In a move that could ultimatety deal a heavy blow to Microsoft’s intellectual property rights, the U.S. Patent and Trademark Office will re-examine patent the company holds on the FAT (file allocation table) file system, a format used for the interchange of media between computers and digital devices.
Microsoft’s FAT Patent Under Review
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
This is something that has been on my mind for a while. The long to short filename conversions of fat32 and ntfs pose a major issue for the ReactOS project. We implement fat32 now but we are in legal limbo as Microsofts patent license is incompatible with the GPL. Also this patent has posed issues for Linux developers that support vfat.
Does anyone know of someone else who can make a claim to have developed the FAT file system? Don’t start a tirade about it being a bad thing, just name the company or individual who developed it.
Shouldn’t the patent have expired by now? This patent seems ridiculous and should fall.
“Shouldn’t the patent have expired by now? This patent seems ridiculous and should fall.”
Except they were granted the patent in 1996.
> This patent seems ridiculous and should fall.
This patent is a masterpiece of technology compared to other patents. Double-click, progress bars, heartbeat, index by date, anyone?
Ugh. Have you bothered to read the actual patents, or are you just talking about the sensationalized summaries on /.? You should realize that if the patents really were as bad as those ridiculous summaries say they are (and they aren’t), other companies would sue to fix the problem.
FAT is such a legacy filesystem as it stands & has been far superceeded by open source alternatives I really don’t see what their problem is.
it’s not like giving away the rights to FAT would help the open source interpolate with Windows because given the size of todays drives who would format with FAT?
Why companies insist on hanging on to worthless technologies so zealously is beyond me.
Mainly, FAT is used in portable devices that don’t need advanced features of NTFS but still need to interoperate easily with Windows machines.
The patent isn’t on FAT itself, but rather on the idea of storing files with a long name on a filesystem that was only designed for short names. FAT itself wasn’t invented
The patent is invalid as several other systems have done it. Linux did it with the UMSDOS FS – it created an extra file to store the extra info, but made the entire process invisible to applications. GEOS stored a header with additional attributes at the beginning of files. Similar to the Linux approach, this was done in the filesystem driver and was transparent to the rest of the system. 4DOS also did similar tricks, but as it was only a layer above DOS, applications had to be aware of it to take advantage.
I don’t know when UMSDOS was written for Linux, but GEOS was doing the long filename support in 1990.
The difference in what Microsoft did is where on the disk they stored the extra information. The idea had been implemented before. The only reason other people didn’t store the extra information within the FAT was to maintain compatibility with Microsoft’s software.
> Have you bothered to read the actual patents,
Yes, but just the European ones. They’re hard enough to decipher.
> if the patents really were as bad
They are.
Why companies insist on hanging on to worthless technologies so zealously is beyond me.
In this case I’d imagine that it has to do with the growing number of consumer devices that use FAT, like USB storage, and mp3 players, each of which is now generating some small amount of income for Microsoft due to the patent, and locking other OSs out (due to a lack of patent licenses) of interoperating with these beasts.
I agree that it’s so old that it’s rediculous to have granted them a patent, but I’m biased. I have no respect for the entire “IP” concept.
“The patent is invalid as several other systems have done it. Linux did it with the UMSDOS FS – it created an extra file to store the extra info, but made the entire process invisible to applications. GEOS stored a header with additional attributes at the beginning of files. Similar to the Linux approach, this was done in the filesystem driver and was transparent to the rest of the system. 4DOS also did similar tricks, but as it was only a layer above DOS, applications had to be aware of it to take advantage.”
And why didn’t any of those others bother to file for a patent? Yup: This would be where the problems begin.
It’s not exactly who did what first; it’s who filed the patent application first that really matters. As a reference look back as far as television and other devices.
Who holds the patents, and who in reality should of are often different things entirely.
“Prior art” as it is commonly called, seems to have very little weight at the Patent Office. Unless the one who demonstrated the art files the complaint, and has a boatload of documentation.
“Who holds the patents, and who in reality should of are often different things entirely.
“Prior art” as it is commonly called, seems to have very little weight at the Patent Office.”
we have immense amount of documentation and can very well show all prior art of umsdos and geos and several such systems. prior art doesnt depend on prior patents. only a clear proof of it
The problem is that FAT32 is precisely the only filesystem that can be used in both Windows and Linux, or any other UNIX for that matter. If I could get a filesystem driver for any of XFS, JFS, ext3, Reiserfs and which ever other to run flawlessly on XP, at least for simple storage, I would convert my FAT partition in no time.
This would be taking the game to MS, although they could just get a service pack that ‘unfortunately’ breaks the driver “DOS ain’t ready till Lotus won’t run” anyone.
But seriously, I think this is a point where the community could galvanize itself into action and provide something tht would help interoperability.
Why FAT? Wouldn’t there be many more important patent claims and already accepted software patents to re-examine?
So – they will do a little re-examination – in order to make software patent critics happier? But will something really happen? Will the US patent office “deal a heavy blow to Microsoft’s intellectual property rights”? I seriously doubt it.
After what stupid (MS) software patents have already been accepted (double clicking, to do lists (= comments) in program code etc.) things don’t look too promising at all.
The US Patent Office, and the US government, don’t seeem to much care for the future of software development, small IT companies, competition and innovation. Money and power got from a few big US software companies is what matters to them. I hope other countries won’t do the same mistakes.
It’s not exactly who did what first; it’s who filed the patent application first that really matters. As a reference look back as far as television and other devices.
Who holds the patents, and who in reality should of are often different things entirely.
“Prior art” as it is commonly called, seems to have very little weight at the Patent Office. Unless the one who demonstrated the art files the complaint, and has a boatload of documentation.
Major flaw in the system. This means that if entity A has implemented something which was done before by entity B and C, while B doesn’t exist anymore and C doesn’t have interest in complaining (there are loads of reasons for this: self-interest in case, lack of money, smear money, not notticed, etc) then a potential big group of people can have a disadvantage because of the patent has been granted to entity A while they’ve got no possibility to fight it directly via the authority (regulators, USPTO in this case).
The danger of a patent isn’t only wether it’s valid or not. A danger which should not be underestimated by us is the threat of an granted (or even requested) patent. Even when it is invalid or ridiculius it still serves as FUD.
In this case, FAT is a milk-cow, forced because FAT is well supported while Windows doesn’t support many other FSes. If Windows supported more, while other OSes would support these too, one of these (depends on implementation ofcourse) would potentially be more of a standard because it is more Free (as in freedom; ie. non-patentable) or simply better.
Nevertheless, kudo’s to PUBPAT.
Here’s more info http://www.groklaw.net/article.php?story=20040611205823281
“If I could get a filesystem driver for any of XFS, JFS, ext3, Reiserfs and which ever other to run flawlessly on XP, at least for simple storage, I would convert my FAT partition in no time.”
I have great news for you!
Ext2FS and Ext3FS run flawless in read/write support with some 3rd party drivers, under Windows NT 4.0 and deritives. A few are for Solaris, DOS or MacOSX too.
http://uranus.it.swin.edu.au/~jn/linux/ext2ifs.htm
http://uranus.it.swin.edu.au/~jn/linux/explore2fs.htm
http://www.it.fht-esslingen.de/~zimmerma/software/ltools.html
http://tarzan.cr.cyco.com/~andreys/ext2fsnt/
http://www.partition-manager.com/n_ext2fs_main.htm (proprietary)
http://sourceforge.net/projects/winlfs/
http://sourceforge.net/projects/ext2fsd/
http://sourceforge.net/projects/ext2fsx/
There’s probably even many more.
READ what and HOW WELL they support it. Expect data loss. But i’ve heard good comments on some of them. I myself had not so much luck with Explore2FS in ~2001. With write support that is. Fsck
Problem is rather: it ain’t native.
Must have been earlier, around ~2000 since IIRC i did run Explore2FS on Windows 95 or Windows 98 back then. Not many of the above support Windows 9x versions (IIRC only 1 or 2) so that might be of interest when you run such version. If you run a NT version, those are well and commonly supported. Especially, there are many software packages which have the full features on (ext2, 3, read, write, etc) and work on 2000/XP. I’ve noticed that at least on the Freshmeat.net entry where i found Explore2FS when i weas searching for the above software links, write support was still deprecated.
“we have immense amount of documentation and can very well show all prior art of umsdos and geos and several such systems. prior art doesnt depend on prior patents. only a clear proof of it”
By all means provide some details to enrich your tale. Inquiring minds want details, and your post as Anon, certainly isn’t a show of force in this field.
Major League flawed and sucking wind…… BUT, it’s a geek hating machine out there incase you ain’t already noticed.
MS owns a huge number of patents. Are they / have they ever enforced a single one?????????
Name one.
there is a very good ext2 file system driver for windows which support read and write. And it is actively maintained. http://ext2.yeah.net
@Don Elings: there’s one for you: Virtualdub and ASF support.
anything that’s encumbered is like the sword of damocles hanging over your head.
just because MS hasn’t sued anyone in the past doesn’t mean they won’t resort to that in the future (if the time comes that they feel sufficiently threatened, like by open source)
THe ASF case is a bit disturbing..
What’s the big deal?
Why not just sell portable devices unformatted and let the user format them in any way he/she choose?
Why not just sell portable devices unformatted and let the user format them in any way he/she choose?
It’s possible, but it will require some degree of cooperation from the manufacturers. For example digital cameras will have to understand many different file systems, so that the user can have some choice. Currently i can format a memory card used by my camera in, say, ext2, but then the camera will not know how to save pictures in such filesustem. It only knows FAT. MP3 players have the same problem.
“@Don Elings: there’s one for you: Virtualdub and ASF support.”
Here is a good dose of that story:
http://www.linuxmafia.com/faq/VALinux-kb/asf-support.html
Depending on your point of view, MS actually owns the rights to that format. (and others as well)
Many Codec producers have fought to keep their format protected over the past few years, this is far from an isolated situation.
It’s possible, but it will require some degree of cooperation from the manufacturers. For example digital cameras will have to understand many different file systems, so that the user can have some choice. Currently i can format a memory card used by my camera in, say, ext2, but then the camera will not know how to save pictures in such filesustem. It only knows FAT. MP3 players have the same problem.
This is unfortunate. Both of those devices can normally be update with firmware updates. Perhaps that may help.
Support for alternative open file systems should definitely be included in future devices. I’m sure if all the device manufacturers got together they could design fully working file system drivers that could be installed onto Windows at the same time as the device software/drivers. After all they would be the ones to immediately benefit from not having to worry about MS licensing.
Things might not be as Plug and Play as they are now but that will only be taking as back a few years. If this were successful then MS would eventually have to support this FS by default. As a nice side affect we would have a fully reliable cross platform FS.
No one seems to complain about having to install extra software in order to access the HFS+ file system on Ipods anyway.
Oh the dreams we make when we have too much time on weekends.
The problem is that FAT32 is precisely the only filesystem that can be used in both Windows and Linux, or any other UNIX for that matter.
The best bet is to set up a second machine with a large disk running Linux/BSD with samba. You can access your files via samba in windows, or nfs in Linux et al. Plus, you can reinstall stuff at will without moving a ton of data around.
I have asked the exact same question on Slashdot two weeks ago. You can read the topic here: http://ask.slashdot.org/askslashdot/04/06/01/1810201.shtml?tid=137&…
In two words, I was answered that there’s no working alternative to NTFS or FAT. Reasons are that
– the dev kit for writing FS driver is expensive ;
– writing a filesystem driver is hard ;
– writing a filesystem is boring.
in this jusrisdiction, microsoft cannot clain owner ship of 8.3, “double-click” or anything else.
“in this jusrisdiction, microsoft cannot clain owner ship of 8.3, “double-click” or anything else”
So what, yet, or how come?
Believe me when I tell you this U.S.patent “thing” has global impact.
This deal with the U.S. Patent Office will cross a lot of borders not yet realized. It’s just the way it works………
If you think for a second the U.K. is immune from events in the U.S.A. Please take a another look, and rethink your posititon.
“Does anyone know of someone else who can make a claim to have developed the FAT file system? Don’t start a tirade about it being a bad thing, just name the company or individual who developed it. ”
Doesn’t matter, it’s over 20 years old. Filing for a patent that many years after you made your invention popular is not vaild. Patents are to be filed after invention, not after you realize you can kill your competition or get rich.