Longhorn to be Released in 2006, Sans WinFS

skillio writes "Everyone's favorite OS maven, Bill Gates, announced a release date for Longhorn on Friday. He confirms what many had suspected - Microsoft will attempt to complete this release in calendar year 2006. The most notable element of this announcement was Gates' admission that WinFS, Microsoft's next-generation file system, would not be complete in time for this release - surprising, since this was the most hyped component of the next iteration of Windows."

Comments

Re:the later the better

[LO0G]

Ok, lets get this straight, once and for all - WinFS IS NOT A NEW FILESYSTEM!

It's a set of technologies that allow you to store metadata in a SQL-like database, and query for that information.

Think of it as content indexing on steroids.

So you winamp album metadata could be put in WinFS and then winamp (or WMP, or Soniq, or iTunes) could build virtual playlists from that metadata.

Or your picture keywords could be put in and you'd be able to search that metadata using a single common API.

It's NOT a new filesystem.

Re:What's so tricky about WinFS?

[Halcyon-X]

WinFS is a way for applications to share data through defacto XML schema. Like the Windows clipboard allows data to be pasted from any application to any other application (in theory), WinFS is supposed to do the same, so any application can request any data through any other application, and it will process it. Sort of like piping in Linux "everything as a file", only they will have hooks for everything not just stdin stdout. I also assume they will tie in NGSCB/Palladium authentication into this. Here is a link explaining this in more detail.

The goal is to make their hard disk search easier, handling all types of data. Another goal is to be like open source, by giving proprietary software more reason not to re-invent the wheel, because they can access the data through another application. They will use meta data to define everything so any application can use any data.

The problem is that 3rd parties all have to agree on a standard, and no doubt patents will be involved, licensing, preventing applications from working well with one another to gain an edge, viruses will have a MUCH easier time doing silly things with your data (this could make distributed data mining a reality if a worm spreads enough), who knows if it will work in practice as well as it should in theory.

This is why WinFS doesn't replace NTFS but cooperates with it, it's a layer of meta data. Needless to say MS have a huge task on their hands.

Re:WinFS Is A Prime Example Of Unneeded Bloat

[SilentChris]

Well, not exactly. I can see how I'd explain this to my grandmother ("Nana, type 'vacation photos from our trip to Italy'" instead of "Nana, search for files with the name DSCITALY001...") That's the ideal implementation any way.

I could also see this being a boon for business. Often when I'm on the phone with someone, I like to pull up all of our email coorespondance. They could do a "spokewheel" implementation: each person would be an axle and various spokes would link to business contact info, personal information, photos of them, etc. Think calling a client, having it pop up and asking "Oh, how was your son's birthday last week?" Again, ideal implementation.

Re:NTFS 5 and 'everyone else'

[lucas teh geek]

Most file utils want you to boot to DOS, Knoppix boots you to Linux, and if you're lucky, you can read, but not write.

It drives me up a freaking wall. I've forced Knoppix to mount an NTFS volume r/w, and made a change to boot.ini once, and I got off lucky.

you do realise knoppix includes a util called captive-ntfs, which allows you to mount ntfs partitions using certian windows files (which it gets from the ntfs partition) for full read/write access? I've used this quite a lot since i found out about it and never had any problems; I'd trust it a whole lot more than I trust the hack-job reverse engineered ntfs write support from the kernel.