Slashdot Mirror
Ballmer Calls Vista 'A Work In Progress'
shanen tips us to a Seattle Post-Intelligencer story about comments from Steve Ballmer at a conference earlier this week during which he referred to Vista as "a work in progress." He also admitted that the 5-year release cycle wasn't a good idea. Despite the approaching deadline for the end of XP sales, Ballmer's remarks about the older operating system were more ambiguous: "Vista is bigger than XP. It's going to stay bigger than XP. We have to make sure it doesn't get bigger still, and that the performance and that the battery life and that the compatibility, we're driving on the things that we need to drive hard to improve. I know we're going to continue to get feedback from people on how long XP should be available. We've got some opinions on that, we've expressed our views. ... I'm always interested in hearing from you on these and other issues."
HEAD/source code = Work in progress
Tagged release/distro = Finished release
There is a difference between always working on a project and releasing crap.
I work as a dev at Microsoft, although I've never worked in the Windows org. Still, I know plenty of development people over there and have heard plenty through the grapevine about what happened with Vista.
It wasn't a victory of marketing over engineering in so much as it was a total failure of engineering and management.
Too many interdependencies, poor project management and Windows' notorious 'cowboy culture' ultimately lead to delays and the cutting of big features like WinFS. What would become Windows Vista was supposed to originally ship in 2003. By 2006, everybody knew that Vista had to be completed so that the nightmare could end and work could begin on the next (hopefully better handled) version. There's a saying at Microsoft that "shipping is a feature". Management cut other features so that Vista could just be done with and ship after 5 years of dragging ass.
Once Vista was pushed out the door, marketing came in and did what they always do--advertise and sell the completed product. Marketing doesn't drive engineering at Microsoft like it might do at other companies. These failures were not about engineers failing to fend off demands from marketing, but rather about engineering from bottom to top scaling back an out of control project enough that it could actually be completed.