Exactly my feeling. My thesis is that FSF achieved its popularity in the heyday of Microsoft monopoly serving as a counter-force, and as Microsoft's hegemony declined (thanks in part to FSF's efforts), it has become less relevant. Open source (I intentionally do not want to use Stallman's language on this) is thriving but FSF is not.
GPL is way too dogmatic, and given the fact that most free software actually is written by people with a paycheck from those hated proprietary software companies (at any rate, companies that definitely do not buy into FSF dogma), there always has been an underlying tension between their philosophy and the vast majority of contributors to open source. Basically Stallman wants them to feel dirty getting that paycheck.
As a result, we are moving to a different phase in the open source world - focus on code, not on the politics.
I want to add an illustration of the perverse effects GPL produced in practice. MySQL - now in the hands of Larry Ellison ... there must be some irony there - perfectly shows how to use GPL in a very "interesting" way. I have dealt with MySQL salesmen on several occasions, so I know how they "use" GPL. Basically they would assert with full confidence that your use-case (whatever it may be) would violate GPL, so you have to pony up unless you want all your code to be GPL. They knew how to exploit "GPL fear" in companies, but would back down in most cases once you bring a lawyer.
That is the practical real world effect of dogma. My prediction: MySQL will be orphaned intentionally by Oracle. No one else is going to touch it because forking the MySQL code will not let anyone escape the famous Oracle salesmen.
"GPL is way too dogmatic, and given the fact that most free software actually is written by people with a paycheck"
I don't think anyone, even Stallman, thinks making money from free software is a bad thing.
I worked at a company (that you've heard of) who basically ran their business on Free Software saving them God knows how much time, money and effort. The CEO announced one day (in public) that he didn't like the GPL as it was business unfriendly. He made no effort to hide the fact that he wanted to take code written by others (often for fun) and convert it to profit without giving _anything_ back. The GPL is a wonderful license.
I don't understand your mysql example. It sounds like the GPL did its job in that case.
It's been hard for me to pin down RMS's exact thoughts on these kinds of things. I think the whole "copyleft" made him and the FSF as a whole seem more communist or anarcho-syndicalist or something, when I doubt that's the actual intent. But I can't say for sure.
My most recent computer purchase was a modest dual-core PC with 350 GB hard drive. It cost only $350 CDN without an operating system - but it was surprisingly difficult to find an assembled computer for sale that didn't come bundled with an OS.
I've been thinking about this for some time. Microsoft these days is pilloried (rightly IMHO) for being closed and proprietary; but the reason DOS originally won is that it had the cheapest, most open and permissive licence - i.e. it was the closest (at the time) to free software.
It was this openness that earned Microsoft its market monopoly on the OS and consequent positive network externalities - benefits that the company has leveraged to maintain its dominance of the desktop for two decades.
So what I'm proposing is this: perhaps the FSF did well during the reign of Microsoft not as a foil to the big, evil monopolistic corporation, but rather because the FSF is the logical conclusion of the very strategy Microsoft itself has used to establish and maintain its market share.
I can see your point but I think the GPL is more important than ever. I'm not sure "most" free software is written by people from hated proprietary companies. It's a little more complex than that. For example I love IBM, they sell proprietary software and also contribute heavily to free software. In what sense is the FSF not thriving?
Many of the open-source licenses (I say open-source deliberately to distinguish them from GPL) are commercial friendly by design, .eg. ASL. Since we all need to earn a living we make compromises. Many corporations for better or worse prefer these licenses as they enable them to benefit from the collaborative model without having to give back the value they add. Some do this in a very ethical fashion by employing major contributors to projects, thereby by giving and getting, all good. However I would argue that these serve the interests of companies more than individual programmers.
So a question I have is, if the GPL is no longer relevant as many now claim, why was Git released as GPL?
I can totally see Oracle sales folks playing this FUD argument. As a programmer I typically look at sales people as what they are. They will say anything to make a deal.
That may all be true but we still have a valuable code base to fork off of if we so please and plenty of independent companies and people who will provide professional support for it independently of Oracle.
What happened to MySQL happens to plenty of smaller projects whose original devs / maintainers get snapped up by companies to work on a commercial closed-source version.
There is no license which will insure the future of a project but if the code is popular enough you will always find someone to pick up the mantle.
I clarified my point above: forking MySQL will not solve the "GPL-in-the-hands-of-Oracle-salesmen" problem. Any forks will still be bound by MySQL's original terms, which in practice amounted to "You need to pay us if you ever ask us about the license, particularly if you are a rich company, unless you bring a lawyer to argue with us." Keep in mind that Oracle salesmen are 100 times better at it than MySQL salesmen.
My broader point is really about the problems of GPL in the real world, not open source in general. I am a huge fan of BSD/MIT/Apache, and thankfully a lot of contributors these days seem to like those licenses.
I have always viewed open source contributions as a form of charity. GPL is a highly encumbered form of charity (it is still charity, but think of charities tied closely to particular religious points of view). BSD/MIT/Apache is much more freely and liberally given and embody a generosity of spirit ("My charity recipients do not need to follow my religion"). My philosophical preference is the latter.
>BSD/MIT/Apache is much more freely and liberally given and embody a generosity of spirit ("My charity recipients do not need to follow my religion"). My philosophical preference is the latter.
Said in much more succinct terms than I could have mustered.
We use MySQL at work. After a major infrastructural change I think I'll be able to get us to move to Postgres. It might depends on whether or not we have any fires to put out at that point but we'll see. Postgres is better supported by the framework we're moving to. shrugs
I'm not sure what, if any, contributions I can make to this cause. When I have anything worth releasing, I try to release BSD, but I'd really like to be able to contribute to llvm. I don't know that I necessarily have any skills that would be useful to them.
GPLv3 is horridly dogmatic. A great number of the old hackers are on some kind of corporate payroll, frequently Red Hat.
I'm not certain I'm 100% comfortable with the current corporate patronage system because it has allowed people like Ulrich Drepper to dig out his private little Hell with glibc and torment anyone who begs a moment of his time.
The vast majority of computer users, I would suppose, have no concept whatsoever of software freedom. A lot of software is still proprietary, and people are still running into problems just like RMS did years ago with being unable to fix the printer driver software.
The software landscape has changed, but software freedom remains essential. The FSF has been perhaps a bit slow to respond to some of the changes (e.g., the AGPL probably should have been developed years earlier), but the underlying message stays the same, and, I believe, is increasingly important as software usage continues to waft into every conceivable facet of our lives.
The AGPL is a completely unenforceable EULA, both practically and legally. It attempts to restrict your ability to run the software -- freedom 0
The AGPL will never ever succeed. It will only ever be used by egotistical idiots and dying companies that wish to sear a scarlet letter into their code -- noone but I may profit from this work
With a strict definition of "profit" this seems more a criticism of dual-licensed GPL works than the AGPL; software published under the AGPL does not grant any sort of monopoly to the person publishing it -- even if they were selling a proprietary version of the same code, it doesn't restrict third parties making money from it.
And if we weaken our definition of "profit" to mean some sort of benefit, then the argument is even weaker.
Third parties attempting to extend the software and sell access are intended to release source code. The only room for profit is in network effects -- any other attempt to sustain margins more than barely above actual costs would be tenuous, as someone else could come along and undercut you.
The copyright holder(s) that licensed their work under the AGPL are not subject to any such restrictions (no multi-licensing is necessary). There's no way to compete with them fairly.
Okay, so you're essentially describing a situation where the copyright holder offers access to a "pro" version and a "vanilla" version, with only the latter being licensed under the AGPL?
I would certainly concede that this scenario would make it difficult for third parties to maintain feature parity, but that's a long way from saying the AGPL offers "no way of competing" with the copyright holder; writing features that are not present in any version would be an obvious starting point. (You also grossly misrepresent the market for "network effects", but that's somewhat tangential)
I don't know whether it's actually true, but I find it surprising. It sounds like Black Duck Software runs some weird web crawler that collects license data continuously.
Too young? Do you have any idea how old the Teamsters and the IWW are? The IWW was involved in turn-of-the-century populist anarchist agitprop and labor activism.
Fer chrissakes, do you think nothing has changed since then?
I make $20 an hour to code, and a guy I know makes $80/hour (he started at that wage!) to tape drywall and carry out garbage.
No offense, but unless you are a grad student and this is your stipend, you are horribly underpaid. The problem is not unions, the problem is where you have chosen to work.
Depends, if you mean to say that it's going to be replaced by things like the BSD or Apache license, then that's plausible.
But a full-on death for OSS is not going to happen.
The Creative Commons / Free Culture movement is fairly recent and only growing in popularity and it's easy to see how it dovetails with a lot of the FSFs work in software.
People (and very large corporations alike) are clearly still seeing the value in liberally licensing their works so that lots of people can enjoy them.
>Depends, if you mean to say that it's going to be replaced by things like the BSD or Apache license, then that's plausible.
That is precisely what I meant.
>But a full-on death for OSS is not going to happen.
That's ludicrous, not what I said, and not what I want. I would carry the torch all by my lonesome if that actually happened.
>The Creative Commons / Free Culture movement is fairly recent and only growing in popularity
I don't care about either/or. I care about software.
>People (and very large corporations alike) are clearly still seeing the value in liberally licensing their works so that lots of people can enjoy them.
Mostly as a means of marketing/promotion in the hopes that people will pay for the stuff that they can't steal outright. Don't kid yourself otherwise.
Useful, nay, necessary at one time, but have now outgrown their original mandate and now exist unto themselves.