WinFS: What’s up?
Readers who follow Microsoft news and blogs already have probably seen the buzz in the blogosphere concerning the fate of WinFS; the “file system as (or in a) relational database”. Apparently rumors of it’s death are based on the fact that members of the product team said it was dead. But not really. I mean.. they say it will live on in other forms and projects. Scary, eh?
I completely understand some of the frustration out there. We’ve been talking about, demonstrating, and giving out samples of the cool things that it will do. Originally it seemed to be the reason that the next version of the workstation – the future “Longhorn” client (now Windows Vista) even existed. Then it was pulled from that release, but Microsoft said we would release it as a separate add-in feature/service/update/six-pack download-thing. And even as early as TechEd last week we were demonstrating it and talking it up.
“What really happened to it?”
Well.. not being on those teams, I will have to trust, just like all of you, what the blog posts from the people who know it best are saying. And for that, rather than going to the press or other blog speculators, I have to trust the source: the WinFS Team Blog. And even though the writer of the original post was raked over the coals for the WAY he delivered the message originally (and rightly so, in my opinion), you should still check out this blog for the full story. (The post following the original one spells it out quite nicely.)
“But Kevin, what do you really think happened?”
Oh c’mon.. you really want me to add to the hundreds of speculatory, “reading between the lines” posts that are already out there?
Okay. It all started with a really really cool idea. Put database power behind a file system, so that you have all the flexibility of fast search, organization, reporting, backup, whatever, that you have with a database, but for the sake of your files, folders.. pretty much any information at all. “Brilliant!” Now wouldn’t it be cool if we could do this for the next release of our flagship OS? Absolutely!
So while developing on this great idea, decide to start selling it as well. Hey.. we’re proud of what we’re working on, and excited at what people will be able to do, so it’s natural that we should want to excite the world about this new functionality.
Along the way, though, many things happen. And here is where I really don’t know enough to say very much… but I imagine that things such as feature-creep were an issue. Also, there were benefits of what we expected in WinFS (such as the completely indexed and quickly search-able file system) that were now available by other means anyway. Eventually some of the functionality may not have seemed as important or pressing – or perhaps what was left really didn’t make sense as just one product, but rather as parts and new technology improvements in other products or processes.
“So, WinFS is dead?”
Well.. as one writer on an internal blogging email discussion said (and I’m paraphrasing), “No. WinFS is not really dead. She will live on in our memories and in our hearts.” But all joking aside, as a “ship vehicle”, that which we knew of the packaged “WinFS” is “passed on”. Nails are in the coffin. But the flowers growing at the grave represent the new functionalities we will see in later product releases.
And hopefully it will NOT turn up again as a flesh eating zombie-thing trying to break down your door.
What do you think? Do you have any speculations, complaints, “it’s about time”s you’d like to add?