What would have happened, on that fateful day of August 27, 2004, if Microsoft officials had said: "You know what? We messed up with Longhorn. And we're starting over."
Instead, as Microsoft historians know, Microsoft decided to cast its decision to gut the next version of Windows client as a "reset."
"We didn't do much -- just took out WinFS, the Windows File System. Oh yeah -- and back-port some of the stuff that was supposed to be exclusive to Longhorn to Windows XP. Other than that, it's full-steam ahead."
As Microsoft enthusiast Robert McLaws on Windows-Now.com notes, the Longhorn reset was really more of a do-over.
The Longhorn we first heard about as early as 2002 is not the Vista that Microsoft will launch next week on January 29. Fewer of the application-programming interfaces at its core are "managed," as opposed to "native," than Microsoft originally had hoped/expected. The integrated search is less capable and game-changing than the one Microsoft initially touted. In short, the product formerly code-named Longhorn is more evolutionary than revolutionary.
Like McLaws, I am not criticizing Microsoft for changing its course. I agree with him that the big mistake was not coming clean and admitting that Longhorn, as originally outlined, wasn't going to work. The stuff we saw at the Professional Developers Conference in 2003, which was Longhorn's first coming out party, looked snazzy. But Microsoft couldn't pull it off.
Being upfront about Longhorn -- and, as McLaws also suggests -- changing the code-name (Windows "Shorthorn," anyone?) to indicate it was not the same product could have changed the historical course and public perception of Windows Vista.
* the Vista development clock began ticking in August 2004, instead of August 2001? Microsoft could have claimed that Vista took just over two years (instead of five) to develop.
* Microsoft could have tabled WinFS sooner (and stopped spending countless cycles to get it to work well enough to make the centerpiece of Longhorn). The Softies could have sent WinFS to the SQL Server graveyard in 2004 instead of 2006.
* Microsoft could have dedicated some of its Windows development hands to Windows XP Service Pack (SP) 3 at an earlier point in time, thereby releasing the next XP service pack in 2005 or 2006, not in 2008.
Who knows ... Microsoft might even have managed to get Vista out in time for the holiday 2006 buying season if the company had just been up front in 2004 that it was going to release a relatively minor, yet more stable, Windows upgrade two years on the heels of Windows XP SP2. (As Windows chief Jim Allchin himself has said, XP SP2 really was a new version of Windows, not just a traditional service pack.)
Sure it's a lot of should-have/could-have/would haves. But definitely something worth pondering on the eve of the Vista launch.
Update: McLaws has some comebacks on my what-if Vista-history timeline.