![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
Tech Industry Musings
Oct. 19th, 2006 11:40 amThe conventional wisdom among people who aren't Mac zealots is that the Macintosh is a niche product that isn't a real threat to Microsoft, any more than Linux is. Looking at their market share seems to bear this out (it's holding steady at about 3.8% in the U.S., or 2.2% worldwide.) The popularity of the Mac has actually declined in the last ten years.
However, I think we may be at an inflection point where Mac popularity may be about to substantially increase, for several reasons.
It used to be that Macs were great for people doing publishing tasks, like layout or graphic design, or for home users who really didn't need their computer to do much. (Putting aside the Mac OS 8 dark ages in which Macs were great only for people who liked waiting for their PC to respond.) However, those people aren't all that influential in the business world. IT people, computer geeks, avoided Macs in favor of "real computers" -- which is to say Windows PCs. Business managers, who operate under the delusion that obviously they need PCs as good or better than their IT underlings who actually use computers, insisted on the same thing for themselves.
The thing that's surprised me recently is where I've seen Macs showing up in the last year -- software developers' laptops. I just returned from the OWASP AppSec Seattle 2006 conference, and I'd say about 20% of the people there with computers were carrying a Mac. At work, I'm cooperating with a vendor of high-end UNIX-based security information management software, and they all both carry Macs and have Mac desktops back in their headquarters. We had to dig up a Mac VPN client for them, because they didn't have any Windows machines.
It's not that the technogeek segment is really that important market-share-wise -- it's that we're influential. If the sysadmin can use a Mac, then the business manager can admit that he can use one, too.
Why are software developers using Macs now? In short, it's because of two recent changes: 1.) New Macs run on Intel Core 2 CPUs, just like modern PCs, and thus offer usable performance. Contrary to Apple's marketing, the PowerMacs were actually dead slow for any serious processing. 2.) Mac OS X sits on a BSD kernel, which means that you can compile POSIX apps for it and you get a bash command line.
Having used my Linux laptop for several months, I love my bash command line. Whenever I use a Windows command line now I get annoyed at its inflexibility and hideously arcane syntax (not that bash's syntax isn't complicated -- it is -- but at least it's consistent.) When I started doing some web development recently, I started it up on my Linux laptop instead of my (faster) desktop, simply because I wanted a functional command line for doing dev work. (As an aside, you actually can do complex scripting on Windows using WSH, and you can install Cygwin and get a bash command line on Windows, too. But most people don't.)
In addition, the migration to Intel Core processors means that if you want to run a Windows app on a Mac, you can -- either directly, through an emulation layer, or in a Windows window via a virtualization layer (a simulated PC in a window), or by rebooting the machine into a real Windows installation if all else fails. The app-compat issue is basically gone for an advanced user (hence the technogeek adoption), and is at least surmountable even for normal users (or business users who can just hand the thing to their IT department and say "make it work.") App-compat is the primary source of Microsoft's vaunted "monopoly power."
Finally, Microsoft may be handing Apple a big gift in the form of Windows Vista. Vista a.) requires people to buy new computers anyway, b.) requires them to be expensive new computers (lessening the price difference between a PC and a high-markup Mac), and c.) includes various features to annoy the pants off of people, such as draconian DRM-like activation procedures and a memory footprint of over 500 megabytes. The only time people can decide between a PC (which they have) and a Mac (which they don't) is when they're buying a new PC -- Microsoft had better be damn sure they have a good value proposition in Vista (which they may; I haven't used a Vista build in over a year, so I frankly don't know how good it is or isn't) before it makes everyone choose a side again.
The real winners here? Apple, of course, but also Intel. Intel's stock has been hammered lately (down from $34 to $22 in the last year), but they have a great product right now (Core 2 Duo), and it's in both PCs and Macs. In addition, it has no real competition -- it is so much better than AMD's counterparts (Intel's $360 chip blows away AMD's $800 chip in every benchmark, yet consumes half the power and puts out much less heat) that AMD doesn't even have plans to catch up before 2Q 2007. I think it's a mistake for Intel to put the kind of emphasis on quad-core they're trying to (as quads offer almost no performance gain for real-world apps right now, and won't for a year or two), but they still have time to back off on that. Vista will cause people to buy new computers, and whether they buy PC or Mac, they buy Intel. (Actually, considering Intel's current 21 P/E, I think I'll be buying some Intel, too, and I don't mean a Core 2 Duo... though actually I'm going to buy one of those, too.)
All this said, Apple has some real problems to address for wide adoption -- specifically, they have terribly neglected the enterprise space, because they couldn't previously compete in it. If they want to take on Microsoft for real, they need administrative tools, automated deployment, group policy -- in short, they need a directory service like Microsoft's ActiveDirectory and they need the tools and server software to support it. Sure, there are some free UNIX/BSD tools, but that's not the same as having real, vendor-supported, corporate-tested applications. In addition, they need to take a hard look at security features. Apple right now is very complacent about security on account of not being much of a target. They don't have ASLR, they don't have image or heap randomization, they only marginally have stack protection -- one of these days, some group of hackers is going to do some serious fuzzing on Mac OS X's interfaces and find a security bug farm. Sure, their BSD base means most of these compromises won't own the box, so that the thieves will "only" get access to all your files and personal information. I'm sure that'll make business people feel really comfortable.
In any case, I think Apple has the best opportunity they've ever had to make a market share inroad right now. It'll be interesting to see what they do with it -- are they capable of making business machines, or will their own culture prevent it?
However, I think we may be at an inflection point where Mac popularity may be about to substantially increase, for several reasons.
It used to be that Macs were great for people doing publishing tasks, like layout or graphic design, or for home users who really didn't need their computer to do much. (Putting aside the Mac OS 8 dark ages in which Macs were great only for people who liked waiting for their PC to respond.) However, those people aren't all that influential in the business world. IT people, computer geeks, avoided Macs in favor of "real computers" -- which is to say Windows PCs. Business managers, who operate under the delusion that obviously they need PCs as good or better than their IT underlings who actually use computers, insisted on the same thing for themselves.
The thing that's surprised me recently is where I've seen Macs showing up in the last year -- software developers' laptops. I just returned from the OWASP AppSec Seattle 2006 conference, and I'd say about 20% of the people there with computers were carrying a Mac. At work, I'm cooperating with a vendor of high-end UNIX-based security information management software, and they all both carry Macs and have Mac desktops back in their headquarters. We had to dig up a Mac VPN client for them, because they didn't have any Windows machines.
It's not that the technogeek segment is really that important market-share-wise -- it's that we're influential. If the sysadmin can use a Mac, then the business manager can admit that he can use one, too.
Why are software developers using Macs now? In short, it's because of two recent changes: 1.) New Macs run on Intel Core 2 CPUs, just like modern PCs, and thus offer usable performance. Contrary to Apple's marketing, the PowerMacs were actually dead slow for any serious processing. 2.) Mac OS X sits on a BSD kernel, which means that you can compile POSIX apps for it and you get a bash command line.
Having used my Linux laptop for several months, I love my bash command line. Whenever I use a Windows command line now I get annoyed at its inflexibility and hideously arcane syntax (not that bash's syntax isn't complicated -- it is -- but at least it's consistent.) When I started doing some web development recently, I started it up on my Linux laptop instead of my (faster) desktop, simply because I wanted a functional command line for doing dev work. (As an aside, you actually can do complex scripting on Windows using WSH, and you can install Cygwin and get a bash command line on Windows, too. But most people don't.)
In addition, the migration to Intel Core processors means that if you want to run a Windows app on a Mac, you can -- either directly, through an emulation layer, or in a Windows window via a virtualization layer (a simulated PC in a window), or by rebooting the machine into a real Windows installation if all else fails. The app-compat issue is basically gone for an advanced user (hence the technogeek adoption), and is at least surmountable even for normal users (or business users who can just hand the thing to their IT department and say "make it work.") App-compat is the primary source of Microsoft's vaunted "monopoly power."
Finally, Microsoft may be handing Apple a big gift in the form of Windows Vista. Vista a.) requires people to buy new computers anyway, b.) requires them to be expensive new computers (lessening the price difference between a PC and a high-markup Mac), and c.) includes various features to annoy the pants off of people, such as draconian DRM-like activation procedures and a memory footprint of over 500 megabytes. The only time people can decide between a PC (which they have) and a Mac (which they don't) is when they're buying a new PC -- Microsoft had better be damn sure they have a good value proposition in Vista (which they may; I haven't used a Vista build in over a year, so I frankly don't know how good it is or isn't) before it makes everyone choose a side again.
The real winners here? Apple, of course, but also Intel. Intel's stock has been hammered lately (down from $34 to $22 in the last year), but they have a great product right now (Core 2 Duo), and it's in both PCs and Macs. In addition, it has no real competition -- it is so much better than AMD's counterparts (Intel's $360 chip blows away AMD's $800 chip in every benchmark, yet consumes half the power and puts out much less heat) that AMD doesn't even have plans to catch up before 2Q 2007. I think it's a mistake for Intel to put the kind of emphasis on quad-core they're trying to (as quads offer almost no performance gain for real-world apps right now, and won't for a year or two), but they still have time to back off on that. Vista will cause people to buy new computers, and whether they buy PC or Mac, they buy Intel. (Actually, considering Intel's current 21 P/E, I think I'll be buying some Intel, too, and I don't mean a Core 2 Duo... though actually I'm going to buy one of those, too.)
All this said, Apple has some real problems to address for wide adoption -- specifically, they have terribly neglected the enterprise space, because they couldn't previously compete in it. If they want to take on Microsoft for real, they need administrative tools, automated deployment, group policy -- in short, they need a directory service like Microsoft's ActiveDirectory and they need the tools and server software to support it. Sure, there are some free UNIX/BSD tools, but that's not the same as having real, vendor-supported, corporate-tested applications. In addition, they need to take a hard look at security features. Apple right now is very complacent about security on account of not being much of a target. They don't have ASLR, they don't have image or heap randomization, they only marginally have stack protection -- one of these days, some group of hackers is going to do some serious fuzzing on Mac OS X's interfaces and find a security bug farm. Sure, their BSD base means most of these compromises won't own the box, so that the thieves will "only" get access to all your files and personal information. I'm sure that'll make business people feel really comfortable.
In any case, I think Apple has the best opportunity they've ever had to make a market share inroad right now. It'll be interesting to see what they do with it -- are they capable of making business machines, or will their own culture prevent it?
