Sunday, April 23, 2006

Here is why boot camp doesn't help Apple's marketshare.

One area I've read many an article on why Apple can't crack drastic captures in marketshare is because of the corporate environment. This is very true, and boot camp does nothing for this, if not complicating it. Granted I'm sure there are corporate environments out there that are predominantly Macintosh. I'm also sure that all 5 of them are very happy with them. The problem that you face when you are in a corporate environment is "Who's going to support this?". There are probably tens of thousands of people you can hire to administrate a microsoft lan and support windows desktops. Be they from Dell, Gateway, or any generic OEM. This is where Apple falls down, hard. There aren't going to be as many people on hand to support OSX, let alone dual booting.

Apple parades OSX around as a great Unix system, yet Apple's certifications aren't as applicable as say an RHCE or MCSE. It all falls back to OSX being an eccentric server by market standards, and by the Mach kernel's performance in database/high traffic environments stinking with multiple concurrent users. In the previous link to Anandtech, they found that OSX manages load extremely well, but pile on the threads and it just collapses. Thus this begins a vicious cycle. Other companies have undoubtedly found OSX to be an inadequate server, hence no one wants to get OSX server certification because there aren't any jobs for it further compounding the problem Apple has in getting people in the workforce with Apple certifications. The current reality is that most people getting a certification to assist them in their careers will go for an RHCE or MCSE long before getting ACTC or an ACSA. There is currently a much more friendly environment for Linux on the desktop for corporate than for Apple for this very reason. There are a lot of RHCE's out there that could most definitely support an Apple system, but they have linux training specifically and that's what corporate looks at.

What's this you say? Apple allows Windows on the machine so you can dual boot! In a corporate environment having OSX as an alternative OS to dual boot to is actually horribly detrimental. If you look at Linux corporate desktops, they are very seldom dual booting into Windows. The two reasons I can primarily think of are the productivity loss and the fact that you haven't saved an OS license this way. A person having to dual boot into another OS loses productivy (granted that virtualization helps with this, it's still a waste of system resources) and they haven't saved the company any money. If you get a Mac for a desktop user that needs Windows stuff now and then, you've for one purchased a more expensive machine than you could have purchased from Dell that would allow the same amount of work to get done, but you've also created a hassle for that user. If you run Windows in a virtualized environment, you're still wasting that user's time and potentially creating support problems. A virtualized environment is not perfect, and things can go wrong, when they do who's going to support this?

Now I am not a big fan of speculation, and there are COUNTLESS articles on what Apple "should" do. If you follow the logic it does become simple. Apple needs to have some success in getting OSX Server out there and performing really great. One avenue is to ditch the Mach kernel and go to another one. LowEndMac has a great article making the case to go to the Linux kernel. In that article Dan goes into a lot of semantics as to why Apple isn't already on Linux, that go way back to the NeXT days when they decided to go with the Mach kernel in the first place. He even makes mention of a Wikipedia entry detailing why Microkernels generally underperform by comparison. He's absolutely right on technical merit, but we run into a brick wall there. The GPL, Apple is very unlikely to use GPL kernels because then they have to document it's integration with the above functioning software layers much akin to how Darwin functions with the Mach kernel. If Apple pitched a fit about Asteroid going public, cows are quite a bit more likely to fly before that would open up. However there is the full BSD kernel, and it is no dog. In fact BSD holds its own quite well as you can see. It appears to perform great in scaling as well, which was a downfall for the Mach kernel.

Apple may already be going in this direction too, with the departure of Avie, who else will be making the pitch to Jobs to keep the Mach? One thing is sure though, performance has to increase.

Boot camp is something that can be done, so it is being done. People that will make the most use out of boot camp are very similar to the users that dual boot linux and windows. Let me just pose a question. If dual booting didn't send a visually superior, faster, leaner, more functional Linux into vast marketshare growth why is it expected to do the same for Apple?

Back in 2003 there were many pushes from Linux vendors, primarily Suse, that dual booting made the ideal workstation. Dual booting may have made more Windows users try Linux out, but they didn't have to buy a whole new machine for it. My notion is that boot camp is just going to make more Windows users out of Mac users, but at least they are doing it on a Mac, so good for Apple. The reason that it won't crack open desktop corporate environments is the support nightmare and lack of people in the workforce to support those systems. Expect it to be this way until we see real technical merit for running an OSX web/sql server.

Technorati Tags:


Post a Comment

Subscribe to Post Comments [Atom]

Links to this post:

Create a Link

<< Home