back
112 comments
I've worked at Microsoft and can attest that while it's OK to use open source tools internally the IP lawyers don't want anyone touching GPL v3 tools to do anything. If you suspect a project might have been even the slightest contact with anything GPL 3 you must call for an immediate code scan and legal review. You cant even use GPL 3 inside of an internal web tool. I imagine Apple and other proprietary software publishers have the same mandate to their troops.

So yes, GPL 3 is hurting open source more than helping it.

Disclaimer: I don't speak for Microsoft. These are my personal observations. And I no longer work there.

> So yes, GPL 3 is hurting open source more than helping it.

The GPL and the organization behind it, the Free Software Foundation, do not care about "open source". They care about free Software, free as in freedom.

/edit: before i get a ton of downvotes: i don't fully agree with with the FSFs policies or definition of "free softare", im just pointing it out.

Yet they restrict what you can do with their software. True freedom comes from licenses like MIT and BSD.
> So yes, GPL 3 is hurting open source more than helping it.

Why is it bad for the FSF to prioritize their interests ahead of Apple and Microsoft's?

There's a strong willingness to adopt open source tools in every large company but GPL 3 is poisoning the well, and that's impeding universal adoption of the open source ecosystem. Sure FSF doesn't care, and that's unfortunate, because that stance is actually keeping proprietary enterprise software on life support. Unintended consequences I suppose.
> I've worked at Microsoft and can attest that while it's OK to use open source tools internally the IP lawyers don't want anyone touching GPL v3 tools to do anything.

That's because they're Microsoft.

> I imagine Apple and other proprietary software publishers have the same mandate to their troops.

I have yet to hear any rational justification for this, yet people keep repeating it.

While Apple has not publicly said anything about it, it's pretty clear from their actions that they consider touching anything under the GPLv3 verboten.

Every single piece of software that has new releases under the GPLv3 they have stopped updating. A few of the most important ones, like GCC, GDB, and Samba, they've even gone as far as to pour substantial amounts of money and effort into funding replacements (clang, lldb, Apple's smbd). Others that they don't consider very important they've just let languish, like bash, make, and other GNU utilities that they include on their systems.

I agree with lambda that it is probably the patent clause.

Suppose I'm distributing GPLv3 software as part of my commercial product. I'm fully complying with GPLv3.

Some third party comes along who is in the business of distributing software, and tells me that they have a patent covering the GPLv3 software I'm distributing.

They want me to buy a license for their patent for $1/copy sold of my commercial product. This license would cover the use of the GPLv3 software in that copy of my product. It would not cover copies or derivative works made from those copies.

GPLv3 says that I must reject this licensing deal. I can only accept a licensing deal that results in all of the copies and derivative works made from those copies or their down streams and so on receiving patent licenses.

If I cannot arrange for this open ended licensing, I have to either stop distributing the GPLv3 code, or I'll have to try to defeat the patent.

No way am I going to distribute GPLv3 code as an essential part of my product with that "all or nothing" patent licensing restriction hanging over it.

> So yes, GPL 3 is hurting open source more than helping it.

It comes at no surprise that Microsoft, a company that vehemently spoke out against Free and Open Source Software for so long, would have these views.

GPL'd code is about keeping that code open source forever. Hence clauses in GPLv3 like anti-Tivoization. Little sympathy for companies that want to take without giving back.

I don't speak for Microsoft. I can only say there was plenty of willingness to use and contribute to open source but the whole GPL 3 question makes that confusing.
Before GPLv3, Microsoft extorted millions from companies by threatening them with undisclosed patents that claimed to cover all linux distributions.

GPLv3 was explicitly drafted to prevent this, and as by magic, Microsoft stop issuing the threats as soon the new license was released.

So no, GPL 3 is not hurting open source. It prevented a racketeering from Micosoft that was unethical, legally grey zone then, illegal today, and direct harmful to open source in every possible way.

>So yes, GPL 3 is hurting open source more than helping it.

The Linux kernel would not be the success it is without a viral license

And the Linux kernel stuck with GPL v2 for a reason...
Well, the GPLv3 is probably helping open source, and harming free software ;)

I would disagree with this stance though. The aim of free software is a moral one, it has nothing to do with being the most-used or the most popular.

FreeBSD et al seem to be doing fine without GPL 3.
Really not much to see here, as the author originally did not know Apple uses stone age FOSS tools because of GPLv3.

Sources here: http://opensource.apple.com

Apple doesn't seem timely in getting sources put up. Wayback Machine shows that 10.9.5 was posted around October 7th even though that update was distributed to users on September 17 (~20 day gap), so they're not exactly adhering to the license as they should.

Also there is no sources for 10.10 and does iOS 7 and 8 really use no GPL'd or other FOSS'd code?

> they're not exactly adhering to the license as they should

Correct me if I'm wrong, but the licenses don't require Apple to proactively publish sources, they just need to be able to supply them upon request.

Indeed, that's what the GPL requires. There is absolutely no necessity to put the sources online like they do.

Accompany it with a written offer, valid for at least three years, to give any third party, for a charge no more than your cost of physically performing source distribution, a complete machine-readable copy of the corresponding source code, to be distributed under the terms of Sections 1 and 2 above on a medium customarily used for software interchange; or,

Source: https://www.gnu.org/licenses/gpl-2.0.html

Apple considers iOS to be a branch of OS X, so I imagine they consider it covered by these releases. I can't really imagine what FOSS code they would feel compelled to include in iOS that's not already in OS X, anyway.
I find it very mildly ironic that the author is comparing these "ancient" versions to MacPorts. Am I naïve, or do the overwhelming majority of OS X developers use Homebrew nowadays? I've always thought Homebrew to be much more robust than MacPorts. However, I've never actually used MacPorts heavily (and probably never will).
There is a tool we use at work (rpmbuild, I believe) that's not available via homebrew and only through macports. It's a nightmare, because if you install macports, it interferes with brew (or at least, brew will complain a lot). I've taken to running a VM in lieu of macports.
Most developers I know who use OSX use brew; fink used to be the big thing, then macports. I know for some certain packages, macports is still preferred.
Disclaimer: I heavily contribute to Homebrew.

Homebrew actually merged in 4 unique bash patches in 6 days after Shellshock broke as well, and forced all users to recompile to ensure that hole was closed on our end. The author of this article would have probably been best checking both MacPorts and Homebrew, and potentially Fink as well.

Every developer I know in SF uses Homebrew, never seen anyone use MacPorts.
I used MacPorts way back in the day; Homebrew is about a gazillion times better.
Not mentioned and highly significant: Apple hasn't shipped anything which is licensed under GPL v3. That cuts them off from a lot of upstream development, including particularly upstream versions of bash.

I'm not sure they've ever published a reason for this, but outside speculation centers around the v3 patent grant and anti-tivoization clauses.

> I'm not sure they've ever published a reason for this, but outside speculation centers around the v3 patent grant and anti-tivoization clauses.

I've heard this more than once but it doesn't make any sense. GPLv2 says this: "Each time you redistribute the Program (or any work based on the Program), the recipient automatically receives a license from the original licensor to copy, distribute or modify the Program subject to these terms and conditions."

Ask your lawyer what happens if you license a piece of software to someone and then try to sue them for patent infringement for using that software.

And OS X doesn't stop you from compiling your own bash (or whatever else) and replacing the one Apple shipped, so I don't see how the anti-tivoization clause would affect them either. It just doesn't make any sense.

I half suspect that there are a few IP lawyers (and/or Microsoft) who don't like the GPL for ideological reasons and go around spreading FUD about it to discourage people from using it.

Is it significant though? I can’t really see why bash of all software should be critical. As pointed out elsewhere in this thread, other operating systems do not rely on bash being installed.
>Basically, the cause of the aging Unix tools in OS X is GPL v3.

That makes it sound like it's the GPL's fault that OS X users don't get up-to-date utilities. Blame Apple's bad policy, not the GPL.

It is the GPL's fault. There was enough change in the social implications of v3 that Apple, and a lot of other stakeholders, don't want to mess with it anymore. The FreeBSD team, for example, has very similar reasons for not including GPL3-infected software, and are actively replacing all GPL software with BSD-licensed software in the base install.
Apple can't, because they don't want to open source their code, which GPLv3 would require.

So, yeah, it is in this specific case the GPL's fault.

OS X wasn't vulnerable because it shipped an old version of bash. It was vulnerable because it relied on bash at all.
Yes. This point also highlights the distinction between interactive shell and system shell. You probably want a bunch of featutes in your interactive shell. You probably want a small, readable code base in your system shell. These two goals are generally at odds.
I was curious why OpenBSD didn't ship Bash by default. Perhaps they weren't happy with the code quality.
The way the article is framed, you'd think that it's somehow the GPL's fault that these applications aren't being updated. Seriously? Why doesnt Apple write their own implementations of those applications? Why shift blame to the license? What a crock of shit
that's how Apple works. They don't make mistakes, they remedy mistakes that others have caused--loudly--while hinting at someone else owning responsibility for the issue.

Sort of like the 'You're holding it wrong" response to iPhone reception back a few years ago.

http://www.cnn.com/2010/TECH/mobile/06/25/iphone.problems.re...

I kind of want Apple to steal most of the userland tools from FreeBSD like they have done in the past...
Seconded. If FreeBSD can replace GPL software, so can Apple.
As someone in the comments section of the original post put it, "newer doesn't mean better". Obviously software with security, stability, or performance issues should be re-written, but software doesn't "age" in the usual sense of the word; the source code can be re-compiled for brand-new machines/processors. A 2004 version of grep will always do everything I need grep to do.
Am I misunderstanding what the author means by 'real release' in terms of OpenSSL? OpenSSL themselves pushed 0.9.8za and Apple later adopted it after a significant amount of nagging in the developer's forums on the issue.
also make, which is stuck at 3.81 (2006)

sort appears to be (2005), and doesn't have the -R flag for shuffling

also missing: head -n -NUM (all but last NUM lines) and tail -n +NUM (all but the first NUM lines)

Is it a meta joke with the internal server error response?
Came here to post this.
It's amusing that Apple products ship with emacs inside.
I don't think it's amusing - I think it's great. I use emacs every day, and have for many decades.
Indeed. I love to show off tetris-mode to unknowing OS X users :)