Mac OS X Darwin x86 Still Open?

[MacRumors]

http://www.macrumors.com/images/macrumorsthreadlogo.gif

In response to reports last week that Apple had closed the source to the Intel version of Mac OS X kernel, an Apple employee responded saying that nothing has been announced yet, so all discussion remains speculative.

Just to be clear, Tom Yager was *speculating* about why we have -- so far -- not released the source code of the kernel for Intel-based Macintoshes. We continue to release *all* the Darwin sources for our PowerPC systems, and so far has released all the non-kernel Darwin sources for Intel.

Nothing has been announced, so he (and everyone else) certainly has the right to speculate. But please don't confuse "speculation" with "fact."

[aricher]

"But please don't confuse "speculation" with "fact."

Classic. Very interesting nevertheless.

[hcorf]

looks like we all jumped the gun on this one

[EricNau]

While he didn't confirm the rumor, he didn't deny it either.

[Analog Kid]

Interesting. Apple employee posting by name to an Apple mailing list. Wonder if he entered the fray because he knows something different than the speculation, or because he's been taking a different side of an internal argument and hasn't given up yet.

I guess this could mean "delayed release". The code that's available always a generation behind production code? (Opps! More speculation... =P)

[Thom_Edwards]

But please don't confuse "speculation" with "fact."

tom yager is obviously not a macrumors regular!

[aegisdesign]

What's up with the rumourmill these last two weeks or so?

MacRumors, AppleInsider, ThinkSecret... all reporting stuff the news sites have posted days ago.

Ars had it Monday morning http://arstechnica.com/journals/appl...2006/5/22/4055 and Daring Fireball the day before.

[apachie2k]

wow, i dont' even know what ANY of that means

[longofest]

Originally Posted by hcorf

looks like we all jumped the gun on this one

Not necessarily. His quote isn't really all that informative. He's simply saying that Apple hasn't announced anything, and so until they do, it's simply speculation.

[dr_lha]

Originally Posted by Analog Kid

Interesting. Apple employee posting by name to an Apple mailing list.

Why? The do this all the time.

[ChrisA]

Originally Posted by Analog Kid

... The code that's available always a generation behind production code? (Opps! More speculation... =P)

No need to speculate here. Open source code for 10.4.6 is on Apple's web site
http://www.opensource.apple.com/darw...ce/10.4.6.x86/

[Doctor Q]

Originally Posted by Thom_Edwards

tom yager is obviously not a macrumors regular!

Unless he's using a different name and speculating on his own non-story!

[AnimeUnrivaled]

Originally Posted by ChrisA

No need to speculate here. Open source code for 10.4.6 is on Apple's web site
http://www.opensource.apple.com/darw...ce/10.4.6.x86/

As the original quote from the Apple employee stated, All non-kernel sources for Darwin have been posted. Analog Kid's "speculation" was merely that it's possible that there will be delayed release for kernel sources.

[fproulx]

OK, let me "speculate" for a moment. Haven't we heard about Apple who might "possibly" be in relation with Sun for integrating the ZFS file system ? Could they be integrating stuff in the new kernel, but since the deal is not closed with Sun, they don't want to people to play around with the code just now and that they will re-open the source once it's cool. After all, up to now they used the "Bazaar" style open source for their kernel, they might be thinking of switching to "Cathedral" style (where the source code is only available "by milestone").....

Food for thought here.

[dr_lha]

Originally Posted by fproulx

OK, let me "speculate" for a moment. Haven't we heard about Apple who might "possibly" be in relation with Sun for integrating the ZFS file system ? Could they be integrating stuff in the new kernel, but since the deal is not closed with Sun, they don't want to people to play around with the code just now and that they will re-open the source once it's cool. After all, up to now they used the "Bazaar" style open source for their kernel, they might be thinking of switching to "Cathedral" style (where the source code is only available "by milestone").....

Food for thought here.

Doubt it. That would not be CPU dependent, so if they are working on that stuff it would be in the PPC branch of the XNU kernel as well, which is publicly available.

[ncook06]

It just says "still open," not "to remain open"

[Lollypop]

Originally Posted by dr_lha

Doubt it. That would not be CPU dependent, so if they are working on that stuff it would be in the PPC branch of the XNU kernel as well, which is publicly available.

My thoughs exactly, I dont think apple would make ZFS intel only either, the thing that I would like to know however is why the kernel isnt realeased along with the PPC version, what is in it (or not in it) that doesnt allow it to be opened.

[bigandy]

i don't make anything of this, because it just says "it's still open" and offers no insight whether or not it will stay open.

[eSnow]

We continue to release *all* the Darwin sources for our PowerPC systems, and so far has released all the non-kernel Darwin sources for Intel.

So, in a way he underscores Yagers point - all PPC code and anything x86 except the kernel. Nice move, Apple. Not.

[kingtj]

When the original story broke, www.slashdot.org had a story and discussion about it. To me, that discussion seemed FAR more informative than the talk I've read about this elsewhere, because it came from the viewpoint of Linux/Unix fans, many of who were actively involved in software development and some who wrote code that Apple's team shared while working on OS X and its apps (Safari, etc.).

It seemed to me that it boiled down to this article being, at best, "old news", and probably misleading -- making people think Apple suddenly locked down some code that used to be freely available.

The developers who discussed this whole thing said that they knew from day 1 that Apple was not going to release full source code for the x86 kernel for OS X. This wasn't some sort of mid-stream change by Apple. In fact, they say it's VERY difficult and confusing to compile up a working Darwin kernel from Apple's provided source for the PPC version - because header files and things have to be located, piecemeal, from all over the place to make it work.

Stop and think about it. Why is this a big deal to 99.999% of the Mac using community? The single biggest reason someone would be interested in source to an OS X x86 kernel is to recompile the latest versions, stripped of security measures, to make them run on non Apple hardware. Anyone legally using OS X on real Apple products won't benefit from that. The argument that "scientific applications in OS X might need custom tweaking of the kernel to make it leaner and meaner" is rather lame too. Seriously, OS X has never been a good choice for a "lean and mean" customized OS where timing is critical, in some sort of research/science setting. A better choice there would be an OS like Linux or BSD. OS X on XServe racks scales quite poorly in things like SQL Server transactions too. It's a great product for user-friendliness and ease of administration, and total integration with a Mac workstation environment. It's *not* your end-all, be-all, fastest solution if speed is more important than a nice GUI, ease-of-use, etc.

Originally Posted by eSnow

So, in a way he underscores Yagers point - all PPC code and anything x86 except the kernel. Nice move, Apple. Not.

[peharri]

There is no speculation.

The source code to recent versions of XNU for Intel is not there. That's counter to Apple's until recent policy of releasing sources for XNU in updated versions of Darwin. The answer to the headline is "No", no, it isn't fully open today in the same way as it has been in the past.

The comment doesn't add any new information. It looks to me like damage limitation. Get people speculating that pro-OSS employees inside Apple may be able to change Apple's mind, and hopefully if they do release code at some point in the future, everyone will forget the incident occurred. But it's sophistry to claim that it's unknown as to whether XNU for Intel is open today. It isn't. You can't download the source for it past Darwin 8.01, and given the fact you can for other components of the same OS for the current and previous versions, you'd expect to. At best you can argue that Darwin 8.01 is, today, open.

The same argument that "XNU/Intel for Darwin 8.1+ isn't closed today because they might release the source code one day" could be used to imply that Mac OS X is open source, or Microsoft Windows. After all, nobody's denied that the source for those operating systems might, one day, be released as well...

[rastejante]

Originally Posted by kingtj

Stop and think about it. Why is this a big deal to 99.999% of the Mac using community? The single biggest reason someone would be interested in source to an OS X x86 kernel is to recompile the latest versions, stripped of security measures, to make them run on non Apple hardware. Anyone legally using OS X on real Apple products won't benefit from that.

I don't agree with you. For example as a Pennetration Tester i need the syscal used to spoof the mac address of wireless NICS (Apple removes it since 1.4.3), if the kernel was open-source i would have made the modifications myself.
Ok... maybe i'm one of the 0.001%!
If Apple made the great steep of start using open-source, i don't understand why they go backwards now! And one of the major reasons that was pointed was the security improvement that a open-source politic will bring for the OS.
I my opinion, this is --ANOTHER-- shame on Apple.