With the arrival this year of Apple's Intel-based Macs, those calculations involve a new set of questions -- the same kinds of questions enterprises face on an ongoing basis.
Typically, Mac-centric IT staffers simply focus on replacing the oldest Macs on hand -- either directly or by replacing their high-end Macs and letting those trickle down to replace slightly older models until the very oldest are phased out. With Mac OS X's highly portable nature, systems administrators need not necessarily develop new deployment disk images for those newer Macs because the old image will usually boot most Mac models (although some updating and tweaking may be needed).
However, Intel-based Macs -- which now include everything but Apple's top-end desktop systems -- pose two problems. First, they will require a separate set of disk images for deployment. Apple has yet to release a Universal version of Mac OS X (though the company has stated that this is an eventual goal, most likely to be introduced next year in Leopard - Mac OS X 10.5). As a result, Mac OS X images are no longer portable between older Power PC Macs and and the newest crop of Apple hardware.
Some users and administrators have managed to craft universal Mac OS X configurations and images. However, the potential problems of this unsupported Mac OS X configuration (which doesn't function properly with Software Update) should keep them out of a live production environment. While maintaining multiple image sets may be a lot of work, it is by far the best approach in terms of stability, security and troubleshooting any operating system or application problems -- regardless of whether the problems are a result of the home-grown configuration.
That means extra thought needs to go into any allocation of Intel Macs or any decision to postpone a Mac upgrade until 2007 in the hopes that a universal Mac OS X release will be available. Delaying upgrades presents its own problems as Power PC-based Macs become scarce. It may be difficult by then to locate PowerBook G4s, iBooks, iMac G5s and Mac minis. Older minis have pretty much disappeared from Apple's order book and even though eMacs are still available to education, they're reportedly in short supply and could soon be replaced with a similarly designed Intel eMac.
So, what should sysadmins keep in mind when allocating new Macs in an educational infrastructure? There are two major approaches.
First, consider consolidated replacement. While typical purchasing decisions (particularly for K-12 districts) might see new Macs distributed throughout various grade levels, classrooms or departments, this approach will create extra headaches for administrators because it will require two disk images that are more or less identical. Focusing hardware replacement in specific areas, where an entire high school or department gets all new computers, is much more effective because the administrator only needs to have a single image for that batch of hardware. That yields a significant time savings in a hardware rollout -- as well as a savings in the resources to maintain that image.
This also gives you the opportunity to look at departments or classrooms that might benefit from additional changes in the process. You may notice that one lab could also do with the introduction of Apple Remote Desktop for monitoring, or that a middle or that a high school art class that works with photography could benefit from iPhoto for class projects.
Another option is identifying which grades or departments get Intel Macs. Typically, new computers are give to those classrooms that will get the most use out of the increased power and features. Media labs and graphics design classrooms in colleges are often first on the list for new Macs. However, if these labs will be working with applications not yet available as Universal binaries, then installing new Intel Macs might not offer much gain in performance and could even wind up slowing things down. While a majority of non-Universal applications may be in use throughout schools, those that require more computing power, such as Adobe Photoshop or Illustrator, will be the ones hit hardest by relying on Rosetta emulation.
This is also an ideal time to look at the applications in use. Several may not be available as Universal binaries or may still require the "classic" Mac OS to run. This is particularly true of lower elementary grads that tend to stay with older educational games year after year. However, there may be new options that you can offer that are universal. This can give you a chance to work with teachers to not only develop a strategy for integrating new Macs but adding newer tools in the classroom as well.
Yet another possibility is whether to offer a combined Mac/Windows lab using Apple's Boot Camp or Parallels' virtualization software. There remain unresolved issues about using these tools: Boot camp is still in beta, and Parallels' final release only emerged recently. However, the ability to make a lab or classroom completely cross platform with a single purchase is likely to be an attractive option.
Given the number of issues involved in deploying Macs that offer both Mac OS X and Windows, you may want to wait for better tools -- as well as final releases of the software needed. It's better to make this option part of a longer term technology plan. In fact, it can mean choosing between buying a limited number of Intel Macs this year to meet basic needs) -- or a larger number next year, when dual-platform Macs and a universal Mac OS X release could offer the best of both worlds.
Ryan Faas is a freelance writer and IT consultant specializing in Mac and multiplatform network design and troubleshooting. He is the co-author of Essential Mac OS X Panther Server Administration and the author of Troubleshooting, Maintaining, and Repairing Macs. He is a regular contributor to Inform IT and is the mobile technology correspondent for Suite 101. For more information, visit RyanFaas.com.