One of the hardest decisions in business is to know when to stop. It's difficult enough killing a project before launch, when the ramifications are purely internal, but it's doubly hard once it's been released to the world. Yet it's also one of the most important decisions to make, especially when your organisation needs focus and clarity to maintain its competitive edge.
Kudos to Sun therefore, for its decision to can the Java Desktop System. From the out, it was a peculiar beast — SuSE Linux dressed up as a Java platform — and all the stranger for a company with a great deal of other Unix and thin-client expertise.
It might have found a niche. There is some justification for the basic idea of a cheap, easy to deploy, uniform network of clients and servers biased towards Java applications, but it was harder to understand why Sun would want to take on the task of supporting something that so closely paralleled its other efforts.
If you want to be a serious Linux shop, you have to seek out and employ some respectable programming firepower and be clear about the support you're offering to your customers. Without that sort of visible focus and investment in your product, you won't see returns. Open source software may encompass intellectual freedom, but it obeys the same commercial rules as anything else.
The downside in abandoning a product is the damage done to the community of customers it leaves behind. Here, open source does not obey the same rules as proprietary software — a factor that may have made Sun's decision easier to take. With the code base visible and expertise easily available on the open market, Sun's JDS users will find it possible to arrange alternative support and product development if they wish. There is also nothing to stop them arranging a staged migration of data and software to a new platform. It might not be a welcome turn of events, but it's by no means the end of the road.
The same cannot be said for Windows 2000 users, who now find themselves with a product that's going nowhere and only two options — buy a brand-new operating system from a company which will turn off the life support again in the future, or switch to one without the built-in limited lifespan.
At a time when Microsoft has done a good job of moulding people's perceptions of open source as just another alternative, Sun's decision to stop JDS is an ironic yet welcome reminder that this just isn't true. Deciding to stop remains a hard choice, even with open source, but at least it remains yours to take.