Mac OS X Kernel Source Now Closed

littleghoti writes "Macworld is reporting that "Thanks to pirates, or rather the fear of them, the Intel edition of Apple's OS X is now a proprietary operating system." Mac developers and power users no longer have the freedom to alter, rebuild, and replace the OS X kernel from source code."

Software pirates won't care

[Myria]

Software pirates will just use IDA Pro [datarescue.com] instead of GCC to get the job done. The part they've always cared about (Don't Steal Mac OS.kext) was never open source anyway.

If Apple says that software pirates are the only reason, don't believe them.

Melissa

BSD vs GPL

[Dan Ost]

And now we know (well, we always knew) why Apple chose to use BSD
userland vs a GPL userland.

This should add more fuel to the debate of the merits of BSD vs GPL
lisencing.

Re:Apple might be worse than MS.

[shotfeel]

So you really do believe that open source software should be viral in that anyone who uses it should be forced to open up all their code?

Not that there's anything wrong with that -there are good arguments in favor of that view. The problem is that corporate reality sucks sometimes.

Obligatory: Will someone fork the last open code?

[multiOSfreak]

I have to wonder if some group or other won't go back to the last open version of the kernel code and fork it into a new project or maybe some alternative to Darwin? Also, what does this mean for the Darwin project?

Would something like that even be worth it without some vendor support or tie-in? It seems a shame to let such a nice chunk of code go to waste.

Initial impression

[rjamestaylor]

Mac developers and power users no longer have the freedom to alter, rebuild, and replace the OS X kernel from source code

Good, because we know how that's helped Windows reach it's apex of security.

"If your OS is secured by keeping the code private, pray it's never, ever, released." Only takes one slip into the public to break that "security model."

Then there's those OSes that *assume* publicity of the source code and have different expectations for ensuring security. These "published" OSes also happen to be the "more secure" OSes available.

Go figure.

P.S. I'm not only referring to GPL'ed and BSD'ed OSes. There are other published OSes, the source of which are publicly accessible.

Disclosure: Mac OS X user here. Linux user here. Reluctant Windows user here.

Re:BSD vs GPL

[alistair]

I'm no expert, but I'm not sure this is true. The kernel for Macs and BSD is very different, the MACH Kernel is no BSD kernel. The parts Apple took from BSD relate to Networking and the user tools we often use from a shell, i.e. the shell and common unix commands most Mac users play with from time to time. The diaply code (Quartz, Aqua etc.) was their own and I think they have kept this closed.

For the BSD stuff they took, they wern't required to post anything back to the BSD communitity but my imprssion is that they have in every case. I don't think this would have been any different if they had taken a GPL equivelent, unless the GPL prevented them linking to a closed source kernel.

The code they have taken for Safari was GPL and I think they have contributed back to this. There have been numerous discussions around this as they did make huge changes optimed for Power PC which they contributed back but were of very little use to Linux on Intel and I would be interested to hear what people think now they have contributed back their Intel code.

I have to say that I am no expert in this, working mainly in the identity and directory field. However Apple's work with Directory Servers and Clients [apple.com] is on a par with the open source contributions of SUN, Novell and OpenDirectory and something I watch with great interest (and far beyond what I would expect from a company which mainly makes home based Macs and iPods.

Re:Good one Apple ...

[goofyheadedpunk]

Even though I own an iBook G4 I don't want to!

I don't like closed systems. I don't like being given the One Way to Do Things, even if that One Way has obviously been well thought out. I don't like farting around the internet looking for software. I don't like installing all of the cruft that comes with Xcode just to get gcc. I don't like being unable to link the command line with the GUI.

I like my Freedom. I like being able to dick around with my system (at any level) when I'm bored if I feel like it. I like the feeling of doing clever things with source code. I like having centralized repositories of software. I like using a system that's been designed for ease of development.

I like GNU, I like Linux and I like being in control of _my_ computer. Granted I'm not a typical user by far, but we're out there.

Re:Sad day

[Space cowboy]

The deal here is that Apple is a company, not a bunch of (talented) hackers. They're here to make money, pay employees and execs, and (hopefully, 'cos I'm one) raise value for their shareholders. They're not a charity and they're not there to make the world a better place (well, Steve may disagree... let me rephrase: they're not there to make the world a better place for free).

The reason I'm replying to you is that you say "But they did grasp how to utilize open source to their advantage, but it was always in a way that was really not quite in the spirit of the open source community". I think that's unfair. Just because they don't want to lose control over *one* piece of s/w doesn't mean they don't get it - indeed they may "get it" all too well, if they're planning on releasing server-based machines in the near future... you don't really need much more than Darwin to have a server, so they probably would lose money to people self-building and self-installing "clone" machines...

Where they see there is an upside for them, I think they've been reasonably generous - Webkit (despite some initial negative feedback, they responded and made things better); there's a story about how to use Quicktime Streaming Server to get MythTV on your cellphone elsewhere on the main page; they put a lot of effort into gcc; etc. etc.

I don't think you can expect much more from a company - so it's not a 'sad day', they do indeed 'get it', and as you say - it's their right to do things as they see fit. I think they do more than most...

Simon

This fucking pisses me off ..

[Sonic McTails]

I know I'm in a minority, but I used Darwin/x86 quite a bit since it had NetInfo support so I could use it for shared login, and while I could switch to everything to LDAP, it wasn't worth the effort. I currently got an Intel Macintosh, but maybe my next purchase won't be a Mac, because I do/did use Darwin quite a bit. That being said, the Macworld UK article doesn't cite sources, so where is it getting this info? I still see the xnu sources on OpenDarwin's site:

http://darwinsource.opendarwin.org/10.4.6.ppc/ [opendarwin.org]

Re:TPM

[GweeDo]

A link to them having included TPM would be helpful. I was under the impression that Apple didn't include TPM in the new Intel Mac's.

Maybe a change to solaris eventually?

[Master Of Ninja]

If they are locking the source down, I for one would hope that it might indicate making a move to solaris. Especially after the recent news that they were porting some of Solaris' file system over to MacOS, moving over to full solaris may leverage the best of unix and the mac os GUI system.

Appel.org

[nanojath]

If they did it would be appel.org

Hmmm... http://www.appel.org/ [appel.org]

Anyway, precisely. Apple's business model is basically to be Sony (Expensive component systems that only talk to their own kind) but they get away with it because the stuff works in a way Sony only dreams, they have this ironclad against-the mainstream, shinyfunhappy thing going (sorta like VW), and they leave the most important points of generic interoperability (i.e. iTunes and iPods play MP3s) open. Darwin was not one of these.

Re:Duh!

[Pius II.]

Nope. If you didn't see this coming, that's because it happened months ago. And just like months ago, the xnu ppc source code is still being published.

What's more, I have personally used the ppc kernel source to compile an x86 kernel. I haven't tested it, since I lack an ICBM, but I'd assume it would work.
The only difference I noticed was that the official x86 kernel includes Rosetta, while my self-built kernel didn't[1]. If I was to take a wild guess, it'd be that Apple does not have the right to distribute the in-kernel parts of Rosetta, and accordingly cannot distribute their x86 xnu branch.

[1] Note that there may well be other differences, but I hacked on binary loading stuff, so that one really caught my eye.

Re:so what you're really sayiing is...

[argent]

So what you're really saying, based on all your comments attached to this story, is that OS X was never really suited to high-performance/server/scientific computing?

Anyone who's used an XServe knows that. It's not and never has been a high performance beast, it's a decent small server that's targeted towards Windows-class admins with fewer Windows-class annoyances.

useful purpose

[r00t]

I happen to like fixing kernel bugs. It's fun, and it makes the bugs go away. (not suggesting that Apple should delibrately add extra bugs just for the thrill of fixing them though) Kernel source is educational too.

Oh well. I can still judge a Mac on hardware alone, and then install Linux if I get a Mac. That's what I did last time I bought a computer.

It sure irritates me to see BSD groups actually helping proprietary vendors compete against open source. Thanks buddy. Stallman got at least one thing right.

Did anyone ever actually recompile?

[ickoonite]

This Macworld writer is a fucking idiot. I'm willing to bet that the number of people who actually recompiled their kernel on Mac OS X can be counted on the fingers of, say, two hands. For that reason, this is a total non-issue. And as others have noted, this has been the case for quite a while and, well, most of the source is still open anyway.

Would that we could concentrate on some real news for a change.

iqu :|

Re:I don't believe this is because of pirates...

[blakestah]

I thought that too. Cringely says, here in his 420 column [pbs.org], that Apple has Darwin running with a full Windows API implemented in-house, and has the rights to release this from a prior Microsoft/Apple agreement. The possibility is that all Windows apps would run in OS X natively. Closing the Intel-Darwin source under a "security" excuse could be exactly what they need to do to upgrade everyone's machine to run Darwin with a Windows API, possibly on a native BSD instead of a BSD over Mach...

minor clarification on 'piracy'.

[nblender]

Some comments here indicate that people aren't entirely clear on the state of OSX86 piracy at the moment. Basically, it amounts to groups of people having started with the early release OSX86 code, and hand-integrating the Apple released patches back into it, as well as adding whatever drivers could be found in the Darwin sources, and/or binary patching Apple's kext's to produce, what today, is a 10.4.6 bootable/installable DVD that works on lots of non-Apple hardware. Indeed, I installed it on a corporate Dell laptop that my employer insists that I use.

Here's the problem, performance sucks relative to my Intel 20" iMac, it hangs frequently, and the network driver can't read the mac-addr. I also can't set the mac-address using ifconfig, so end result, is no networking. Screen resolution is also not able to match what the screen is capable of so the aspect ratio is wrong.

In short, while it's a cute hack and the novelty of seeing OS X running on Dell hardware is certainly nifty, it's far from production ready. Why did I dare to anger the Apple gods by trying to pirate OS X? I'm ok with it personally. I own 4 Mac's personally, have a G5 tower on my desk at work. My employer makes me carry this 20lb Dell around when I travel and I'm certainly not going to add weight by putting my powerbook in my luggage as well. So if I can have a few of the comforts of home-computing on the road with me, then I'll do it. It may not be completely legal, but I'm not taking any money out of anyone's pockets and I'm only using one instance of my OS X 86 license at a time.

Re:Irony

[Ohreally_factor]

I would suspect that the legal purchasers that are hindered and annoyed by this, if it turns out to be true, would be on the order of about 10. No, not 10 per cent. 10.

Comparing it to Sony's rootkit is beyond absurd.

Re:Software pirates won't care

[Anonymous Coward]

Huh?
a) There is no IDA Pro for Mac OS X.
b) GCC? Did you mean gdb?
c) How is debugging relevant to an open source kernel?
d) Don't Steal Mac OS.kext is an empty kernel extension that just contains a bit of text.

Could someone explain why this is moderated "Interesting"?

Re:Obligatory: Will someone fork the last open cod

[aristotle-dude]

Apple owns the copyright to the kernel and it is licensed under the Apple Public source License [apple.com]. Since you do not own the copyright, you do not have the right to relicense it under the GPL.

What would give you the impression that you should have the right to fork something you do not own and did not contribute anything to?

Given that nobody bothered to contribute to the XNU kernel when it was available, what makes you think someone would contribute to a fork? Technically, it might be possible to fork but in reality, you would have to publish any changes back to Apple anyway and assign them copyright under the APSL so your code branch would become a dead end quickly as they would release a kernel with any desirable changes you made as well as changes they made internally.

Welcome Back To The World of Closed Unix

[linguae]

It looks like OS X is taking a few tips from the 80s. Most Unix developers are accustomed to having access to the source code for the system; this dates back to the mid-1970s when universities bought Unix licenses from AT&T including source code to study. This practice ended in the 1980s when source code licenses from AT&T started to cost nearly a quarter of a million dollars. Then, in the 1990s and 2000s, we get BSD, Linux, OpenSolaris, and even the original Unix sources (from Caldera). Having access to the source code of the kernel is useful for understanding how the system works, creating device drivers, and optimizing the performance (research experiments, for example). Removing the kernel source code is a loss. As a FreeBSD user, closed-source Unix just doesn't make sense to me, and this removes one incentive of the Mac (although I'm still planning on getting one).

Then again, NeXTSTEP and OPENSTEP were completely closed source (but that was due to AT&T licensing; BSD wasn't fully unencumbered until about 1994), so I guess most NeXTSTEP/OPENSTEP users who switched to the Mac have no concept of having access to source code.

Re:Extremely old, and misleading, news

[Eil]

Mac OS X has ALWAYS been closed. It's Darwin that has been open. And "Darwin" is more than a bootable OS: Darwin is Apple's open source strategy AND an OS; but the usefulness has always come from the open source components of the OS, not the usefulness of Darwin as an OS itself. Darwin's usefulness as an OS is, shall we say, "limited" at best, and always has been.

RTFM. The kernel of Darwin/OS X was always open source, now it's not. The only open source parts now are the Unixy userland (and I'm not even sure about that), which is really just FreeBSD. "Limited" is an understatement. Darwin is now totally useless to the open source community.

For awhile there, it looked like Apple was going to be a friend to open source, but time and again they've shown that they're only interested in taking rather than cooperating. (The Konqueror/Safari debacle, for instance.) Rather reminds me of another large operating system company that likes open source, but only the kind of open source that they're not obligated to contribute back to.

Re:Extremely old, and misleading, news

[Quantum Fizz]

But besides that, I have to say, it's been proven with hard facts and my own experience that MacOSX is not an efficient OS. I don't know why they would even want to spend time hacking the kernel, or use MacOSX for a massive grid.

What has been your experience with Apple's XServe Clusters [apple.com]?

But regareding your hard facts and your experience, what do you know that the technicians who built the following systems don't?

• The Virginia Tech Terascale Computing Facility [vt.edu] containing 1100 Xserve G5 cluster nodes running OS X Panther
• The Human Genetics Grid Cluster [pitt.edu] with 250 PPC G5's running OS X
• The COLSA MACH 5 Cluster [technewsworld.com] using 1566 cluster nodes, which should be running OS X as well. More info here [apple.com].
• Additionally, the following people [apple.com] using OS X clusters to do biomedical research, and having won awards from Apple. But of course Apple wouldn't be unbiased in their opinions of OS X.

So can you describe your experience with Xgrid [apple.com] and why you think it's so bad. And regarding software, what problems do you see with the following software packages, or have you not used any?

• iNquiry bioinformatics [apple.com]
• grid Mathematica [apple.com]
• MATLAB's Distributed Computing Engine [apple.com]

Re:I don't care

[Anonymous Coward]

On the other hand, if you don't use it because now you can't modify the kernel to lose some of its inexactness it's completely legitimate. I'm only familiar with a slightly higher API space in OS X but it's enough to know, for instance, that OS X has a few "features" that make it awful for critical real time applications or high memory throughput. Maybe I'm wrong but I would think those flaws stem from the kernel. I'd myself thought about how if you could change the memory management and timing mechanisms, you'd have a real kicker of a scientific OS (though at the time it was partially because I found optimizing for Altivec to be extremely easy). If this change affects that space, then the potential usefulness of OS X has dropped dramatically. How much it's actual usefulness dropped I don't know - and as a percentage of users it is relatively small. Still, for me it kills the dream of OS X becoming a nirvana of scientific application development.

Re:Who cares, really?

[Anonymous Coward]

Strange to see a "Who cares" post labeled as flamebait.

While I have no strong feelings on this particular subject, I find it humorous, and perhaps slightly worrying, that you seem to think of apathy towards an issue as an inherently uncontroversial stance.

Pirates

[orasio]

Pirates attack ships at sea.

It does everybody a disservice to call copyright infringers with a term that is used for actual crimes. In fact, the DMCA does in fact make you a criminal in some copyright infringement issues, but that is just stupid. In fact, the choice of the word 'pirate' is convenient for the people who like restricted distrbution, because it implies that copyright infringers are criminals, and that kind of concept grows in people.