Yet Another Microsoft Data Indexing Project

Plowing ahead towards Windows 7, Microsoft tossed up two more job listings explicitly looking for software engineers who want to “shape the vision of browsing, searching, indexing and visualizing data.” Given Microsoft’s track record with previous insane data-fluidity projects, I’m not too hopeful for where this might end up. Here’s a quick look back on what Microsoft unsuccessfully tried to bring to the masses:

• Cairo was a Microsoft project which revolved around Object Oriented Everything. This massive undertaking encompassed a number of ideas, such as using distributed computing for easy access to data anywhere. Most of what made up Cairo found its way into Windows later on, though there was one concept which never truly materialized on its own: the Object File Store (OFS). OFS was designed on top of an object oriented database allowing for the retrieval of information based on content.
• WinFS was an attempted revival of the Object File Store. The idea this time was to organize everything, including pictures, music, sound clips, documents, PowerPoint presentations, PDFs, etc. in such a way where a user or application could quickly and easily retrieve and work with any data on the system regardless of location. WinFS was intended to work its magic through a relational database. Because of this, a number of insane usage scenarios suddenly seemed possible. For example, inter-related data could be searched and retrieved within a single query (say, finding the pictures of anyone who called you two days ago who also happened to send a work-related email to you within the past month).
  Sadly, the project was drawn and quartered, with the remaining pieces integrated into other projects such as SQL Server 2008, ADO.NET 3.5, and Windows Desktop Search.
• Virtual Folders weren’t nearly as ambitious as the previous two, which only meant that it had much more promise of reaching fruition. It worked via xml files which wove different locations and files into one “folder.” This xml file would appear as a translucent-blue Windows folder but would function no differently from any other folder on the drive, despite the fact that the folder doesn’t exist as a folder on the file system. The virtual folder itself would be pieced together in real time when Explorer loaded the xml file, thus allowing it to change dynamically depending on the way the virtual folder was built by the user. For example. a user could have a virtual folder which would show all pictures no older than thirty days ago, and the contents of the folder would change as the days wore on and pictures got older.
  Unfortunately, the Windows Vista beta testers as well as the consumer research teams at Microsoft (remember the Maryland family from the Windows Vista Launch? People like them) decided that this was too confusing. Instead, they butchered virtual folders and carved “search folders” out of the remains. Search folders works largely on the same premise except for the fact that it’s much harder to do something useful for people with more experience than a typical salesperson.

These new job listings only have one major difference: assuming this isn’t a typo like the word “a” further down in the listings, one is searching for an engineer for the next version of Windows (238893), whereas the other is in search of someone for future versions of Windows (238899). Besides this, the two are completely identical, which leads us to the substance:

Want to shape the vision of browsing, searching, indexing and visualizing data?
The Windows group is looking for engineering staff to be a part of history by reinventing the way people think about their data. Our team is responsible for developing the Windows Explorer. In the next version[s] of Windows we will be breaking new ground enabling users to act on their items, irrespective of where they are stored and delivering a [sic] indexing technology to provide wicked fast views over their local data. [emphasis added]

First of all, the engineer hired for the task will work alongside the Windows Explorer team, which could immediately knock out the possibility of WinFS making a comeback. There is also no word of applications being able to interact with data as there was with WinFS. Since Virtual Folders were an Explorer-centric feature meant mostly for users, this could just be a revival of the old Virtual Folders. Given how popular Virtual Folders were amongst power users when Beta 1 of Windows Vista went live, I wouldn’t be surprised.

Just so long as this doesn’t turn out to be more roasted yam dip (har har) like the previous projects. everyone will be happy.

Update (9-13-08 11:04pm edt): looks like I was dead-on with the Virtual Folders; they’ll be reincarnated in Windows 7 as Libraries. In short, they’ll “let you access and arrange items from one or more folders quickly without moving them from their original locations.”

Sounds like Virtual Folders to me; I’m looking forward to it!