Longhorn in 2006, Avalon + Indigo on XP/2003

I’m currently in Bellevue, WA, attending a Microsoft event for us developer evangelists, and I’m just back into my hotel room from a meeting with Jim Allchin who gave us an update on the Longhorn roadmap – you can read the official press release here. (You can, as expected, find the first derisive comments already, for example here. 🙂

What does it mean? The bad news is that the great Longhorn vision which was presented at the PDC in 2003 has to wait a little longer to become reality in its entirety. The Longhorn shell will probably still be pixel-based, and we will still organize our information in hierarchical folders. Also, it remains to be seen what happens to the plans to merge the recently canceled ObjectSpaces into WinFS.

The good news is that

  • Longhorn will ship in 2006,
  • WinFS will probably be quite complete (span the network, e.g., not being restricted to the local machine, as was recently speculated) when it ships, hopefully with Longhorn Server in 2007,
  • Avalon + Indigo will be available on XP and Server 2003 – isn’t that what almost everyone has been asking for??
  • the wishes and demands of lots of customers have been taken into account when making these decisions.

So, I personally think that the good aspects outweigh the bad ones.

Now I’ll go and update my Longhorn slide decks… 🙂

Comments (5)

  1. Pixel based? I thought Avalon was all about vector graphics, XAML and 3D acceleration. This can hardly be called Avalon anymore if the whole vector concept has to go! :/

  2. Josh says:

    I’m not sure why it is good news that Longhorn will ship in 2006. Sure, it is good news that we will have Avalon + Indigo in 2006, on our XP/2003 machines. But why should we care about Longhorn? Microsoft will have to do a lot of re-education about what Longhorn is. Up until now, it has been explained as pillars: WinFS, Avalon, Indigo. Now we see none of them will be tied to Longhorn. So…

  3. David Morris says:

    This may sound like an idiot simple answer, but why are the development team conforming strictly to the planned release date? Wouldn’t it be better to wait a little longer for a release that was more like what was originally proposed?

  4. FrankPr says:


    There are techn(olog)ical aspects, and there are economic aspects involved in the decision when to ship Longhorn. The economic ones are clearly primary in this announcement. Remember, Windows XP, the current version, shipped in 2001, and 5 years (to 2006) are an incredibly long time in this industry already. We simply can have people wait even longer before they get the next version.


    Yes, we were a little too ambitious in planning to deeply integrate WinFX into Longhorn. This will have to wait a little longer, so the Windows team is probably focusing on incorporating new features into the existing XP codebase. That does not have to be a bad thing per se, since we’ve got a lot of improvements with SP2.

    So, why Longhorn? Because, despite the absence of large parts of WinFX, you will see a lot of improvements in security, manageability, user experience, performance, etc.


    Yes, pixel based. Avalon will of course be vector-based, but the Longhorn shell will *not* be based on Avalon in this first release, so it will be pixel-based. (My speculation.) Does this mean it won’t be cool? Not at all – look at comparable shells like Mac OS X, e.g. We’ll do a lot more in Longhorn.