Everybody buying a laptop this month is being told to wait for Vista, and I think that is bad advice for almost everyone receiving it.
The reasoning behind the waiting argument is not stupid. Microsoft has said the consumer release lands at the end of January, which is five months away, and nobody enjoys spending fifteen hundred dollars on a machine that ships with an operating system about to be superseded. If you are buying for a student who starts in September, though, five months is most of a semester, and the alternative is doing coursework on whatever is currently in the house.
The stronger reason not to wait is that the hardware just got genuinely good. Core 2 Duo arrived at the end of July and it is a much larger jump than the naming suggests. The previous generation of mobile chips were fine and ran warm and drank battery. These are meaningfully faster and meaningfully more efficient at the same time, which is the combination you almost never get in one release. A laptop bought this month has a processor that will still feel adequate in three years, and that is not something I have been able to say about a back-to-school recommendation for a while.
Vista will run on it. That is the part being lost in the argument. The machines shipping now are not going to be excluded, and the graphics requirements for the transparent window effects are the only place where a cheap laptop might fall short. Buy something with discrete graphics or at least a current integrated part and the upgrade path exists whenever you want it.
What I would actually worry about is buying at the very bottom of the market, where manufacturers are clearing stock of the previous processor generation at prices that look excellent. A five hundred dollar laptop with last year’s chip, a gigabyte of memory and a screen you have to squint at is not a bargain, it is a machine somebody will replace in eighteen months. Memory is the single cheapest upgrade that makes the largest difference and the one every budget configuration skimps on.
On the Apple side the calculation is different and simpler. The Intel transition is essentially complete, the current machines are the first Apple laptops in years that are competitive on raw performance rather than merely nice to use, and there is no equivalent operating system cliff to wait behind. The thing to check is whether the software a course requires actually runs, because a fair amount of Mac software is still being rebuilt for the new architecture and running the old versions through translation is slower than the specification implies.
The accessory advice is the boring advice and it is the advice people ignore. A second battery is worth more than any specification upgrade to somebody spending days in a building with contested power sockets. An external drive to back up onto costs less than a textbook and will eventually save a piece of work worth considerably more. Nobody buys either until after the first disaster.
Buy now, buy the current processor generation, spend the extra on memory rather than on the badge, and ignore anybody telling you to hold out until January. The operating system will still be there when you are ready for it. The semester will not wait.