Enterprise Mac | Tom Yager | InfoWorld | Apple responds to my "proprietary OS X kernel" column | May 19, 2006 02:12 PM

MORE ENTRIES

Apple responds to my "proprietary OS X kernel" column

May 19, 2006 | Comments: (0)

Apple responds to my "proprietary OS X kernel" column

I was serendipitously in San Jose when my column on Apple's closed-source OS X x86 kernel posted on-line. Apple had set up a day of meetings for me to discuss MacBook (nice!), Final Cut Express HD (overdue, underpriced and very powerful) and Apple's WorldWide Developer Conference (I'm jazzed). After the column ran, my agenda expanded to include a discussion with leaders of Apple's OS and open source programs. The gist of their message was that I had sensationalized and misrepresented a topic of interest only to me and an almost immeasurably small group of "geeks." I was told that the question of OS X kernel openness is still an open question, and while they concede that my column was accurate as a whole, anyone who didn't read past the second paragraph might misconstrue my meaning.

I told Apple that I'd address these objections in a blog post. So here it is.

I stand by every point I made in my column as it appeared in print in InfoWorld magazine and on-line at infoworld.com. I have nothing to do with any other sources for or interpretations of my column. Go to the link above and read the whole story. A thorough read obviates the need for clarification. However, I will address the objections and observations that Apple expressed during our meeting, because the discussion itself was enlightening. Apple's remarks are set in bold type.

My column didn't tell Apple anything it didn't already know. Where my Mac editorial is concerned, I write for my readers and for Apple's prospective commercial and professional customers, as well as Mac developers. I write as a proxy for vendors' customers, not for vendors.

Apple hasn't made public any decision to open source or keep proprietary the OS X x86 kernel. I allowed Apple plenty of time to publicize a decision. Steve Jobs announced the delivery of the first Intel-based Macs during the first week of January. My column ran in the second week of May. That interval reflects a promise that I made to Apple back in February (expressed in a 2/22 blog entry on OS X x86 syncing with open source Darwin x86) to delay my column by several weeks to give the company ample time to state its position first.

When Apple lands on an open source kernel policy, neither the decision nor the timing will have anything to do with me, my column or other external pressure. This was presumably to head off any gloating I might do over an Apple decision to open its code. Noted.

A lot of people linking to, paraphrasing or interpreting my column are getting the wrong idea because they failed to read past my column's second paragraph. I have higher expectations of my readers' attention spans than others do, and I don't break every point I make down to the molecular level to aid those who just don't understand the issue.

I misrepresented Apple's motivation for keeping the OS X x86 kernel closed (to slow piracy). In the May meeting, Apple said that this "represented the agenda of one person," who happened to be the one Apple person who would speak to me on this subject in February. This is the same individual for whom I held this story until now.

My concern over Apple's commitment to keeping the OS X x86 kernel open is held only by me and a fraction of a fraction of Apple's customers, and a fraction of InfoWorld's readers. Not so; I've gotten plenty of feedback on this subject from people who agree and disagree, but they were all concerned enough to write.

I did bring something to the table that gave pause to those present: If the OS X x86 kernel stays closed, commercial Linux will have an easy time pushing OS X out of competitive bids by asserting that Linux is fully open, while Apple quietly transitioned OS X x86's privileged code to closed source.

Commercial, professional, scientific and academic Mac users are the only ones I care about. I also don't concern myself with what's happening now; that's reflected in my column's title (Ahead of the Curve). Consumers buying today's Macs would be just as well served by their Macs if OS X were completely proprietary. Server and workstation customers are making plans and setting budgets now for the kinds of machines that Apple will start selling in the fall. They're not flipping a coin or making an Intel-based Xserve impulse buy while they're out shopping for an iPod skin. It's my job to lay out facts and opinions for people who have real money riding on Apple's solutions so they can select and spend wisely. And the fact is that Apple's OS X kernel for Intel-based Macs is proprietary. It's no surprise that Apple knew that I knew what they knew and so forth. It is a wonder that Apple figures you, unless you're in that tiny fraction of a fraction of my geekiest readers, don't care how open an open-source OS is.

All that having been said, I came away from that discussion fairly confident that Apple will open the OS X x86 kernel. The fact that Apple was concerned about having its motives misread was a fair tip-off. If you hadn't made it all the way to the bottom of this story, you'd have missed this paragraph!

Posted by Tom Yager on May 19, 2006 02:12 PM

RATE THIS ARTICLE:

COMMENTS

It may well be true that only a small minority of users care whether the kernel is open, but that minority is: developers.

Posted by: Brian Most at May 24, 2006 09:35 AM

"My concern over Apple's commitment to keeping the OS X x86 kernel open is held only by me and a fraction of a fraction of Apple's customers, and a fraction of InfoWorld's readers. Not so; I've gotten plenty of feedback on this subject from people who agree and disagree, but they were all concerned enough to write."

How much feedback did you get? 50,000 emails? 500,000?

I'll wager its fewer than 100. And if you disagree that's a fraction of a fraction of Apple's customers, and a fraction of InfoWorld's readers, I'm glad you're not my accountant.

Posted by: David Fell at May 25, 2006 08:08 AM

I like it when Apple is persnickety about the details of their products -- the user interfaces, the fit and finish of their hardware, color schemes and icons and all that stuff. But when they get snippy about whether or not OS X is open source, they lose my support. It either IS or ISN'T open source, and Apple should be clear on this matter either way. I can see dis/advantages in each approach, but you need to pick one and go with it. Beating up on journalists for pointing out the Emperor has no clothes doesn't sell any products. How hard would it be to make a decision here and be clear about it?

Posted by: John Proffitt at May 25, 2006 03:56 PM

David Fell makes a valid point from one perspective. Statistically with a levelset of 0, all of the commentary on feedback is a Numbers Game. But there are incidents where the Numbers Game does not truly equal.

We need to remember that Who the E-mail is from can lend weight to How many. If one of the respondents is the CIO for Mayo Clinic, it is more important then a programmer working for a single doctor.

Who it is matters. And companies that do not pay attention to the Single E-mail with Weight can suffer.

I do not expect Tom to divulge his respondents, because that would be default of confidentiality. I do not agree with Brian Most that the community that counts most is Developers. Developers do what they are told if they want to be paid. The day of "Development for development sake" is ended. Except for achidemic endevors, Nothing happens without Money involved.

Posted by: jjmcdonald at June 1, 2006 05:42 AM

jjmcdonald -- I never said developers mattered most. But they do matter, and Apple can make decisions supporting developers without hurting other users. You don't need to make groups of users exclusive in your thinking.

As you say, professional developers do as they're told. One of the things we're told is to find the best platform for business solutions. Most professional programmers don't work on shrink-wrapped software and can steer a business to run on whatever platform they can most economically service. An open kernel doesn't close the deal by itself, but it's a comfortable plus when trying to determine why an OS function is acting differently than documented or expected.

Posted by: Brian Most at June 9, 2006 03:37 PM

Technology White Papers

InfoWorld Technology Marketplace

» Technology White Papers Library

Technology White Papers by Topic

Technology White Papers E-mail Alert