Slashdot Mirror
Apple Offers Solution to IT Roadmap Complaints
daria42 writes "Apple has admitted that enterprise IT users complain a lot about not being able to find out what its product roadmap is ahead of time. The Apple answer to this problem? Sign a non-disclosure agreement and go to Apple's annual worldwide developer conference, to be held in August this year in San Francisco. IT users can apparently get plans of Apple's roadmap up to 18 months ahead."
FTFA, Its software roadmap they are primarily talking about. "Those attending the conference get a clear roadmap of Apple's software development plans up to 18 months ahead"
Only one Darwin component closed: the intel port of xnu (the kernel).
It only closed because it was being primarily used to enable people to use stolen software on hardware for which it wasn't licensed.
It only takes a few idiots shitting in the pool to make the lifeguard kick everybody out.
Apple's own MacOS X Product Manager has pointed out that they've only not released the Intel xnu code yet.
This follows Apple's pattern of getting the code working, then packaging it up when they get a chance.
There has been no official closing of anything, just one overheated journalist's rumor-mongering.
For a great rebuttel to Yager's blathering I recommend reading The 'Mac OS X Closed by Pirates' Myth.
In the meantime don't repeat rumor and assumption as fact.
I don't read ACs: If a post isn't worth so much as a nom de plume to its author then I wont bother either.
To reiterate, or, indeed, repeat, something I've told you before:
The source code for Windows has never been open[1]; the argument is not the same.
The source for xnu has been open, continues to be open on PowerPC, and is available in an earlier incarnation for x86 (parity with Mac OS X 10.4.0). Therefore, saying that a final decision might not have been made on current iterations of xnu on x86 is perfectly reasonable.
Further, if anything, MORE source is now released than previously: x86 sources for all non-kernel components are released with parity with Mac OS X releases for PowerPC and x86; previously, x86 sources, including the kernel, were only available with major releases, e.g., 10.x.0.
The bottom line is, while Intel xnu is closed *right now* (and I have never disputed that fact), it's also accurate, given all of the information we know and can infer[2] right now, that the decision isn't final, and indeed may only be temporary.
Further, it's disingenuous of Yager (and anyone else) to paint this as a bigger issue, given that the majority of utility many enterprise customers have gotten from "Darwin" has been from other the many other open source projects and components that continue to be open. That is an indisputable fact, not opinion. Does this mean that no one benefited from and/or used the kernel source? No, of course not. It means exactly what I said: that the MAJORITY of the utility of Darwin has come from the other projects. Not from the kernel source, nor from the ability to build Darwin as a bootable OS. This does not diminish anyone's need or desire for the x86 kernel source; it's simply stating a fact.
For the record, I completely agree that Apple should have made some specific statement. But I think it's pretty clear from what we know that they simply haven't decided yet. While I would have loved a statement, what would they have said? "We are temporarily closing xnu on Intel, and it may or may not be permanent"? "We think we might want to close xnu, but want to test the waters first"? "We are closing xnu on x86 temporarily because of some licensing issues that need to be resolved for some components of xnu on x86"? The fact is, we really don't know why xnu source on x86 is currently unavailable, as you state.
[1] Ridiculous academic source agreements aside.
[2] Since two separate development trees are being maintained for Mac OS X 10.4.x, and since we won't have any news on Leopard (Mac OS X 10.5x) until WWDC, it might do well to give Apple the benefit of the doubt on this topic at least until WWDC. Because Apple has publicly stated that Mac OS X 10.5 will be unified across PowerPC and Intel, it would stand to reason that Apple's intentions for xnu will become clear once a unified OS (Leopard) is released.
Apple has not, and still does not, get the enterprise.
While Apple has been *extremely slowly* working to provide enterprise services (Apple's enterprise-specific phone support options are extremely improved, particularly the Preferred and Alliance levels), there are other areas that are still sorely lacking. Currently, we use AppleCare Enterprise Help Desk support, and have been fairly satisfied.
Apple does now offer 24x7 and 4 hour on-site service and support plans, and matches fairly well, most of the time, with our other vendors (primarily Dell, Sun, and IBM).
Where we get killed is on any kind of roadmap or planning information.
At Macworld San Francisco 2004, Lawrence Livermore National Laboratory, which used to be the largest Mac site in the world, talked about what they learned integrating Macs in their enterprise. I've made the presentation available here. I recommend people take a look at it, as it covers other issues as well. Like many organizations, Apple suffered during the late 90s, when it wasn't clear what Apple's direction was, and when Apple's future seemed uncertain. This meant that LLNL went from having 14000 Macs in 1997 to having about half that in 2003. In the meantime, the Windows installed base increased commensurately. (I believe that since then, we've gone to being one of the largest Mac sites outside of Apple, with well over 10000 institutionally-owned Macs, and many Apple server and storage products both deployed centrally and around the campus. Some examples.)
What is one of LLNL's top recommendations for Apple?
Develop a working balance between Apple's needed "confidentiality" and Corporate IT's need of "roadmap" information
This balance, or lack thereof, is also listed as one of their top "difficulties" when working with Apple. And I couldn't agree more.
You touched on some other issues related to software development, integration of OSS components into the OS, acknowledgment of and tracking of bugs, bug fixes being pushed out to next major releases of the operating system (e.g., 10.3.x -> 10.4), and so on. One small victory has been that Apple does now provide semi-detailed information about security updates, and does provide security updates for the previous major version of the OS.
However, the list of deficiencies is much longer. At WWDC, pretty much the only information we get is with regard to software development (and to be fair, that's all the original article actually refers to). We get virtually no information on hardware futures. We don't need to see pictures or know exactly what speed something will be. We want to know where Apple's headed. What form will the Intel servers take? We don't want to find out the DAY they ship. Will they use multiple cores? How many? Which architectures? Will they finally have redundant power supplies? How many drive bays will they have? How many expansion slots, and what kind of expansion? Will the Xserve RAID transition to SATA? Will Apple provide onboard video on the Xserve? Will there be an expansion beyond light-duty servers? How will they integrate into our existing management infrastructure? Will Mac OS X Server make provisions for virtualization of multiple instances of Mac OS X/Mac OS X Server? Where is Apple going with Darwin? What is the EOL schedule for Mac OS X/Mac OS X Server? (Apple still makes NO INFORMATION available about official end-of-life or end-of-support for any versions of Mac OS X or Mac OS X Server. We just have to guess that the previous major version of Mac OS X is what's supported.) How long will PowerPC be supported? This list goes on and on and on and on.
Yes, you can glean and infer some of this stuff unofficially from things happening in th