Leopard early adopters suffer for the rest of us

I can't decide whether early adopters are saints or fools.

Mac OS X Leopard, the latest version of Apple's operating system, turns one week old today. An estimated 9 percent of the Mac OS X installed base had already signaled their intention to upgrade last weekend, and those numbers presumably grew by some degree over the last few days.

Most Leopard users seem satisfied. But there have been a fair amount of complaints from those who were first down the road to Leopard. Most are relatively minor, some were quite annoying, and a few raise questions about how Apple's operating system strategy might be different when it's time to ship the next release.

The "blue screen" problem perhaps got the most attention because it made for a delightful comparison to the infamous Windows "blue screen of death," and because it was frustrating as hell for those who had gone through the entire installation process only to derail right at the end. MacFixIt is keeping a running tally of other problems reported with incompatible applications, wireless networking and .Mac accounts, and user accounts.

There was grumbling among some of the design geeks regarding the aesthetic choices made by Apple's designers, the new folder design appearing to take the brunt of the criticism. The whole debate about transparent docks or folder icons is a little beyond me, but I do think the wavy Stacks thing is a little off-putting, and some of the icons are more confusing now than in Tiger.

And some in the developer community, most audibly Java developers, are up in arms over Apple's failure to include support for Java 6 in Leopard. They see it as another example of Apple's iron-fisted control over its developers, in that the company assumed the Father Knows Best role even though Java 6 is the latest and greatest version of that development environment.

Can you compare Leopard's first week to Tiger's, or perhaps (if you're in a sardonic mood) to Vista's? There certainly was no Leopard-related disaster, and if you measure it by sales, it was a success. As I read all of the reviews, nits, and obsequious literary odes to Leopard trotted out over the past week, I kept coming back to a few things.

First of all, the point I referenced in the opening: let's give thanks to the early adopters, however masochistic they may be. You can do all the QA in the world before releasing an operating system, and it's not going to compare to what happens when the unwashed masses get their hands on the product.

Microsoft's Windows Vista had years of developer releases, and was released to manufacturing several weeks before it went on sale to the general public. Still, compatibility problems cropped up because it's extremely difficult to anticipate what people are running, and in what combination. It's easier for Apple because it tightly controls its hardware and software, and because there are fewer potential combinations in the wild, but it's still a Herculean task.

Those people willing to be first-on-their-block with an operating system bear a disproportionate brunt of the slings and arrows to identify the problems that will be fixed in time for the rest of us. But did they go through more pain than was necessary?

Leopard had already been delayed several months to make way for the iPhone. The last-minute removal of some features promised for Leopard, such as Time Machine support for external hard drives plugged into a wireless router, could have been a signal that Apple was hard-pressed to make the late October deadline for Leopard.

That brings me to my second point: Apple's culture of secrecy can backfire. One of the reasons Apple has such a hold on the tech industry is that it keeps any thoughts it might have about the future of computing to itself. Most tech companies fall over themselves trying to trumpet their vision for what's over the next bend. Apple and CEO Steve Jobs come down from on high four or five times a year to introduce new products and hint at the future, and that's about it.

This, of course, creates the whirl of speculation and buzz that, along with the fact that they've mostly been releasing things that people want to buy, helps sell Apple gear. But it also creates distrust and paranoia, especially among those that depend on Apple's products for their livelihoods.

Many of the developer-related complaints stem from the fact that Apple held the "Gold Master" (or Golden Master, depending on your tolerance for semantics) final release of Leopard until the day it officially launched. Presumably, the company wanted to avoid a repeat of previous leaks: copies of both Leopard and Tiger appeared on file-sharing sites well ahead of their release, and the leak of a final version would be much more serious.

Apple's desire to prevent Leopard piracy shouldn't strike anyone as odd. But developers depend on that final release to make sure their applications will work properly with the final code, and if they don't have enough time to do that, problems will crop up during the upgrade process.

So, perhaps it's a question of balance: is the threat of a few leaked copies of Leopard greater than the pain suffered by early adopters slogging their way through a buggy upgrade? It's not hard to imagine Apple taking the first point more seriously, but the most likely scenario is probably that Jobs made October a "final" deadline for Leopard's release, and anything that wasn't ready by then would just have to wait until a later point release.

That would mean there was probably a scramble to set the final release, which wasn't formally released to developers until the day it was released to everyone. That also ensured that developers had no way of knowing whether the bugs they pointed out during the seeding process (where incremental builds of the operating system are released to developers under NDA) were fixed in the final copy until it hit stores last Friday. That's not the sort of thing that makes them all excited about working with Apple.

By the time the next Big Cat comes along, some may wonder whether Apple will have to change its spots when it comes to taking care of its developers. If the Mac really does start to take serious chunks of market share, will Apple have to be more friendly to developers, even if that comes at the risk of leaks?

I'm still not sure we're anywhere near that point. Apple's pitch for Macs revolves around its own lifestyle applications, like iLife and iTunes. It doesn't have the same incentive to court developers that Microsoft does (click here for Microsoft's view of developers) because it doesn't build operating systems for developers, PC companies, or IT departments. It builds them for people.

As long as Jobs is around, I can't see a new touchy-feely Apple on the horizon. And right now, there's no need for them to change: they've proven they can get to this point (mostly) on their own. It will, however, be very interesting to see if they can get to a higher place without more help from the outside, or whether they even want to take that step.

Mac OS X 10.5.1 will likely arrive around Thanksgiving with fixes for most of the problems encountered by the brave early adopters. I still don't quite understand what motivates people to go through that process, but I do wonder if more people would be willing to take the early plunge if they could be assured of fewer problems. Next time, however, Apple might not have spent the entire year launching next products, and can take more time to focus on the release itself.