News is spreading that Longhorn is getting closer. Several online publications are indicating that Microsoft is going to be announcing details this afternoon. This is great.
It sounds like Microsoft is still targetting a 2006 ship date for the Windows Longhorn Client, but in order to make this happen, I understand that some features explored in the current Longhorn pre-alpha builds are being re-prioritized.
First and foremost, security continues to be at the top of the list. That’s what SP2 has been all about. As SP2 illustrates, detailed, security-minded work takes time–lot’s of time–especially when they are merged in with all the customer requested enhancements–so not everything that’s been talked about is going to make it into the first stage of Longhorn. No surprise here. That’s the way software gets written and done.
So what are some changes? First, it sounds like the new file system WinFS is going to slip a bit. It’ll be available in beta when the Longhorn Client ships. This makes a lot of sense because for enterprises that want to leverage WinFS they will need it to work with the server products, but the server version of Windows won’t be finished until after the Longhorn Client. Taking the right time to design WinFS for Client and Server versions makes a lot of sense.
Some may be disappointed in the delay on WinFS. I’m not. High-performance searching and retrieval of information on its own is tough. A new file system is tough too. Merge the two and you have a big challenge. The information problem that people are encountering is still being explored. It’s not just a database problem. Indexing, syncrhonization, searching are all substantial problems–in fact, I think these are the real problems. I look forward to seeing WinFS evolve, but I for one am glad I don’t have to be any more of a guinea pig as Microsoft explores how to best provide information retrieval capabilities than I have to.
One huge “win” for developers is that the WinFX graphics progamming model and Indigo web services are going to be made available not only in Longhorn, but also for Windows XP and Server 2003. I have no idea if these will be trimmed down versions of these Longhorn-designed features, but I wouldn’t be surprised. We’ll have to see.
Either way, it will provide an easier way for developers to transition into Longhorn development as Avalon and Indigo make their way to the mainstream.
As a developer I’m not surprised by the Longhorn announcement and I’m actually quite pleased. It sounds like Longhorn is getting real–a natural part of the development cycle. Who knows when the Longhorn Client will actually ship, but whatever date it is, I’m going to guess that it would have been possibly years delayed without reshuffling the roadmap.
It’s official, here’s the press release on the updated Longhorn roadmap.