Some days ago, I was rumbling against the confusion regarding the availability of Service Pack 3 for Windows XP. After that, the confusion continued. One of our admins mailed me an article about Microsoft’s decision to pull SP3. He wanted to tell me that this service pack might not be ready for prime time and that we’d better wait some time before we start deploying it. What I found interesting is that he didn’t notice the reason why Microsoft stopped delivering XP SP3. There is an incompatibility with Microsoft Dynamics RMS. Well, we don’t use this software and that certainly also applies to 99.99% of all Windows XP customers.

I think, my colleague was not the only one who perceived Microsoft’s back-and-forth behavior this way. It just makes an unprofessional impression if a company has to withdraw a new product because of technical issues. Some commentators wondered if Microsoft doesn’t test Windows service packs with their own products.

You might remember that when Vista SP1 was released, we had exactly the same situation. Microsoft was forced to postpone its public release because there were incompatibilities with a couple of device drivers from important hardware vendors. Like with the recent incident, only a very small minority of Microsoft customers were affected by those incompatibilities.

I have been complaining a lot before about Microsoft’s information policy regarding service pack release dates. However, I’m beginning to have doubts if those complaints were really justified. I am also seeing now a connection to the slow Vista adoption. The often unfair media coverage might be one explanation. But I think the real reason is that the whole Windows ecosystem is much more complex than it was six years ago when Windows XP was released. There are now so many dependencies between Windows and the rest of the software and hardware industry that even the release of an absolutely harmless service pack for a very mature operating system becomes a real challenge for the biggest software company in town.

There are quite a few analysts who have criticized that Windows Vista is too fat. But I don’t think that this is the real problem. These accusations were also quite popular when Windows 2000/XP came out. I still hear all those voices saying that it doesn’t make sense to unite the consumer version of Windows (95/98/ME) with the corporate edition (Windows NT) because the whole thing will just get too bloated. How can an operating system good for gaming also be an option for the corporate network? All these complaints are long forgotten and Windows XP is everybody’s darling now.

No, I don’t think that Windows is too bloated. The only thing that is too bloated is the whole Windows market. This is the real and the only reason for the slow Vista adoption. Journalists started bashing Vista only after the first adoption numbers were released. That made them feel safe and secured them lots of applause from the early adopters who had to fight with incompatibilities. The reason why those incompatibilities were so problematic is simply because it takes much more time than ever until this giant Windows ecosystem can digest a new operating system version or even a tiny service pack such as XP SP3. And this is the real problem because it makes radical changes (some call them innovations) very difficult. The most interesting question certainly is what can Microsoft do against this bloated Windows market? I must admit, I have no answer to this one.