WinFS Beta 1 Released Early

Mouldy Punk writes "Infoworld is reporting that WinFS Beta 1 has been released. The new relational file system for Windows is posted on MSDN Subscriber Downloads. This release is designed to offer developers a preview of WinFS capabilities. WinFS will be in beta when Windows Vista ships and will RTM afterwords. WinFS, when it ships, will be available for download for Windows Vista and possible support for Windows XP is being considered. The distribution mechanism for WinFS will be through an add-on download much like the .NET framework is today. Tom Rizzo also notes that there is a new blog dedicated to Win FS."

Comments

I wanna know

[Anonymous Coward]

What should Hans has to say on this fs.

First post.

[Anonymous Coward]

MS just found the backup disk.

Is this really a file system?

[jbplou]

A file system that you get by an add-on? What good will that do, most desktops in Windows have partion set to ntfs under XP what do you do with it once you added it on. Is this really a file system or is it a indexer of files.

Re:Is this really a file system?

[Anonymous Coward]

You can convert ex2fs to ext3fs, and you can convert fat32 to ntfs. There is a good chance you will be able to upgrade ntfs to winfs.

Re:Is this really a file system?

[Tackhead]

> A file system that you get by an add-on? What good will that do, most desktops in Windows have partion set to ntfs under XP what do you do with it once you added it on. Is this really a file system or is it a indexer of files.

The bu^H^Hfeature is that you no longer get^H^H^Hneed to know where your files are.

Some idiot UI designer probably wrote a paper about how Windows users are confused as to where their files are located.

Rather than addressing the root of the problem -- the even bigger idiot UI designer for Windows 95 who decided to (a) by default, hide the full path to the file and (b) again by default, also hide the file extension, and (c) when users turn off "hide file extensions", still hide some file extensions like .SHS, etc -- and whose mistake was propagated to Windows 98, 98SE, ME, NT, 2K, XP, and 2K3, effectively making it impossible for nontechnical users to ever learn where their files were located...

Ahem. Rather than addressing the real problem of why nontechnical users had trouble finding where their files were, the idiot UI designer for WinFS decided to take idiocy to its most proper level: at no time should a user ever be able to find a file. At no time should a user ever be able to choose a file's location. Teh desktop is like teh Intarweb, the user should have to goo^H^H^Huse some sort of MSN Desktop Search tool in order to find "content".

Microsoft UI: Dumber than advertised, and making sure our users stay that way.

Re:Is this really a file system?

[xygorn]

Take a look at http://channel9.msdn.com/showpost.aspx?postid=1063 56
for more information.

Basically, it sounds like the files are stored at the low level as ntfs files, with a relational database wrapping around them, allowing you to treat them as .NET objects.

Re:Is this really a file system?

[Matt2k]

> and whose mistake was propagated to Windows 98, 98SE, ME, NT, 2K, XP, and 2K3, effectively making it impossible for nontechnical users to ever learn where their files were located...

The root of the problem is that most people do not care where their files are located. They just want it to work.

By the way, I think something is wrong with your keyboard.

Re:Is this really a file system?

[Anonymous Coward]

By the way, I think something is wrong with your keyboard.

You mean the ^H's? He's probably using Lunix. It's 2005 and they still can't get the freaking backspace key to work.

Re:Is this really a file system?

[bhtooefr]

As I understand it, WinFS is an overlay on top of NTFS, adding metadata, much like how VFAT is an overlay on FAT, adding long filename support.

Trivia bit: Before NT4, you couldn't install NT on an NTFS partition. FAT was the only way to go. The install WOULD immediately convert the partition to NTFS on first boot, but it wouldn't actually install as NTFS.

Re:Is this really a file system?

[globalar]

WinFS is essentially an intelligent metadata layer. In Windows OS parlance, an executive subsystem that utilizes an existing NTFS volume. The idea is to extend the traditional data model for files/folders and scraps of metadata into object-oriented patterns that the entire system can use (and hopefully reuse). Sort of like an object manager for the filesystem.

It's more than a file indexer for a developer, but just that for the enduser. Right now, it seems Microsoft really just wants feedback on the API's. If any real innovation for endusers is going to come from this, Microsoft seems to hope developers will figure it out.

ext3 was essentially an add-on for ext2. Point being, some of the better improvements don't take reinventing everything.

Re:Is this really a file system?

[aussie_a]

The bu^H^Hfeature is that you no longer get^H^H^Hneed to know where your files are.

One of these days they're going to invent an operating system that recognises the Backspace button. The possibilities will be endless! They might even have cars that run on electricity in that future age.

Re:Is this really a file system?

[ciroknight]

To be honest, I rather like the model of files having no direct location, rather, just a byte-heap in a database. For a long time, this is actually how I've organized my files on my disk, but the problem is, every now and again, your mind changes how you want to lay out all of the files, and it takes a few hours to refile everything in the correct folders.

With folders going the way of the highway, you can just heap whatever files you want, wherever you want, without all of that path confusion. Deal with namespace collisions either with longer, more descriptive file names, unique file identifiers, or a mixture of the two.

You might find it idiotic, but I find it as the best way to organize my files and find what I want, as fast as possible. Pair it with a program that can rip my files apart for all of the metadata that it can give up, index that along side the files, and no file is ever more than a few mouseclicks away. Best yet, instead of having to delete and move files around, which thrashes the disk and makes the filesystem a disaster, the filesystem can effeciently use space because it can know exactly how big the files are, and start sticking files right up next to each other. And if I were designing the UI for this thing, you'd be able to change over to a pane, change the SQL query, and poof, the folder displays what you want.

No more rediculous symlinks. No more folder paths, executable paths, etc. Better isolation of executable files and libraries and configurations, verses userspace files. Honestly, the advantages outweigh the disadvantages IMO.

Re:Is this really a file system?

[DigiShaman]

I totally agree. I mean... when I search for "Porn" on my drive, it would be as though I did a search for *.*

Re:Is this really a file system?

[Anonymous Coward]

> You might find it idiotic, but I find it as the best way to organize my files and find what I want, as fast as possible. Pair it with a program that can rip my files apart for all of the metadata that it can give up, index that along side the files, and no file is ever more than a few mouseclicks away.

That's a reasonable solution for a technical user.

...to consistently and correctly enter and update the metadata about their files?

It's hard enough to get people to ID3V1 or V2-tag their MP3 files.

But is even the most diehard of us going to bother to add the appropriate "pr0n_actiontype=[ clothed | b00bies | fullfrontal | hardcore | facial | [single|double|triple|OMGshesahunkaswisscheese]pen etration] ]" tag to each of the images in the 100-image series we just finished downloading and group-tagging with "pr0n_model1_haircolor=blonde" pr0n_model1_b00bsize=38DD"?

My goofy example illustrates my main point: is it reasonable to expect 100,000,000 non-technical users, the overwhelming majority of whom currently have trouble understanding metadata fields like file length, format, and name... in other words, people who store stuff like this:

C:\Documents and Settings\Foo\My Documents\New Report.doc
C:\Documents and Settings\Foo\My Documents\Report.doc
C:\Documents and Settings\Foo\My Documents\Report2 .doc
C:\Documents and Settings\Foo\Application Data\Adobe Acrobat 6.0\New Report I Said Make it a PDF.pdf
C:\Documents and Settings\Foo\Desktop\New Report.lnk
C:\Documents and Settings\Foo\Desktop\New Report in PDF format.lnk
C:\Documents and Settings\Foo\Desktop\New Report I Said Make it a PDF.lnk

...to consistently and correctly enter (and update!) metadata about their files?

If Grand Moff Tarkin had "I think you overestimate their chances" in response, the Death Star would still be with us.

Re:Is this really a file system?

[Skreems]

That post just made my day :-) Hilarious. 90% of users don't understand that the Desktop is actually a file in a heirarchical structure. If you stopped hiding that fact from them, sat them down and said "look, here's how things are structured"... you could fix all of this in about 10 minutes. Perpetuating the "hide things from the stupid user" UI philosophy only makes people less willing to learn, and thus increases the need for stupid workarounds to fix it. Vicious cycle.

Re:Is this really a file system?

[tsa]

The root of the problem is that most people do not care where their files are located. They just want it to work.

That attitude (of the most people you are talking about) to me is just like, for instance: ``I don't want to learn about strings and notes, I just want to play the guitar!''

People will have to get it into their heads that computers are complicated things and you need some basic understanding of how they work before being able to use them. Have you ever seen a `My Documents' folder of someone who doesn't want to know about computers? No wonder they're always complaining that `it doesn't work.'

Re:Is this really a file system?

[GuyWithLag]

The thing with the underscores is that in a CLI you don't need to quote them, whereas a space needs at least an extra character to quote it.

Some of us have grown up, and still prefer the CLI. What can you, as a person, improve easier? CLI typing speed or GUI mouse accurracy?

The GUI's strangth is providing you with information. The CLI's strength is in receiving commands.

Now if I could only merge gnome-terminal with nautilus...

Re:Is this really a file system?

[Vanders]

..to a point where interaction with the technology is as simple as pushing against a door to open it.

General purpose computing devices will never, ever be as simple as you wish. Notice that all the devices you list are single-function devices; they have a finite set of states and are capable of doing only one job.

A computer is nothing like a toaster. It is a general purpose device with an almost inifinite number of states. It's "jobs" are mearly abstractions. It can do one, many, or no jobs at any one given moment. It can not and will never have a simple interface, because it's job is not simple.

Now creating simple specific interfaces to common well understood jobs may be possible. That's still a very small subset of what a computer does though.

Re:Is this really a file system?

[xtracto]

Have you ever seen a `My Documents' folder of someone who doesn't want to know about computers? No wonder they're always complaining that `it doesn't work.'

Yes, I have seen the "My Documents" folder of my mother's account. And as you say she has like 500 documents, including MS Explorer saved files AND their corresponding folders to hold images and misc binary files.

Yes I know that for me it is really stupid, as I tend to order every thing on its subfolder. For example let me tell you how I order my music: /mnt/Music/ /Anime /Metal /Dream Theater /Images and Words /... /OST /Boondock Saints /Kill Bill /... /Guitar /Classical /Shred /... ...

blah blah, you get the idea.

And, althoug I have heard the marvelous things that programs as iTunes, Win.Media Player, Winamp Media Library or even MusicMatch jukebox do to order music libraries I still cant get one that I find really useful.

Maybe for a lot of us that is THE way to do it, but see, my mother, as a lot of computer users is just a Biology teacher. She knows the minimum required to do what she NEEDS to do in her computer (Word, Excel, Power Point) you just need to understand that people does not have the model in their heads, I mean, the model of the file system, that you/we automatically recall when we open the Windows Explorer/Knoqueror/etc...

That attitude (of the most people you are talking about) to me is just like, for instance: ``I don't want to learn about strings and notes, I just want to play the guitar!''

Now, as an example, Think about the WinFS like Gmail, I really found the Gmail approach useful, more if I have thousands of mail. If you see, desktop search bars have gained a lot of acceptance these days.

That is because we no longer know what each file in our computer does, and we do not have to care. We need to get exactly the file that we need when we need it, and you can do that searching.

Now before ranting about the facts I gave, just take my last paragraph and replace the word file with mail and instead of a Microsoft technology you will have a Google technology, is it bad? no, I really dont care where all my files go, if I need to have some files classified then a Tag would be great. otherwise I just want the OS to identify it when I ask for it.

NTFS?

[Snoolas]

I wonder if there is a possibility of MS releasing the NTFS specs for the FOSS community once WinFS becomes widely used? That would be great, but seems unlikely.

Re:NTFS?

[sharkey]

So which one is not a file system, the New Technology File System or the Windows File System?

Re:NTFS?

[sharkey]

Really? Let's ask Tom Rizzo, shall we?

GNOME Storage?

[Anonymous Coward]

I realize that this is a story about WinFS, but I'm hoping someone knowledgeable about GNOME Storage is reading.

I'm just wondering if any progress has been made on GNOME Storage or if it's just completely stagnated (a Seth project stagnating? Why I never!). My guess is all he did was some special natural language interface (which should have been an add-on later) and did no real work on a relational file system.

I wish that guy would finish something for once.

Re:GNOME Storage?

[ciroknight]

Beagle is a carbon-copy of Apple's Spotlight, which I noted in my original post. As far as I can tell, inotify was added to the kernel for the explicit purpose of allowing something like this to be created.

That being said, I cannot solicit Beagle, as much as it is a part of GNOME. First of all, it's written in C#, which I am against, but even averting that point, Beagle is slow, it's very, very buggy, and for some insane reason, they decided to go with Lucene as an Index server, instead of a fully qualified SQL server which could be connected through ODBC or any other database abstraction method.

I've said these things before and been modified as troll, with people responding with "if you could do it better, do it yourself". Well, this isn't my capacity at this point in time; I'm simply observing and reporting on the product. I understand that it's deep in alpha right now, and I do have hope that it'll get better, but in the meantime, it's connection to C#, Lucene, and fundamental archetecture problems as to where the program is allowed to index makes me doubt it's future relevance.

My point is that we need a database file system, but that Linux as a whole will be in last place to get one. Beagle is a good attempt, but I can't see it as anything more than a graduate project. I offered to port it to C++, a database agnostic implementation, and to add Kerberos/PAM support to it as my Google Summer-of-Code entry, but as I was declined, and because I do need to stay alive and eat, I can't just code it for free.

Not to sound too offtopic, but...

[Stevyn]

I hope people find it usefull. I tried the Vista beta a month or so ago and I wasn't impressed one bit. Nothing felt different or improved. I don't know if I was expecting some radical changes, but other than the "theme", it looked the same as XP. In fact, judging from "look and feel" it rendered the clear type fonts very blurry compared to xorg on gentoo which I'm currently typing this on.

However, the only thing I can saw I was pleased about was its performance. On a 2.4 ghz celeron with 512 mb of ram, it ran fine, just as fast as XP on the same system.

What did impress me about a week later was when I took that spare HD I used for vista and loaded OSX on it. Now that looked beautiful, ran fast, ran native OSX apps fine, and my conclusion from that week of OS experimentation was that if OSX ever made it to whitebox computers legally (let's not start this discussion again) it would knock Microsoft out of the water.

Let's face it, few home users will switch to Vista legally. Most will get it with a new computer. My school uses Windows 2000 and probably won't switch to even XP for a while. So go figure.

Re:Not to sound too offtopic, but...

[merreborn]

There's one good reason to switch to vista: Microsoft ends support of their OSes after 5 years. Windows 2k, as much as I love it, isn't going to be much fun after a few years without a single patch. XP will go the same way before long.

Re:Not to sound too offtopic, but...

[DrCode]

About performance...

It's somewhat telling that you were pleased that it ran just as fast as XP on the same system. On my Linux box, when I upgrade the kernel or even KDE, I generally expect better performance than before. I get the impression that OSX users expect the same.

diff -u WinXP Vista

[Goalie_Ca]

Seriously, if this is being backported to XP then what will be the difference between XP and Vista? Afaik all the avalon and .net libraries are being backported. All i can think of is a glass looking interface, some toolbars and a bunch of wizards?

WinFS Is *Not* A Filesystem

[WombatControl]

WinFS is not a separate filesystem. It uses NTFS as the filesystem, but then stores metadata on top of that (the same way other filesystems like HFS+ have for years).

You don't need to reform to WinFS, it's not a filesystem, but a relational database that carries metadata about existing files on an NTFS partition.

Umm

[ad0gg]

WinFS has nothing to do with DRM, its just a relational database storage system that indexes by xml meta data. I think Be had something similiar.

WinFS

Vista==XP

[digitalderbs]

When buying a Vista license, you'll be paying for XP a second time ... but you're really saving in the TCO.

Excellent!

[sigmaseven]

Yeah, I should use a Microsoft beta file system, because my files and documents aren't in enough danger as it is....

Seriously, is there an upside to this system to the casual-to-serious user? Or is it mostly a DRM-delivery platform? I read TFAs, but this sentence hurt me: "(Integrated data initiative is a term used to refer to a group of technologies whose goal is to provide better integration for data..." ...aaaand I just bluescreened my brain.

Don't forget DRM.

[plasmacutter]

yep.. it's chock fool of DRM. Requirements for vista logo testing involves mandatory compliance with CGMS-a, AACS, Down-rezzing, bus encryption, and "remote controlled component revocation" programs.

There's a reason Vista took so long to develop and it wasn't the end user interface

Re:What exactly is it?

[xygorn]

http://channel9.msdn.com/showpost.aspx?postid=1063 56 This interview talks about the difference between tagging and the WinFS system. Seems to boil down to a more structured relationship between tags, and the ability for multiple apps to use the same tags and tag relationships.

I wonder if it runs in Wine?

[Anonymous Coward]

If this is an add-on which interfaces to the kernel through an API it should be possible to get it to run under Wine, right? I'm fairly ignorant about how Wine works, so I'm wondering. That would be cool to have WinFS running on Linux.

Re:Too complicated.......

[wbren]

Why even released it at alL? 99% of those idiot windows users don't even know what the heck a partition is..... How can you expect them to reformat the drive to WinFS and re-install Windows on top of it all??? Unless you can convert NTFS into WinFS, theres no hope for most users, but of course... theres always geeksquad.

So much to cover, so little time. Windows bashing. Windows users bashing. GeekSquad bashing. Oh my! If you take what you're saying and shift it back to 1999's context, it's like someone saying, "Windows 98 users are too dumb to use NTFS! None of them will be able to reformat their drive to NTFS and install Windows XP! There's no hope! All is lost! Run for the hills!" and so on. (And that's all before XP was even released)

Well, of course, things went pretty smoothly. Users were able to easily convert their partitions to NTFS when upgrading (even if they didn't know what a partition was). New PCs came with NTFS by default, and Windows XP+NTFS succeeded largely (unless you're a Linux fanboy and don't want to admit it; in that case it never happened, how could it?). The (Windows) world was a better place now that FAT32 was largely a thing of the past. I'm not so sure if WinFS will be all that great, but we'll see.

Windows Vista will be no different than the 98 to XP conversion. NTFS users will be able to easily convert their partitions. Again, they will be able to do it even if they don't know what it is exactly. As long as they know it's recommended, they will keep clicking the Next button. You're worrying about something that will clearly never happen, given Microsoft's track record.

The add-on will likely be via Windows Update and extremely simple to apply. People who buy PCs after the add-on is released won't even have to do that. They will just have WinFS.

I also want to touch upon the phrase "idiot windows users" that you used. Saying something like that only serves to make you sound like an idiot. Windows users are largely novices, but you can't expect everyone to be an expert user able to keep up with the quirks of Linux et al. Calling Windows users idiots is like calling people who drive car's with automatic transmissions idiots. Sure, automatics are easier to learn to use, but that doesn't make those drivers idiots.

Now, I could go on to write a whole article bashing Geek Squad, but that would be pointless since we all know they suck and they overcharge.

but what about the interface to it?

[Gothmolly]

Right now you can add attributes to NTFS files, but there's no decent way to do it. Likewise, ANY DB-style FS is going to be limited to the ways that the vendor (MS) provides for you to access the data. Remember those ridiculous dialogs Winword used to prompt with? Asking all that crap about the author, and topic, etc. etc. until you asked Clippy how to turn the fscking thing off?
The "DB based FS" is only as good as the data that you put in, unless you solely want to make virtual folders of "all my MP3s that I warez'd last week from Rancid", but I'd say those sorts of things are going to be in the minority.. and again, depend on the metadata of said pirate MP3s.
Now there will be code jocks out there who would LOVE this sort of thing, since you could probably use it as a halfway decent free CVS replacement, but I'm thinking more of Joe and Jane Sixpack. How is it going to make their AOL experience better?

Re:but what about the interface to it?

[Keeper]

I'm actually looking forward to using it in a manner that will help me organize my photos. Directory structures + filenames don't cut it, especially when you've got attributes like "christmas vacation, florida, camille, beach, lighthouse, 16:9 crop" that you want to associate with one file.

Of course, as you note, the system is useless if you have poor metadata associated with the files. But with good metadata, the flexability/power available to organize and find the information you are looking for is increased by an order of magnitude via dynamic folder creation.

Re:Give it a rest, OK?

[DrCode]

You are (deliberately?) misunderstanding what WinFS is designed to accomplish.

Locking out Linux/Samba clients and servers?

Re:bleh

[aussie_a]

2) Release an add-on with "BETA" in the title

Works for Google.

Ever been to Cairo?

[AnimeFreak]

When Microsoft first introduced WinFS in 2003, the company said it would include a new synchronization engine that could index a host of disparate Windows files

In 2003? Jesus Christ!

I seem to remember that in 1994, Cairo was all the rage. Hell, it has been an idea since 1991. If I did not toss them out before I moved into my current house, I'd have scans of each individual article in Windows Magazine about Cairo from 1994, 1995, and 1997.

WinFS is not even close to being called "new."

Re:A Complete And Utter Joke

[Anonymous Coward]

I worked at MS on filesystems for a dozen years, and I totally agree with you. WinFS has been kicked around (and cancelled) since OS/2 days at Microsoft. Its a conflation of abstractions that needlessly complicate the data-structure of the file system. System recovery will probably be completely impossible in WinFS. And it can only slow down the system. And what does it buy the user? Faster file search. Except it will be beyond most users comprehension; a recipe for disaster rivalled only by the registry.

Finally!

[Perryman]

...A microsoft supported file sharing program! wait... what does the FS stand for again?

Re:And ...

[GeorgeMcBay]

Someone will probably have at least partial support for it before it ever even gets out of beta.


And then it will stay only partially supported for 10+ years, just like the NTFS support.

Why is this modded funny?

[HishamMuhammad]

One may or may not agree with the guys opinions (especially about his stance on non-technical issues), but the fact is that Hans Reiser is one of the top experts in the field of filesystems.

I for one would like to know what Hans has to say on this fs.

Sounds like an AS/400 to me

[msobkow]

Your description sounds an awful lot like what the AS400 team used to describe when I worked at companies that had good AS400 techies. It hybridized the mainframe-style contiguous file allocations with an integrated RDBMS that tracked the file information, much as the file information pages do with other file systems.

I find it interesting that so many "advances" other systems are making nowadays sound exactly like what the AS400 developers used to talk about. Using databases to store configuration information. Making the database an integral part of the OS. Virtualizing all storage so the system could shuffle files based on size changes and usage patterns to minimize head thrashing. Using wizards/forms for adding new software, changing configurations, etc.

I guess it's all considered "new" because so few people ever actually learned anything about the AS400 internals -- they just used them and counted on the system to do it's job properly.

Re:Sounds like an AS/400 to me

[TheAncientHacker]

It wasn't an "advance" new with the IBM AS/400. The AS/400 series inherited it from the relatively unpopular IBM System/38. The System/38 inherited it from the IBM Future Systems project done in the late '60s and early '70s but that IBM never quite managed to get quite ready enough to actually ship.

You can read more about it at the relevant Wikipedia article.

So then what is Delete

[SuperKendall]

Ok, fine... you have just heaps of data, with a myriad of references to them.

What then is delete? How does a user distinguish between "remove an association from the blob of data" vs "remove this blob of data altogether". Should the blob automatically delete when you remove all metadata around it? If not, how will you find it again? If so, would you really want data vanishing just because you removed a keyword?

What does partial backup look like on a system? How can you have a combination of partial backups and know you have a whole? I can do that with a set of five directories. Let's say you tag a set of files with "project fred". But one small file, that you almost never care about, gets tagged with "project ferd". What good is the ol' Fred backup now?

At some core level these blobs of data that users place on a system need ONE meaningful location where they always "are". You need someplace where the file will always be, no matter what other associations you remove. You need somewhere you know it will be to assure yourself EVERYTHING you care about is backed up or moved between systems.

The perfection you seek can just as easily be obtained with files in directories that allow metadata on top of them and things like smart folders that are essentially queries over the user-defined and automatically extracted metadata. In fact I think that's what WinFS does anyway (just like OS X does today).

If you really like the system you describe nothing is stopping you from storing all your files in a DB and writing an explorer on top of that. Yet all this time, things like that have never taken off in the market.

Some things do not take off because the technology to make the useful has not yet arrived. But some things simply never take off because in practice they are not practical, and the filesystem as a full-fledged database with no default structure is one of those things.

Comment removed

[account_deleted]

Comment removed based on user account deletion

Is Linux Trailing?

[hansreiser]

Reiser4 is technologically ahead of WinFS as a high performance storage layer, see www.namesys.com for details on its design. When you do this layering the way they did it, with the metadata stored in a layer above the FS rather than integrated into it, you lose a lot of performance while gain the advantage of successfully avoiding dealing with a host of technical issues. We are at least 5 years ahead of them technically in the storage layer.

That said, semantic enhancements matter more than performance, and it is better to do something semantically than to do nothing, and what Linux currently is doing is nothing.

The political support for adding semantic enhancements to Linux namespaces is mixed at best. I worry we will see that death by committee rules, and there will be no belief that each FS should try to innovate in its own way and compete with the others until one is proven the right solution. We are in serious danger of having MS implement bad technology, and Linux having to devote large amounts of resources to copying it in 5 years because we were late and chose to trail rather than lead. If the filesystems were free to compete in semantics, we could have one or several of the Linux filesystems leading them instead.

SQL and the relational model is fundamentally the wrong model for semi-structured data. See www.namesys.com/whitepaper.html for why.

Technically, I would worry much more about Apple. Dominic Giampaolo is very bright, and well funded. His chances of delivering on a good set of semantics are high because he and Jobs are very sharp, and neither of them is afraid to go where no one has gone before. Our chances of losing technically to Giampaolo and Jobs are high, because we are frankly not well funded, and a lot of us are complacent with semantics that are still pretty much the same as their father's Unix box.

So, in summary, I would say that we are still ahead but losing speed fast.

Thanks for your kind words Hisham.

Re:Is Linux Trailing?

[team99parody]

adding semantic enhancements to Linux namespaces is mixed at best......If the filesystems were free to compete in semantics, we could have one or several of the Linux filesystems leading them instead.

I could not agree more. I would very much like to see more advances/innovation/inventions out the F/OSS, and here's a place where it has happened but apparently is at the risk of stagnating.

Is there a recommended place (hopefully one of the big distros) where we can get a kernel that supports the hooks you need?

Personally, I'd speculate that these benefits would be a nice point of differentiation for one of the commercial distros; and its proven success in that environment could be a big motivation for the kernel to approve the changes.

Re:Is Linux Trailing?

[IamTheRealMike]

Hi Hans,

I've been watching the fun you've had on lkml and wanted to say don't give up! The work you and your team are doing is wondeful.

If anything, I think you should stop focussing on getting Reiser4 into the kernel and instead start demonstrating the applications of your ideas on semantics. In other words - put what you've built to work outside the kernel and prove to people that they cannot live without a next-generation filing system. It may even mean doing things you have never done before, like creating a new distro derivative.

I know how emotionally draining free software politics can be, we get a lot of that in my own autopackage project. If it gets too much rather than risk burn out, go off and do your own thing for a while. If you really do have a better way people will join your banner ;)

Re:Is Linux Trailing?

[mkro]

So true. I've also been reading the mailing list about this, and I can see Hans banging his head to the wall. I must admit the reasoning against implementing it -- why this-and-this should go in that layer and so on -- goes over my head, but I fear Hans will give up the same way Nemosoft Unv did regarding the PWC driver. Where I feel Nemosoft was wrong (and that solved itself quite nicely), I do not think we can afford losing Hans Reiser. Mike (The real one) has a good point about letting Reiser4 prove itself: Quite a few run non-standard kernels, all you need is something like BEST/KAT (Tenor?) to be successful, and the users will start applying pressure to get Reiser4 included by default.

Re:Is Linux Trailing?

[jrumney]

If by VFS you mean the gnome and kde libraries that add a whole load of explorer-like features to the desktop environments I disagree strongly. Anything implemented there is a workaround, and is only usable by applications that are built to take advantages of those libraries. File systems and the features they offer should be transparent. It shouldn't matter whether I'm using Gnome, KDE, GNUstep or the command line. The file system should be the same.

Re:Where's the Answer?

[petrus4]

At the application level, this, and this are two possible answers, or at least workarounds. On the filesystem level, this could be a possible workaround as well.

I agree however that it would seem people have been caught with their pants down in regards to WinFS though. The usual sentiment about it among Linux peeps from what I've seen is that it either isn't doable, or that it is, but that it'd be horribly slow.

Methinks a change in attitude is called for, however. This could very well be Bill's answer to the One Ring if he gets it out, which is presumably why Microsoft are trying to get a working release ASAP. Forget the coder bias for a minute here, and think about what the implications of this could be from the perspective of ease-of-use...and then think about what a battle we'd have converting people to Linux if we still don't have it when Microsoft does.

Longhorn was intended to be a Linux killer...but of all the elements I've seen, WinFS is the only one which could truly cause us problems...Especially when you consider how difficult back-engineering compatibility with such an FS would probably be.

As I said, I'm aware WinFS hasn't been taken seriously around here so far...but somebody needs to start to.

Re:Rushed?

[jonadab]

> Unlike 'Vista' (I'll always call it longhorn)

Vista is not Longhorn -- at least, not as Longhorn was envisioned at one time.

Longhorn, it was said, will use WinFS as its native filesystem. (It will include support for fat32, ntfs, fat16, iso9660, and possibly fat12, but these will be "legacy" systems, deprecated, and probably not supported for the main filesystem where the OS is installed, only for additional filesystems, such as on removable drives.) Vista will still use ntfs as its primary native filesystem; although WinFS can be added to it later, that is an add-on.

Longhorn, it was said, will include the new shell, Monad. Vista will not.

Oh, and Longhorn, it was said, will ship in 2004. Vista will not.

Vista is the Windows 98 SE of our time -- it's not the big upcoming release Microsoft has talked about for so long, and it doesn't have the capabilities that the big upcoming release was supposed to have, or in fact any new capabilities, and it's not going to be a compelling upgrade, but it has to come out, because it's just plain been *too long* since the last release and the market can't wait until the real thing is done.

Microsoft's release cycle gets lengthier with each passing year. Nine months after they finally release Vista, they'll be talking about the next big release (not the server version of Longhorn, I mean, but the successor to Longhorn), but you won't see *that* one for a good long while. Even giving them the benefit of the doubt and assuming they manage to get Vista out the door in 2006 as they currently say they're fixed to do, that means they'll start talking about Blackcomb in 2007, and by 1Q2008 they'll be predicting they can have it out by "next year) (2009), but the earliest you might possibly see it on store shelves is 2012, and frankly 2015 is more likely.

This is actually good news for the OSS community. It means we have a fairly good idea what the Microsoft desktop is going to look like for the next 7-10 years. Sure, there'll be add-ons, WinFS and eventually Monad, but add-ons are add-ons; if you want add-ons on an OSS system you can have Reiser4 today (though I don't know how stable it is yet -- but WinFS hasn't even been officially released, so I guess we're okay there), and Perl6 is likely to beat Monad to market, or in any case there are a number of excellent scripting languages available today; we haven't had to get by with just a bourne shell for quite some time, to say nothing of making do with the likes of cmd.exe. (Yes, there are people who advocate doing everything in old-school sh for the portability, but they're talking about portability in terms of running on fifteen-year-old systems; the Microsoft equivalent would be writing .bat files that will run under anything from DOS 5.0 upwards.)

I guess what I'm saying is that we know what to expect. Microsoft has grown large enough to become fairly predictable. That's good for the competition.