, they don't really have any incentive to contribute to an AGPL project (in most cases).
The set of information people publish on their own about themselves (like blogs) is almost exactly the same as the set of information they disclose via Facebook.
Anyone can datamine people's blogs for similar surveilance.
.
From Stallman's blog & email history on newsgroups I could work out:
who he knows
what projects he participates in
his age
infer his sexual preference
infer his religion (may even be explicit there)
where he lives
where he was on any day (conference speaker history)
Those concerned with privacy have oodles of crypto-tools to do so with.
It's just people can't be bothered. That's the root problem.
--
Let's just see what can be done without Facebook:
You (pessimizer) have listed your email address pessimizer@xxxxxx .
You have an HN comment history.
Your email pops up in google under Arkansas
You commented on a Django topic.
I could infer your set of HN buddies (maybe not including myself now!).
If you use a pseudonym to hide from those - why don't you use a pseudonym as your facebook account??
--
I haven't listed my email, but I'm pretty sure people could work out who I was and where I lived from my comment history. But I am reasonably happy that the reward of not having to screen all my comments for incriminating info is worth the risk of someone figuring it out.
(edit with an example)
--
You (gldalmaso) don't list your email address, but you haven't used a unique username, so you pop up under twitter (your name is disclosed).
You have a StumbleUpon page which lists your age (27) and your hometown Flxxxxxxxxx, Brazil
You are clearly an Anime Lover.
--
This has all been automated, cached and waiting for me just to do a search.
Facebook has not been involved at all.
http://www.webmasterkb.com/Uwe/Forum.aspx/javascript/35764/For-all-those-jQuery-fanboys
https://github.com/derwiki/redirect_tracker
http://itunes.apple.com/us/app/ltwatch/id482807766
http://angularjs.org/
https://github.com/ludicast/angular-peepcode-todo
etc.
The problem is: FB has all these things together: blog, e-mail, network of people and surfing habits (inside FB, at least). For you and recursively for each friend of yours.
But I have no idea what to do about the people who just don't care.
For example, OKC recommended a person to me recently.
EDIT: Redacted a bit more.
She lists her blackberry pin; her facebook account; an email address; her cell / mobile phone number; and her twitter account.
That's enough information to find her profiles on a wide range of websites.
Remember it's not just IP address, but could be cookies. Facebook can set a cookie that will be stored in your browser and will be sent to facebook each time. So if your laptop moves around, then the facebook cookie follows you.
Since this is at the browser level there are browser extensions that will block this for you if you want.
I prevent the IP address leak by using the Firefox addon RequestPolicy to block cross-origin requests.
Open-source, proprietary or not, gives you control over your computer compared to closed-source software. It's not free software versus proprietary software.
Free software goes beyond open-source, and besides safety gives you freedom.
No it's not. This is just a few persons coming in buses and stopping the entry. If you want to equal it to protest then all the requests have to come from real people, not some bots.
I'm also not agreeing with this: "I won’t use the non-free software at all! I dedicate my effort to getting away from it! So if they stop making it – that would be great!"
This is ridiculous. I understand that the current IP legislation is a load of crap but trying to get ALL software to be free is absurd. How are developers going to live? How about groceries? Can I pay for that? Or that should be free as well?
Now (regarding other more strong notes on this thread) - Stallman has his opinions and he supports them in a very logical way (albeit a bit too passionate). Moreover, he tries to support personal freedom, which in my books is more of an American value than anyone's else. So, trying to dump the real issue here by referring to what RT may be (according to what sounds like post modern cold war psychosis) is at least cheap propaganda.
The term 'free' is very misunderstood in this context. Its free as in freedom/free speech, not (necessarily) free as in beer.
Although sometimes (most times) its both. This doesn't stop you using it to earn a living. It allows you to freely use and modify it to your own purposes. Contrast that with non-free (i.e. locked, obfuscated and proprietary).
I've written freely available stuff that people have (never the less and willingly) paid me rather handsomely for the privilege of using or modding to their own needs. They didn't have to, but people can be inherently decent that way.
I could have made it entirely closed and I think I'd have made less out of it if I had. I would have had to market it for a start -- and that's a fucking headache. I'm not a salesman and don't want to be.
Anyhow, Stallman is 100% right. Everything he's been warning us about for years is already upon us. With much worse to come.
Well it can be way worse than just a simple IP. When you ask for a picture your browser send a bunch of infos about you that can be dangerous because it gives a footprint. If you want to see that go to that page ( hosted by EFF) it shows you how it can be used to fingerprint you: http://panopticlick.eff.org/index.php?action=log&js=yes
Also there are other attacks that can be used to go even further using for example browser cache. The browser cache has a field that can be set by the server and generally is the date for the expiration of content you are asking. But... when first designed it accepts a random string, so for example an UUID... This cache cannot be cleaned with normal procedures and you are tagged without your consent.
Other techniques can be used with Flash and cookie revival has been actively performed by companies like Quantcast. (look it up on the net)
So... what Mr. Stallman has said is true to an extent that only few people know and that's a pretty big deal.
A few points:
1) To Stallman, having the 4 Freedoms[1] to any software you acquire (paid or not) is an Human Right. And of course, you don't violate human rights just because it employs some people. So your question is completely irrelevant to him.
2) It's Free is as in Freedom, not Beer. You can charge for free software and in fact he encourages you to charge as much as you can. Of course, any of those buyers might start distributing it for free, but on the other hand, people could have also bought the Humble Bundle for almost nothing and yet they chose to pay a decent amount.
3) You're discounting the software - possibly most of it - which is produced either in-house or by a company contracting with another to write it. If a company needs some software which doesn't exist yet, or to add some feature to an existing FOSS package, they'll pay.
In fact, I have friends which work on a company which makes money by adapting Free Software to others' needs.
4) You're leaving out value adds. Red Hat makes money, despite CentOS. Reddit makes money, despite having a repository with all the code. If your software depends on a service, you can give away the software and charge for the service.
1) If you are busy protesting by blocking a street this causes you inconvenience as you can only physically be in one place at once. Whereas leaving your computer on to DOS while you go out to a bar isn't exactly a hardship.
2) Many of the DDOSers would not even be remotely aware of what they are doing , see the JS worm that they used recently.
This is ridiculous. I understand that the current IP legislation is a load of crap but trying to get ALL software to be free is absurd. How are developers going to live? How about groceries? Can I pay for that? Or that should be free as well?
You've just opened up a huge topic there, has been discussed on HN lots of times and while I have seen good arguments for copyright-less software in many areas (OS kernels , web frameworks etc) there are others where I don't think anyone has thought of another viable business model (at least not one that isn't even more freedom restricting in some way).
Some would argue that these areas should just disappear or be done only by hobbyists but I think I would miss professionally produced video games for example, indie or AAA.
To clarify though, I doubt that Stallman would support Anonymous or people who want to pirate software. To him any software that is not libre is irrelevant and should be rejected regardless of monetary cost or who distributes it.
You only have a finite amount of bandwidth, and you have to decide how much to use for DDOS and how much for your own use, so, in a way, the analogy of street blocking still holds there.
The Free in Free Software stands for freedom. He is not making a statement on pricing, but on liberty.
Furthermore, I must say that his DDOS argument is a lot more valid than you give it credit. After all, it IS possible to DDOS a site without malicious intent, just ask any celebrity on twitter who tweeted about a site he or she liked. And who is to say that some protests today don't consist of people who have been either paid (western countries) or forced (eastern countries) to attend them?
The whole point is to temporarily break a server someone's paying money for.
I.e you can sell the software, but anyone you sell it too has the right to just give it away for free.
Some business models are impossible without slavery. Does this mean slavery should be allowed?
Stallman's argument here is the same: if the particular software you make can't survive financially without abusing people, then you shouldn't have a right to make that software. And he considers all proprietary software to be abusing people's human rights.
So if you want to counter him, saying that it will put developers out if work is no more valid that saying that abolition of slavery is bad because of the financial hardship it would impose on the cotton industry.
To counter Stallman, you have to come up with arguments showing that non free software does not abuse its users.
let me give you an example the first few pieces of FSF software was distributed by RMS by computer tape and he charged for that and that revenue paid for his living expenses.
The other big example is Redhat/JBoss..everyone of their workers get paid through services charged..
It's only actually completely voluntary if you have infinite money and no issue (and no repercussion from) completely dropping whole social and professional circles out of your life.
And that still requires the network/system is honest about it[0], when Google automagically creates a G+ account for you when you sign up for (supposedly) unrelated services or forces you to create one to access other content, "voluntary" is really debatable. Again, unless you have no issue shedding whole social circles instantly.
[0] http://arstechnica.com/gadgets/news/2012/01/google-doubles-p...
People were social long before Facebook was around, and can remain so without it. Even with an account, the information you share on Facebook is voluntarily given.
I think what Facebook has done most effectively is given people the illusion that their lives are somehow more social because of it.
Although employment and social circles DO limit your range of actions, you accept the tradeoff voluntarily in exchange for benefits. You can find a new job or people to associate with if the benefits do not justify the cost, without your employer or friends being able to legally threaten you with violence. On the other hand if you refuse to allow surveillance or investigation by the law, they are authorized to use as much force as is necessary to seize your property, imprison you, or execute you. It's completely voluntary in that regard.
"Voluntary" can be viewed as an antonym for "coerced", where coercion implies an underlying threat of violence. Forced denial of an individuals life, liberty, and property can be construed as violence or coercion. Such properties do not presuppose the presence of another person or governing body, so may be considered intrinsic to the individual. However, employers or friends denying the deliverance of their property or empathy to you cannot be construed as violence or coercion. Such exchanges presuppose the presence of multiple people, and therefore cannot be considered the denial of a property intrinsic to one individual.
You are not forced to share anything. If you want to use Facebook as strictly a messaging platform, there is no negative effect for doing so.
Characterizing it as "surveillance" is RMS's typical good-natured extremism.
You only have to pick up the phone instead of using the computer.
[edit]
G+ doesn't change the surveillance situation at all. I have a G+ account and I don't do anything with it. Google doesn't have any additional information.
Imagine I start a FREE postal delivery service for writing letters to your friends. But when your letter shows up at the destination, it is full of advertisements that I hand picked for your friend based on the content of your correspondence. Oh yeah, also the police asked for a copy of the letter. Oh yeah, and we also showed the letter to a bunch of your other friends because we changed our delivery rules and figured you wanted to opt-in. Didn't think you would mind.
But you can't complain. The service was free and you signed up voluntarily.
I'm neither pro / con on the Facebook thing. I never know when the real world rules should apply to the electrons. I think we are just making it up as we go.
My personal email. My work email (can infer where I work from this). Many of the people whom I know who searched for me on facebook. Highschool friend links (can infer my highschool and possibly even age from this). I know for a fact that they have photos of me tagged with my name.
I didn't authorize any of this, yet there it is. Never before has it been so easy to aggregate this sort of information, or perhaps even possible.
Also, some might say that GPL software was successful precisely because of the server-side loophole. That's not necessarily bad, because the cost of starting your own Google or Facebook is much, much lower than 10 years ago.
# Block Facebook
127.0.0.1 www.facebook.com
127.0.0.1 facebook.com
127.0.0.1 login.facebook.com
127.0.0.1 www.login.facebook.com
127.0.0.1 fbcdn.net
127.0.0.1 www.fbcdn.net
127.0.0.1 fbcdn.com
127.0.0.1 www.fbcdn.com
127.0.0.1 static.ak.fbcdn.net
127.0.0.1 static.ak.connect.facebook.com
127.0.0.1 connect.facebook.net
127.0.0.1 connect.facebook.com
127.0.0.1 www.connect.facebook.net
127.0.0.1 apps.facebook.com
Sure, you could probably find that out by googling; but Facebook makes it significantly easier.
Saying that Facebook hasn't changed our privacy is like saying that firearms haven't changed the face of war. I mean, it just puts metal through your body just like a bow and arrows.
I also don't want to turn gldalmaso into some sort of cruel hamster to argue a point. Suffice to say the poor guys twitter feed is completely open. You or I could easily deduce all the answers to your questions based on his past tweets and followers.
If he didn't use twitter, then I could use blog comments.
If you visit facebook.com, a cookie will be set, then later when you visit another site with a facebook widget, it WILL send that cookie that was set earlier when it wasn't 3rd party.
Also I don't think where we're at the stage yet where losing your whole internet connection temporarily (or just having it slow down) are that much of an inconvenience to life.
If you have a 10Mb pipe and you could use 90% of it for DOS for an entire day without noticing much inconvenience (unless you want to use torrent or stream HD video).
For example you could plant them in the ground and use them to grow more potatoes or you could chop and fry them into chips, bake and serve them with with chilli etc.
His issue is not really with the cost of software (that is more a side effect of the GPL). He takes issue with the fact that with software you often have artificial restrictions in use and that the manufacturer may include features that are not to your benefit (e.g DRM , spyware , adware) and you can not remove these without breaking the license agreement.
He would liken this more to buying some potatoes that can only be legally used for one purpose and if you wish to use (physically identical) potatoes for another purpose then you must pay a higher fee.
The commercial issue with the GPL is that if you give people the rights to distribute as they see fit there is guarantee that they will give anything back to the original author.
Personally I would love to get applications with source code available that I can modify as I wish (or just fix bugs) but would require that the original author was paid a fee upon re-distribution (of original or modified version) to someone who did not already hold a license. I see this as a very good compromise in many cases.
The problem with this of course is that if I did wish to distribute the software to an unlicensed person then I could easily remove any copy protection methods prior to doing so.
In such a case ironically the best solution might be stronger copyright legislation to protect the rights of open source but non gratis software developers.
"“Free software” does not mean “noncommercial.” A free program must be available for commercial use, commercial development, and commercial distribution. Commercial development of free software is no longer unusual; such free commercial software is very important. You may have paid money to get copies of free software, or you may have obtained copies at no charge. But regardless of how you got your copies, you always have the freedom to copy and change the software, even to sell copies."[1]
So yes, you can farm your own vegetables. But you shouldn't accept that you buy a potato without being able to plant it to make more potatoes. (Ironically, you should check back on that with Monsanto, though[2]).
[1]http://www.gnu.org/philosophy/free-sw.html
[2]http://en.wikipedia.org/wiki/Monsanto#Terminator_seed_contro...
But one can trivially block these autoinclusions.
---------
http://webgraph.com/resources/facebookblocker/
--
My work's firewall directs Facebook.com and Twitter.com to /dev/null . This stops all such tracking.
Individuals could do their own blacklist via Greasemonkey / Charles proxy equivalent.
Paid for OSS means less software shops, to be replaced by a few developers paid to produce some OSS (say, RedHat paid employees, people paid to work on the Linux kernel etc) and most other developers just paid to integrate and customize it (say, programmers from startups to huge enterprises using RedHat).
And with the "Like" buttons you /are/ forced to share your browsing history (no, being able to circumnavigate it by blocking it doesn't count)
Also, singling out Facebook for what every advertiser cookie has done since the mid 90's seems a bit silly.
However, one thing is certain, I think. Some technologies may have a transformative effect: The fact that you could act in manner X to get result Y before the technology gained widespread use does not mean you can continue to do so after it has. I'm sure this was true for the telephone, and I'm pretty sure it's true for Facebook and its replacements.
Those same people are regularly out of the loop in conversations their friends have on Facebook, which at times makes them feel a little isolated.
I've heard of one person who made a huge social faux pas because they couldn't read about someone's breakup on Facebook.
The same are all true of some people I know who live on LiveJournal.
Facebook is a great tool for socialisation, and because of that a lot of people use it. When your social circles use it heavily and you don't, you miss out. Yes, you can do all that socialisation outside of Facebook, but the fact remains that a lot of people don't, and you can't really choose how your friends socialise.
It's not an absolute. You're unlikely to become completely socially ostracised because you're not on Facebook, but it can certainly make some peoples social lives more difficult not being on it.
There /is/ social pressure for some people to use Facebook.
It's the same as asserting that people can communicate without phones, or that they can get news without the internet, the phone or newspapers. Sure, they can, but it's incredibly inefficient, exponentially so when everyone else still does.
(I've still managed to stay away from Facebook, but, then, I don't have many friends.)
For better or worse there is a growing danger in FB regarding the tons of personal data that the same users would just never give away if it weren't for FB. He is just trying to make users aware of that. He's certainly not the first to do it and hopefully not the last. Also he said nothing that could be interpreted as innovation stopper.
Out and over.
Anyway - he is Stallman and he's known for not being extra diplomatic or even smart at politics. Like you, he speaks the way he feels - point is that IMHO there is a growing problem with all the data that we - FB users - carelessly upload.
(Moreover I'm not convinced that FB has all that much to do with innovation any more but that's an entirely different conversation.)
PS: Sorry if I sounded hostile but the unabomber reference drove me mad. My bad.
PS2: this was supposed to be an answer to a comment that's not there any more - anyway.
What "tons" of personal data? I doubt I could myself collect tons or personal data about myself, not to mention data that would be dangerous.
I bet that you're quite computer literal yourself (hard to suppose otherwise since we're talking through HN) so as to be quite cautious as to how much of yourself you'll expose on the net. Unfortunately most people aren't plus they are quite good at documenting their lives (photos,videos,writings, etc) and linking in with people they know and share data and the like.
That's "tons of personal data" - it's not an absolutist firework but it's the main reason that gives FB such a high market value. AFAIK FB is considered a gold-mine in terms of highly targeted advertising and marketing that's because it hosts "tons of personal data" of high granularity and of high relevance to the real identities of the people uploading it. Otherwise wordpress for example would be considered of at least equal value with FB.
Point taken, and partially agreed with, but the fact that facebook is systematically doing this about every aspect of every individual's life would guarantee that getting better info than you got about me, in a shorter time, would be trivial.
I imagine there are data-mining companies that are crawling the interwebs systematically to collect detailed information.
Facebook is the motherlode of course, no doubt about it.
Read their term http://www.facebook.com/legal/terms they are things like "transferable, sub-licensable, royalty-free, worldwide license to use any IP content that you post on or in connection with Facebook"
As Stallman pointed out in the interview, the scale of the monitoring is on a different level of what you can do manually. No matter how hard you try, you will not know what websites pessimizer has visited, FB got a list of his/her visits on any site that contains a like button (doesn't matter if he/she uses FB).
Beside the scale of information disclosed, that's how it can be used or not. I'm pretty sure I can decide what products are endorsed on my blog, not so much on FB, and if we visit pessimizer site, no one will now. On FB, they could use this information to know pessimizer better. Guilty by association ;)
Probably they have some common heuristics to spot a fake, but those cannot be too clever.
---
Re: Scale of monitoring
Does this not just make the haystack bigger for the same number of needles?
But it is not a problem owned and invented by Facebook, it is any network.
I'm not taking sides here, it just occurred to me that portraying DDOS attacks as definite cyber terrorism is a problem in the discussion we have today and I think it's a slippery slope, similar to the "piracy" argument we hear so often.
The problem with getting worked up about DDOS is that it isn't technically possible to make a clear judgment from it - that's what I was stating. Let me put it like this: A real-world protest can be thousands of people standing in front of a building and thus making it hard for them to do business or it can be smashing in their windows. A cyber protest can be linking thousands of people to an article on a website that you don't like and reducing their quality of service - or it can mean causing their servers to melt.
There are shades of gray in this discussion that you exclude and it is not doing the nuanced point that Stallman was making justice.
Warning: If any of the following is incorrect, I would really appreciate someone correcting me. I use the AGPL, and if I'm incorrect, I would like to know.
You never have to push your changes back upstream. I'm not aware of any license that requires you to submit your changes back to the original author. You just have to make it accessible to your users.
Just like the GPL, you can run AGPL code on a private network and not have to publicly release the source code. It's when the network is public that you have to make the source code available.
If I visit your website, enter in some data, a script/lib licensed under the AGPL does something with it and returns a result. You must release the source code of "everything that it touches".
If you have a business and on the company Intranet you have the same script, while the employees (users) are entitled to have a copy of the source code, you are not required to publicly release the source code. Of course, other laws prevent those employees from also releasing the source code to the public and AFAIK they "trump" copyright laws eg.NDA.
Since a lot of OSS code is contributed by companies, they don't really have any incentive to contribute to an AGPL project (in most cases).
I don't know, companies contribute to GPL licensed code even when they don't explicitly have to. Just off the top of my head, tarsnap comes to mind, he contributes back to libarchive even though he is not required to do so.
That doesn't mean I don't see your point though. You could just as easily suggest that if libarchive was licensed under the AGPL that tarsnap wouldn't have used it in the first place due to the fact he would have to release the source of the entire stack.
So I think you argument should really be "Why should companies even use software under the AGPL" not "Why should they contribute back". I don't really have an answer for that, but the same argument was made about the GPL and companies still use GPL licensed software. I'm sure if you asked rms that question he would say something about not caring about companies that want to restrict his freedom....
As for requirements to push back changes - of course not. The only license I know of that does something similar is the SMF - Simple Machines Forum - license[1]. Yuck.
The argument between GPL and AGPL is similar to the one between BSD (or similar) and GPL on the desktop. My own position is that yes, I expect humans to be inherently good, kind and sharing souls and thus the BSD should be the license of choice. However, I have to be pragmatic and say - if we all assume that humans act like that, what's the harm in writing that down into the license? All it can do is make it hard for people who don't want to give back. So in my eyes, that's an important feature to have.
Yeah, sorry. I think that was a poor choice of words on my part.
For instance, if you have a website that uses AGPL code to produce an HTML page and non-AGPL code to provide a PDF from that HTML page, the two packages would indeed touch, but there would be no obligation to release the non-AGPL code, just the AGPL code.
Interesting, in your example, if your application calls the AGPL code that is used to generate the HTML wouldn't your application also have to be released under the AGPL? Or do you only have to make available the AGPL "HTML generation library" and the changes you have made to it (if any)?
I thought that it had already been established that if you use a library under the GPL any code that uses/calls that library must be released under the GPL. The only reason companies can use GPL code server side is because it is not run on the user's machine (technically not "releasing it"), which was the basis for the creation of the AGPL, to "fix" this "loophole".
edit:
Thanks skore, yeah I think your example still stands. Specifically, in the wikipedia article you linked this part:
By contrast, pipes, sockets and command-line arguments are communication mechanisms normally used between two separate programs.
To continue the original example, if your program just runs the HTML generation program and supplies user data via a command line argument then it need not be released under a compatible license. But on the other hand if you were to "copy and paste" the functions from that program into yours, it would need to be licensed under a compatible license.
The AGPL basically establishes the same concept on the internet that the GPL has established on the desktop.
I see, interesting idea unfortunately I really can't see it taking off, most companies would just prefer to work internally and then release some bits of code as LGPL/BSD when they are ready. What advantage could AGPL give the developer (assuming they are writing OSS for commercial rather than ideological reasons)?
The commercial advantage is that it levels the playing field: If I create a library that can be used for a website and release it as FOSS under a license that is not as restrictive as the AGPL, I basically create a disadvantage to my own business: Another company could use my code in a closed service and thus benefit from my work without giving me and their users the courtesy of benefiting from their work in return.
I write commercial GPL software and will, in my next refactor, license certain parts of the software as AGPL. An example would be a library that connects an invoicing system to an online payment processor (ie. creates a full checkout process for invoice X at amount Y). I have a lot of work invested in the code (basically 5 years of my life) and the better and more reusable I make it, the easier it is to just retool it for another application. I would naturally love to see websites utilize the code, even if it is just in a SaaS setting. But since it is my work, I have decided to set the rules and if people want to use it, I think it's fair to ask that they should promote the same liberties that they received from me.
Wikipedia has a pretty good section on the more broad legal background:
http://en.wikipedia.org/wiki/GNU_General_Public_License#Comm...
I wonder what would happen if you modify the GPLed command line program. Would that mean you'd be required to release such changes? I would say no.
I would be more concerned about the "viral" nature of the GPL (I assume this applies to AGPL as well) i.e I would be happy to use AGPL code in my application and release any modifications I make (or create AGPL code and have others do the same for me) but there are parts of my applications that I really wouldn't want to have to release (because they are very industry specific and would be useful to competitors without modification and not useful at all to the general populance).
A similar license but one based on LGPL would be preferable.
Of course, if you accept patches from others, that's no longer true, unless if "force" people to dual license their code, with the second license being an exception just for you.
A similar license based on the LGPL wouldn't make sense. The LGPL is basically the GPL with an addition that says that you don't have to be that strict on what license the software you use it in uses.
Problem is I'm not enough of a lawyer to know what would and wouldn't be compatible with the GPL (and confident enough to tell other stakeholders). With something like a BSD license it's easier to point to concrete examples of it's co existance (OS X for example).
There is another special case, but it's selling software in spite of the GPL; home routers, set top boxes, etc. Were they to use GPLv3 software, they'd have no way to protect against another company using their (potentially substantial) work on the software, building/copying the hardware design and creating cheap knock-offs within a few weeks of release, making it too costly to continue.
If a group of companies and individuals wanted to come together to build a new router platform where all would contribute back to it, but they could differentiate themselves on edge features, management interfaces, etc, the BSD license is good enough; and probably a better bet than the GPL.
There seems to be a phenomenon on HN of hardcore anti-copyright or pro-piracy posters who reply antagonistically to posts that they appear to have either deliberately misread or not read at all.
Software that used to be sold commercially but was subsequently GPL'd , starving artist type programmers scraping by on donations (i.e earning significantly less than the median programmer salary for someone of their skills) or companies that produce GPL software but make money selling either support or software/hardware based around their GPL software (that itself is not GPL) don't count.
Companies that sell software generally have services divisions as well. My company sells software for large amounts of money, and we have a separate division of the company that sells services.
If a company exists to make money, why would you discount one of their profit centers as invalid because of another of their profit centers?
I assume that the software your company sells for large amounts of money is not GPL , if it is pure GPL I would be very interested to know who they are and how they manage to get people to actually pay for it when they could just download it for free online.
If a company exists to make money, why would you discount one of their profit centers as invalid because of another of their profit centers?
That's not what I'm saying, what I'm saying is that I've never seen a company create GPL software as a profit center in itself and I really can't think of a business plan that would make that feasible.
This means that there are categories of software where GPL does not really fit at all, games for example.
The reason we need to use a license like that and not a libre license is that some of our software is very niche and would only be useful to a handful of companies around the world.
It's quite likely that none of these other companies employ any programmers at all bar the occasional contractor. Basically we are trying to disrupt an industry (in a small way) by getting more efficiency by doing stuff through tech that they are doing by employing legions of admin staff and using the old-boys network (it's a very old school industry).
If the source for certain parts of our software was available to them then they would most likely simply use it (employing an Indian contractor or similar for a couple of weeks to set it up) to wipe out our advantage and not contribute much if anything back to us.
There are however other parts of the software that could be generally useful to a wide number of people who are much more likely to contribute code back.
We aren't planning to distribute any of our software so we get along fine just using LGPL / BSD licensed stuff and a few commercial libs for the most part.
Something like AGPL is potentially interesting but if it involves having to hire lawyers we'll probably keep things in-house or maybe release a few bits under LGPL a few years down the line.
That is blatantly false. Who gives a shit if they are selling software, or selling "support" for software? Either way people are making money with GPL software.
There seems to be a phenomenon on HN of absurd pedants deliberately derailing conversations.
"You do realize that people sell GPL'd software all the time, right?"
You specifically said sell and I asked for examples.
"sell"
So sue me.
They sell licenses and services, and you can pay for quicker updates to firewall rules and the like. As commercial customers, you can have their software engineers onsite tuning things to your environment, etc.
Covalent is a company that sells Apache. Obviously, they didn't manufacture Apache webservers, but, like Xamp or Wamp, they sell value added bundles that include Apache configurations pre-compiled with PHP or Perl or what have you.
Again, they also have a services division, but I don't personally know much about it.
For the record, no, my company does not sell GPL software. It was proffered as an example that non-GPL-based companies rely on both software sales and services, and it doesn't really matter whether or not the software is GPLd. I don't know why you'd hold it against RedHat that their software is freely downloadable when they're still selling more than a few copies of RHEL.
Not so familiar with Covalent but it appears they have some things available as GPL but if you want all the enterprise features then it isn't GPL.
Redhat is the same really, they also make their money from support. I believe you need to buy RHEL if you want the support from them, although this might have changed now.
I never suggested that GPL cannot be part of a business plan , I was just disagreeing with the statement that there are companies that sell GPL software, in all cases I can find what you are really paying for is something else not the software itself.
My point is more that there are many areas of software that this does not work in , games or most consumer software being an example.
http://www.gnu.org/licenses/why-not-lgpl.html
"Using the ordinary GPL for a library gives free software developers an advantage over proprietary developers: a library that they can use, while proprietary developers cannot use it."
Sourcefire SELLS the product, flat out. If anything, they're more of a 'freemium' offering where the free product is a GPL product, and the 'premium' is a faster release cycle for patches / upgrades / definitions.
Covalent sells Apache. I've dealt with Covalent products at a number of federal installations, and I've never once seen a Covalent services rep. If their aim is to sell services, they're not doing a good job.
There are a cornucopia of other examples as well, but these are the three most fitting the description you claim doesn't exist. They do in fact exist, and are making money selling GPL software. That they also have services divisions has nothing to do with whether or not they're making money from selling software.
GPL software is in no way a barrier to earning money. It's, in fact, crucial to most software today, even if indirectly.
Your correct that a huge amount of software today uses libraries that are distributed under some form of libre licensing although in most cases this is LGPL , BSD or Apache style licenses rather than GPL. Sure perhaps they run some stuff on a GNU/Linux server but in most cases this could just as easily be a BSD , Solaris or even Windows server.
There are many types of software where GPL is absolutely a barrier , namely just about anything where the software is the end product. For example pretty much any video game or for example Adobe Photoshop/Illustrator. Some products simply don't lend themselves well to selling support , very few people are going to pay $1/min to get help running a program that is simple enough to figure out on their own from reading the manual and that should 99% of the time just install and run especially if it's not something mission critical that requires 6 nines uptime.
Could be, but isn't [1]. There's a bunch of embedded GPL code in routers and various little bits of all kinds of systems that enable our current lifestyle, too.
I won't argue that for most (though not all) actual software intended to be sold, GPL probably isn't the best of ideas. But that doesn't prevent earning money while using GPL software.
[1] And certainly hasn't. I won't overlook the historical effect GPL'd code has had.
This is an important distinction as I don't believe they could do what they do under GPL which is provide proprietary stuff on top of the open source core.
Sourcefire is selling the faster access to releases of rulesets etc, the GPL software is simply the carrot.
RHEL contains software that is not GPL and I doubt that they would stay in business without selling support.
The GPL license was designed specifically to stop people developing proprietary software on top of GPL software.
As for the other two, you're still picking nits. The companies wouldn't exist without the software, and they sell it. There's still no good reason for WHY you would discount them as being irrelevent just because they're also selling other services.
The crux of my argument is that with most current OSS licenses (GPL especially which is what Stallman advocates) companies can only be commercially viable by supplementing their OSS offerings by providing other products that are proprietary (i.e you are not free to redistribute) or by providing additional services on top.
I think it is a stretch to say you are selling GPL software when what people are paying you for is actually something else that is supplemental and if you didn't provide the extra then they wouldn't pay you anything. There are many good open source based businesses that is not in doubt but equally there are entire parts of the software business that simply couldn't practically release their flagship software under GPL, the games industry of course being the biggest example that springs to mind.