WinFS slips into oblivion

The news broke late last week. Microsoft's Bob Muglia (the guy in charge of Windows Server) made it clear in an interview that the next-generation Windows file system, code-named WinFS, won't be in Longhorn Client next year. Well, we already knew that. But he also said it won't be in Longhorn Server in 2007. It might not even make it into the Longhorn Update, circa 2009. In fact, he doesn't even seem real sure we'll be seeing WinFS in Blackcomb, the version of Windows destined to ship in 2010 or (given the reality of how Microsoft's schedules slip) well beyond.

Let's review. At the PDC in late 2003, Microsoft showed off WinFS in public. This is to be a new metadata layer placed on top of NTFS, allowing all sorts of spiffy searching and object storage scenarios. Conference attendees got their hands on an ultra early alpha (aka "broken") build of Longhorn to try out this stuff themselves, and I know some developers who have invested substantial time in understanding WinFS in its current incarnation.

But the story doesn't start there, oh no. It was way back in 1988 that those of us who were paying attention first heard about Storage+, the new unified storage system that was going to be based on SQL Server and ship in Windows "Cairo." No, that's not a typo: this upgrade to the tried-and-true file system was first announced as part of the Windows DNA architecture, 16 years ago. Since then, under various names including Relational File System, RFS, the new storage system has been a planned part of every single version of Windows server. At least, it's been a planned part until ship dates started to creep closer and then, like the bear who went over the mountain, it saw another mountain to move to.

At this point, I just pity a few people. For starters, there are the external developers who've spent time with the WinFS code in Longhorn. Forget it; if this thing isn't going to ship for years, anything you've seen already is irrelevant. But even more, I pity those who've managed to get on to the WinFS team inside Microsoft. Somehow it's fitting to have a cursed project associated with an Egyptian code name right from the start. If I was inside the halls of Redmond, I'd fight tooth and nail to stay off that team. Seems like a bad choice from a career standpoint.

Of course, from out here we don't have any real idea why WinFS has slipped yet again. Perhaps, as some have speculated, it's just not getting its fair share of resources. Perhaps it's run afoul of the "second system" syndrome, in which everything gets so over-designed and over-engineered that it can't possibly be completed. Perhaps this all just reflects some arcane Microsoft turf war.

But you know what? It doesn't matter. Microsoft is not the only one with access to the data stored on your hard drive, and the recent success of applications like the Google desktop search and X1 make it clear that they're not the only ones who can search it either. Long before 2010, I expect some competitor will have rolled out an indexing and search application with its own algorithms for locating and cataloguing metadata that works so well that WinFS becomes simply irrelevant. This hare has taken so long to get started in the race that some upstart tortoise is going to plod right by it to the finish line.

About the Author

Mike Gunderloy has been developing software for a quarter-century now, and writing about it for nearly as long. He walked away from a .NET development career in 2006 and has been a happy Rails user ever since. Mike blogs at A Fresh Cup.

Upcoming Events