I took a walk down memory lane today, in the form of test driving the "Longhorn Reloaded" project. There are some people out there that really miss the "Pillars of Longhorn", and want to see the original concept completed. So they're risking being castrated by MSLegal to stabilize and release the Longhorn build that was put out at WinHEC 2004. It allowed me the opportunity to look at what Longhorn could have been, and what actually came to be, all on the same machine. (For the record, I think the project is a waste of time, but it was fun to poke around for 30 minutes or so.)
It reminded me of the way things were before Vista. There was a time not so long ago, where regular people didn't have a say in the way that software was developed. Microsoft (in particular) developers assumed that they knew more than you did about the way Windows should work. They went underground, building whatever they felt you wanted, never seeing the light of day. Then one day, they would reappear to show the world what they had built. And you were going to LIKE IT, DAMNIT! Cause you didn't have a choice in the matter.
Then PDC 2003 came around, and Microsoft shared more about Windows earlier than they ever had before. And people got really excited about the vision. Over the course of the ensuing three years, as software often does, the plan changed a couple times. The tech press lambasted Microsoft for not being able to accomplish their vision, and Longhorn was maligned ad nauseam. Most of this was simply disappointment that the vision was being built on a foundation of sand, and it wasn't possible to align development schedules on the massive scale Microsoft was attempting.
The problem with having something completely finished when it is announced (AKA the "Steve Jobs 'Reveal'") is that it doesn't allow people to get their feedback into the product. I can't tell you how many times during the process of testing "Whistler" (Windows XP) when my feedback was closed as "By Design", and no changes were made. If you didn't like the way things were... tough crap, deal with it.
The Vista beta process was a totally different beast. Sure, I still had my share of bugs closed as "not reproducible" in 30 seconds without any tester notes whatsoever. But there are plenty of places where my feedback made a difference (Remember the Network Center in Longhorn Beta 2? It was BAD). Sure, it took a little longer, but the result was an outstanding user experience that is centered around user-driven feedback.
So in looking at the Longhorn Reloaded project today, it made me realize that for all the eye candy that disappeared when Avalon was yanked from the Longhorn core, Vista is much better that the original Longhorn vision. I mean, can the "return" of WinFS really compensate for the lack of SideShow, Web Services for Devices, the networking improvements, DirectX 10, Windows Media Center, Aero, Flip3D, ASLR, the Games Center, the speech improvements, Digital Cable Tuners... I mean the list goes on and on. After about 15 minutes, Longhorn reloaded is worthless. An animated login screen is cool, but compared to Vista... there's no comparison.
So here is my point. Microsoft's cancellation of PDC 2007 today is just more evidence that Microsoft is using the "Steve Jobs Circle Jerk" as proof that Microsoft shouldn't talk about products until they have a concrete execution plan, and tangible results. That strategy is just wrong. Concrete usually equals unchangeable, and I don't want to have to fight tooth and nail to get
Some people argue that the strategy works for Office, and it's true that Sinofsky has stuck to a solid ship schedule. But before Office 2007, the Office platform hasn't been compelling since Office XP... as evidenced by the fact that Office 2003 adoption was relatively nowhere near as strong as it should have been. Compare that to Office 2007, the most compelling release to date whose sales have been fantastic. So the 18 month ship schedule is not all it's cracked up to be.
Microsoft answer to completely open source development was to keep the source, but open the development process up to the people. They had a bunch of kinks to work out, but it was a good answer. The new management regime at Microsoft wants to go back to the old way, because it caught them less flak in the press. But if they stay this course, the result could be far worse: people who are dissatisfied with their products, and have no recourse (because their feedback arrives too late in the dev cycle to change anything). It's an idea that is just as bad as this press release. I remember that time. It sucked. I don't want to go back there... do you?
Instead of being silent, Microsoft should be involving us from day one, showing a select group of beta testers what they want to focus on, and letting us give feedback, before they even start coding. That would allow them to ensure that they don't waste time on stuff that could get pulled later (WinFS, I'm talking to you). Screw the talking heads, and screw the tech press. 90% of Microsoft's customers don't read the rags that bashed Longhorn anyways. Keep the process open.. you'll save money, and your users will thank you.