Slashdot Mirror
GPL, Copyleft On the Rise
paxcoder writes "Contrary to earlier analyses that predicted a decline of copyleft software share to as little as 50% this year, John Sullivan, the executive director of the Free Software Foundation, claims the opposite has happened: In his talk at FOSDEM 2012 titled 'Is Copyleft Being Framed?,' Sullivan presented evidence (PDF) of a consistent increase of usage of copyleft licenses in relation to the usage of permissive licenses in free software projects over the past few years. Using publicly available package information provided by the Debian project, his study showed that the number of packages using the GPL family in that distribution this year reached a share of 93% of all packages with (L)GPLv3 usage rising 400% between the last two Debian versions."
The earlier study looked at a much broader base of projects, not just cherry-picking by limiting itself to packages in a distro.
Let's call it what it is, Anti-Social Media.
The reports are all wrong and based off personal bias because there really isn't any good data. But we put it in PDF forum so that makes it a "paper" not blog jaw jacking. Keep reading suckas!!!
IMO, if you're writing or releasing software, the GPL is preferrable. You benefit from patches, even being able to take those people don't intentionally contribute. You keep your code unusuable to those competitors who follow a closed management model. You also get to use it as advertisement if you're willing to offer an alternate license for money.
If you're looking to use somebody else's software though, of course the BSD is best. But the thing is that once you spent a few months working on code, a BSD license can be a bit of a hard sell for anything important, because you have nothing of the above. I think for most people some degree of attachment and desire of control develops after spending a lot of time on something.
So one study, which looks at the wide ecosystem of open source software finds copyleft is on the decline. But a study which only focuses on a Linux distribution which has a strong focus on GPL finds copyleft is increasing? Isn't that a bit like going to a Green Peace rally and saying a majority of people surveyed support saving whales?
I've seen so many developers just slap the GPL on their code because it's perceived as the "default" choice. When asked why they chose to use the GPL, they can't even explain its basic provisions. When told how it works, many of those same developers will say "oh, that's not really my intent." Sadly, because of the original "default" perception, a ton of code gets licensed this way.
I aggressively support the right to license something any way creators see fit, and happen to license my most of my stuff under the BSD and Artistic licenses. That said, people really need to understand what different licenses provide before they run off using them. When in any doubt whatsoever regarding any of it, it wouldn't be a terrible idea to pay for an hour of a lawyer's time (if possible).
Write failed: Broken pipe
So one study, which looks at the wide ecosystem of open source software finds copyleft is on the decline. But a study which only focuses on a Linux distribution which has a strong focus on GPL finds copyleft is increasing? Isn't that a bit like going to a Green Peace rally and saying a majority of people surveyed support saving whales?
You left out the part where the pro-GPL study comes from the authors and advocates of the GPL.
I think the FSF might be a bit biased. Don't you.
While non-copyleft licenses like the Mozilla, Apache, and LGPLv3 are quite popular for core services and libraries, most applications I've used over the years were copyleft/GPL type licenses.
If you're building a core service, you want it used by as many people and projects as possible. But if you're developing a tool, utility, or application, often your concern is more to prevent any one company or individual from seizing that work and selling it as their own product.
Personally I use both LGPLv3 and GPLv3 licenses as a result, because the goals of the different software components are not the same.
I do not fail; I succeed at finding out what does not work.
Who would have thought?
That doesn't apply to software that's intended to be sold to end users, which can take advantage of the GPL. For the rest, there's the AGPL. Google doesn't seem to like it.
Actually that's in a way a benefit. Part of what I like about the GPL is precisely the situation with the kernel. By mixing together so much code from authors that would disagree with a change, are unavailable, dead, etc, it'd take the rewrite of a huge amount of code to relicense the kernel, to the point it's not worth trying. It exists in a weird category of its own where nobody really owns it, and nobody can ever become the owner. I consider that state to be desirable, even if there are problems like with the GPL3.
They like it better for sure, but why would be that a good thing for me? Darwin last time I looked at it was unusable and pretty much dead, for instance.
BSD projects benefit from patches and contributions, both from individuals and corporations.
Exactly, which is why all of MacOS X is BSD licensed. Wait.
You clearly must have missed the part of the GP that says "even being able to take those people don't intentionally contribute."
GPL is perfectly usable in a closed management model when the code is used internally, for example when you provide a service not a software product like google.
Except that if you're concerned about that then you use the AGPL.
Second, it is a political belief, not a fact, that denying access to the close management model is beneficial.
Your opinion regarding the empirical consequences of closed management is rejected on the grounds of it not being a fact.
Your license it for money under an alternative license argument is in conflict with your patches from 3rd parties argument, you can not license code that others own the copyright to - look at the Linux kernel being locked into GPL v2 because all the contributors of patches and new features/functionality can't/won't authorize a switch to GPL v3.
OK, they're in conflict. Which one does that disprove? It's neither of them, isn't it? It's that you get a choice between reincorporating changes from unwilling contributors (without copyright assignment, as with the Linux kernel) or requiring copyright assignment from contributors before incorporating their contributions so that you can license them on different terms while preventing prospective licensees from just taking the code and incorporate it into proprietary software without your consent. You don't get to do either of those with BSD.
There's absolutely no point in releasing it under both the GPL and the BSD, since if you use one of the new BSD licenses (like the one used by FreeBSD), it's completely GPL compatible.
Dilbert RSS feed
it isolated and emasculated the platform in one easy step.
when the gpl was first applied to the linux kernel it was still a hobby project, :) and perhaps it might've; but that doesn't mean it hasn't achieved much under the gpl, even if quite a lot of computer users don't know of it. popularity is not equivalent to achievment.
to compare linux today to linux then is hardly to see an "emasculated" linux, whether on the server or desktop, you might think that if linux had a different license it would rule the world - mwahaha
snake
The GPL poisons commercial code -- intentionally -- and that keeps GPL'd software from ever bringing mainstream software [...] Those big packages everyone wants, [..] simply can't afford to mix in with that kind of licensing.
Yes, that's so right. Look at how Oracle became free software straight after they ported it to Linux. Bankrupted the company too.
Ok, I know, here comes the mod-bombing, lol. :)
I know I know. The mods here; so damn biased. What next? Discrimination against Goatse posters?? I think you both have an equally valid reason to demand to be modded up. In fact I'm surprised that the BSA hasn't managed to put through legislation to guarantee that for you.
=~ s,(.*),<sarcasm>$1</sarcasm>,g if any_point_you_wish();
Actually, for a commercial entity, the GPL (or other copyleft license) is really the best choice--for releasing their own code! It means that your competitors can't make and sell a version with proprietary enhancements and gain a competitive edge on you based on your own code. Any improvements made by your competitors have to be shared with you, meaning that both of you gain an advantage over any third competitor who isn't participating in the sharing.
Ooh look, a liar.
Good way to completely incorrectly representing how the GPL works.
Bullshit, plain and simple. There are LOTS of non-GPL packages, proprietary packages even, that run on Linux.
And for so blatantly lying and deliberately misrepresenting the GPL you deserve it.
If the University or project is publicly funded, i.e. funded by taxpayers, then you should not go the GPL route or dual licensed GPL / paid license route
Why should Japanese users benefit from a work funded by British taxpayers? Could something be dual licensed where royalties for a paid licenes are due only for copies distributed outside the funding government's jurisdiction?
I would choose a license that is very easy to understand in a few sentences, like the BSD license.
Ever considered the Sleepycat license? It's a copyleft like the GPL, but it's short like the BSD license.
-- This is why the "year of the linux desktop" never comes.
The year of the linux desktop arrived over a decade ago. The second most popular Unix desktop is a free variant, Linux, which was the original goal to offer an alternative to Unix desktops. The most common server OS, a huge player in the embedded space and the dominant OS for supercomputing ain't nothing to sneeze at either.
I don't know of any UNIX which has been as successful on this many fronts. Desktop is hard because the first and 2nd place commercial offerings are excellent.
MidnightBSD currently has 2754 ports. Of those, 1194 are under some GNU license (gpl2, gpl3, lgpl variations). A good chunk of that is for GNOME, KDE or GNUStep. Only 127 are under GPLv3.
I believe them that GNU licenses are as popular as ever, but I doubt that GPLv3 will be the most popular for some time. Many projects have been downgrading their licensing to GPLv3 including core GNU projects, Samba, etc. There's still a lot of old code or code that hasn't been updated in awhile under GPLv2. Frankly, the linux kernel isn't even under GPLv3 which is huge.
Disclaimer on the stats.. the way our licensing framework works, anything that's dual licensed is unlikely to show up. Perl code is dual licensed under artistic and gplv1 and that's not in the stats. Also something like firefox is always under the superior MPL license. I avoid GPL when possible for accounting purposes. It's a defect in the system that it doesn't support multiple license types (and uses agg for that)
Outside of the GPL, I've seen a lot of new code coming in under the Apache 2 license. I think that's the most popular outside of GNU circles now and possible surpassing the BSDL.
MidnightBSD: The BSD for Everyone
A great example of mixing proprietary software with GPL software is VMware ESXi. The hypervisor rides on Linux, even the Virtual Machine File System is proprietary, yet the virtual disk file is freely available and implemented by multiple vendors. There are ton of patents that VMware owns for the ESX hypervisor and ESX can peacefully coexist with Linux.
It is also worth noting that VMware makes contributions to the Linux kernel, voluntarily.
Did I hear someone say, "Industrial commons for the digital age"? I think I did.
The diversity and expression of human opinion is essential to human survival.
the GPL was the very worst thing that ever happened to linux -- it isolated and emasculated the platform in one easy step.
And in one fell swoop, the troll is slain. Emasculated indeed. ha
The soylentnews experiment has been a dismal failure.
Research labs I have worked in are often hampered by the viral nature of the GPL. The GPL seems to assume that the value is in the software, not what it produces. We develop biomedical research software that we would like to open source to non-commercial users, while having a license fee for commercial users. Companies might make make significant amounts of money on the output of the software, without ever reselling it or even needing to modifying it. To us, it seems fair that they contribute back with license fees to help fund more development. The viral nature of the GPL, along with no one being sure what use is infectious, causes us to avoid build with GPLed software. Frustrating when there is some fine GPLed software that would be useful and to which we would gladly contribute changes.
> ESX can peacefully coexist with Linux.
There's also the fact that Linux is (and will forever be) GPL2, and not GPL3. That's not always the case with other FLOSS.
There's a big one everyone knows about: Android.
Buy an Android phone, and you get the source code to the kernel and that's about it. The rest of it? Nope, it's not under the GPL. Google does eventually release most stuff to AOSP, but they don't need to. It's not a requirement. And the code they do release isn't GPL'd, it's under the Apache license. So much for the "viral" GPL.
Actually no, it doesn't, although it's an easy misconception. The hypervisor runs on the bare metal. in ESX you have a (redhat-derived) custom VM that is used as the system console for controlling the hypervisor. The console VM isn't viewable or manageable like the guest VMs but its underlying implementation is the same. Because (in ESX 3.0 and earlier) significant configuration procedures could only be performed through the Vmware commands accessed through a bash shell (as opposed to VIClient and/or VCentre), VMware admins generally assumed that the hypervisor ran on Linux, but the reverse is true. Since ESXi4, the system console has been stripped down to a minimal VM with a busybox interface, and configuration is performed via either VI Client or the VMware PowerShell extensions. In VMware 5, the ESX version with the full Linux console VM is no longer even sold and only the ESXi version with the stripped down Busybox console VM is available.
Laissez lire, et laissez danser; ces deux amusements ne feront jamais de mal au monde. - Voltaire
If you are talking OSs, there is no kernel that is GPL3 - maybe Hurd will be, once it's released. Aside from that, the GNU userland is GPL3, but that doesn't get used by VMware. So for what VMware needs, Linux plays the role just fine. I'm sure that something like NetBSD or OpenBSD would have, as well.
This is the FSF's position, but some people think the requirement to preserve the BSD license text is a "further restriction" relative to GPLv2.
Even if those people are correct (one might argue that the text the BSD license says must be preserved falls under "appropriate copyright notice", or is close enough that a court would not find it a substantial breach of the GPL), explicitly dual licensing is useful to satisfy said people's paranoia, especially if you want your code merged into their project.
Of course, if we're going to start interpreting the GPL that strictly to the letter, section 2a of GPLv2 gets violated constantly...
I'll reply to you.
You nailed a key issue: If a company wants to write their own 100% proprietary blob, sure. Have at it.
But if they swipe "free code" (aka GPL) then whine about losing "competitive advantage" if they release their end result, that's the abuse we should stop.
"Hi. GPL is enforced by a version of Copyright Law. Remember how much fun you had with that? Now pay up. Or, MAYBE if we are nice, drop the code you swiped and White-Room it from scratch."
My first Journal Entry ever, in 8 years! http://slashdot.org/journal/365947/aphelion-scifi-fantasy-horror-poetry-webzine
You're not kidding about how easy the misconception is. I've installed ESX many times thinking that I'm installing something derived from Red Hat Linux. I did a little more research and now I can get a better sense of what they have accomplished using Busybox. I guess the only question left in mind that would be relevant to this thread is this: does VMware use the GPL version of Busybox or the BSD version that I've read about in the press?
The diversity and expression of human opinion is essential to human survival.
The year of the linux desktop arrived over a decade ago.
What year was the year of the linux desktop?
His whole response was a typical response of someone who doesn't understand the issues. Linking may still be a derivative work despite what the intention of the licence is. The FSF has been unwilling to clarify that this does not make a derivative work, and that it will take no legal action in such cases, last I heard. So this means that, no, the LGPL is not enough, if you want to absolutely sure. There's also the issue of bits of compilers and parts of headers getting compiled into the program.
a} OSX becoming UNIX based, caused Linux to become a solution in need of a problem. Linux is still seeing some use in the embedded space, but that's about it. The only reasons for anyone using Linux on the desktop are a} being a Stallmanite fanatic, or b} somehow being ignorant of the infinitely more desirable alternatives.
I expect the usual enraged responses to this from Ubuntards in particular; but I'd encourage you to save your energy, guys. Your distro sucks, and no amount of foaming at the mouth on your part is going to change that.
b} With the release of version 3 of the GPL, the Free Software Foundation effectively committed suicide. At the time, Linux was already on its' way to becoming less mainstream than the fandom of Battlefield Earth, and all that really did was hammer the final nail into the coffin.
As a result, we get no new Windows games ported to Linux, Ogg Vorbis is considered the domain of autistic nerds, and producing open source hardware drivers is considered more pointless and unnecessary than ever.
Way to go, FSF. I get the feeling that if you'd actually been trying to destroy FOSS' chances, you wouldn't have been able to do a better job.