Jon Udell responded to my post from last week about WinFS and social computing.
... we need to understand that blogs, wikis, and social tagging have succeeded not because they've engineered data structures and access patterns to the nth degree, but because they provide an environment in which people, doing what they are naturally inclined to do, co-create the data structures and access patterns. Turbocharging that process of co-creation, on all platforms and for differing personalities, isn't only -- and perhaps isn't even primarily -- a database engineering challenge.
I think we agree :-) I should clarify, to be sure. When I wrote that I saw WinFS as the logical successor to the NSF file, the Groove data store, etc, I should have reinforced what I had written in the previous sentence: that it is by leveraging WinFS and XML -- together -- that I see traditional collaboration merging coherently with traditional collaboration. The analogy in terms of NSF is that WinFS replaces the services provided by the NSF file, while XML replaces the document within the NSF file.
The beauty of NSF was (and is) that it did (and does), in fact, allow people to co-create data structures and access patterns. Initially, anyone with a Notes client could create a database, or modify forms and views in an existing database. (And the schema-less nature of the Notes data store allowed that to happen without having to take into account such troubling things as consistency and integrity of data :-))
Yes... I know... I'm ignoring the fact that access controls on the database and server could, and normally did, limit who had the authority to play with the design of a Notes database -- but the regular Notes client contained all the design tools. It wasn't until IBM started aggressively bringing down the price of the Notes client that we had the concept of end-user licenses versus designer licenses, and eventually a separate Domino Designer client. It could therefore be argued that Lotus Notes, until circa 1996, pointed the way to the very co-creative process that Jon describes. I should add that It also pointed out the pitfalls: namely uncontrolled proliferation of large numbers of poorly designed, non-integrated applications in many early Notes shops.
In terms of architectural fundamentals, WinFS plus XML will have little or no intrinsic advantage over NSF plus Notes documents. WinFS's vendor lock-in will be a problem for certain people, no doubt, but that hasn't stopped the Office juggernaut, nor significantly slowed the growth of Outlook/Exchange in the past decade. It may begin to have an impact in the next few years, but being standards-based and not being saddled with 20 years worth of design decisions, however, will work powerfully in favor of WinFS/XML. A fundamental architecture that finally really competes with the underlying power of Notes and Domino, and XML staking a claim for that architecture being both standards-based and friendly to social collaboration could, and very likely would, do a lot to help Microsoft further entrench themselves on the corporate desktop.