WinFS - Death and Resurrection

WinFS


Quentin Clark, PUM for WinFS has just posted something which will probably show up on Slashdot in a few hours. I already see a storm brewing in Techmeme. You know it is important when Mini links to it :-)

Disclaimer - I've never worked on WinFS and my only involvement was lunchtime conversations about it with Pooja back when she was in the WinFS team.

I understand the need for changing the ship strategy but there's this little geek in me that wishes otherwise. After the tantalizing glimpses of WinFS at PDC 2003, my mouth watered at the prospect of a filesystem in a database.

Why? Because it throws opens possibilities that are a pain using today's filesystem metaphors. The idea of 'linking' an item to another item anywhere across the operating system was worth dreaming about.

This thinking lead me to start working on a Thinkweek paper for BillG last year. My basic premise was that there was value in investigating whether such a database model for a filesystem would work on your typical Windows Mobile device. My basic thrust was around the fact that these devices have primarily Outlook-like data all the time. However, as a developer, I jump through hoops of abstractions to get at this data (the hoops are much less in number if what you want is inside Outlook Mobile).

However, this paper never materialized unfortunately. Something called 'work', as is its wont, got in the way and I got distracted by my other Thinkweek paper (which I managed to see through to completion).

However, all is not lost. Ado.Net vNext 'Entities' slices through a lot of the abstraction problems that WinFS was supposed to solve and DLinq, in my opinion, is probably the biggest data programming innovation of the decade.

Back to WinFS.Will we see a database filesystem someday? I'm sure we will. When? I don't know. Perhaps sometime next decade - I don't know. But the prospect is too mouth-watering for any geek to resist (and the evidence for this is the number of attempts Microsoft has made at this - from Cairo to WinFS).

And Microsoft has a lot of geeks. Someone will make it happen.

It is just a question of time.

Comments:
``The idea of 'linking' an item to another item anywhere across the operating system was worth dreaming about.''

the idea of symbolically linking files across the filesystem has been part of unix for quite a long time. Why do you find it so innovative?
 
Because you are talking about *data* and not *files.

Like linking an email message to a person. Like linking the person to all the photos they were in.

For me, symbolic links only redirect from one file to another. Apple's new filesystem does do some of the stuff I speak of.
 
Post a Comment



<< Home

Archives

November 2004   January 2006   June 2006   July 2006   August 2006   September 2006   October 2006   November 2006   December 2006   January 2007   February 2007   March 2007   April 2007   May 2007   June 2007   July 2007   August 2007   September 2007   October 2007   December 2007   January 2008   February 2008   March 2008   April 2008   May 2008