OpenBSD Journal

Microsoft to Charge for FAT File System

Contributed by jose on from the intellectual-property dept.

anonymous writes : "According to Microsoft, the Redmond company is going to charge a license fee for any product that is formatted in FAT by the manufacturer. Any manufacturer of compact flash memory cards or digital cameras may end up paying Microsoft as much as $250,000 for the use of the file format. The FAT File System is covered by several US patents.

Since OpenBSD implement FAT, we are now illegal ..."

The Microsoft intellectual property site is located here . I'd wait until their position is clear and its determined how they are going to enforce this before worrying.

(Comments are closed)

  1. By Tabsels () on

    Sounds like a great time to get that whole loadable kernel modules stuff rolling. From lkm(4): <br> <br> "Virtual file systems may be added via the LKM interface." <br>

    1. By Ben () on

      Compiled in or via loadable kernel module. Still illegal if Microsoft is going to push this.

      - Ben

    2. By anonymous () on

      actually, it should be possible to have this outside the kernel (in userspace). and other types of filesystems too. you are really not going to use fat for anything related to performance, so loss of performance is not an issue.

      1. By Fábio Olivé Leite () olive at on mailto:olive at

        Mtools has been around forever.

        1. By anonymous () on

          well, i still want the files in my namespace, not only accessible through some tools interface.
          so i still want it mounted.

  2. By Simon () on

    It is a license on storage devices that are formatted in FAT, not on the software used to due this. Nor is it on the software used to read FAT formatted storage. Of cause that could be the next step for Microsoft.

    I would agree that this action is worrying, but let us deal with the problems as they arises. OpenBSD is still perfectly legal.

    1. By Anonymous Coward () on

      This is superficially true, however, the idea of OpenBSD is that people can use it for whatever they want, so let's say somebody makes an access point based on OpenBSD using a flashdrive with FAT, they would then be subject to license costs, in contradiction to OpenBSD's goals.

      1. By Anonymous Coward () on

        but this has nothing to do with OpenBSD. It's not OBSD that made it ilegal, but the flash card of choice.
        If I want to store ilegal drugs on my safe and keep the combination of it on my OBSD encrypted driver, and the police cannot open the safe because of this, OBSD is helping me to hide my drugs.
        But that does not make OpenBSD ilegal, even though, helping to hide evidence is.

      2. By Anonymous Hero () on

        Bull. Full freedom to do with OpenBSD what one wants does not exist. Or are you claiming that it is legal to use an OpenBSD box to DoS someone? Illegal by law in most countries.

  3. By bolo () on

    The terms of the license are:

    * A license for removable solid state media manufacturers to preformat the media, such as compact flash memory cards, to the Microsoft FAT file system format, and to preload data onto such preformatted media using the Microsoft FAT file system format. Pricing for this license is US$0.25 per unit with a cap on total royalties of $250,000 per manufacturer.
    * A license for manufacturers of certain consumer electronics devices. Pricing for this license is US$0.25 per unit for each of the following types of devices that use removable solid state media to store data: portable digital still cameras; portable digital video cameras; portable digital still/video cameras; portable digital audio players; portable digital video players; portable digital audio/video players; multifunction printers; electronic photo frames; electronic musical instruments; and standard televisions. Pricing for this license is US$0.25 per unit with a cap on total royalties of $250,000 per licensee. Pricing for other device types can be negotiated with Microsoft.

    Since OBSD doesn't manufacture 'removable solid state media', I don't see that this is going to affect OBSD. And, if the FAT filesystem sub-system is, or was, made through a 'clean-room' process, there is little chance that M$ will come after OBSD. And I'm not even going to go into the grounds whereby the underlying patents could be overturned due to prior art or the widespread use by the applicant or industry for commercial use _before_ a patent was applied for it. Just my $.02.

    1. By Bruce () on

      A 'clean-room' implementation does not allow you to produce and distribute something which someone else has patented. Please get your propery rights straight.

      1. By bolo () on

        A patent on an idea or invention does not mean that _any_ means of doing the same thing is also covered by the same thing. Speaking as an ex-USPTO employee (God, did I hate it), a patent has "claims". These claims have to, in the case of software (the area that I worked in), have some tangible application (they have to be change the physical operation and be based on, in this case, computer hardware). Because someone implements the same functionality, but in a different manner, on a new invention, does not mean that the granted patent is applicable to the new invention. If it can be proved that the prior patent does not apply to _all_ the points of an invention, then the patent does not apply to the invention. Therefore, if a "clean-room" implementation of the same functionality of a FAT filesystem is created, and it can be shown to be "clean-room" or "black-box" or whatever you want to call it, then the patent does not apply to the "clean-room" FAT filesystem invention.

        :P*pththththth* So take that!

        1. By Anonymous Coward () on

          Clean room design processes are used to circumvent copyrights, they have nothing to do with patents:

          Patents can be circumvented by new designs which are sufficiently different from the original patent claims. How easy this is depends on how broad the claims are. There is no requirement that the second designer not have access to the original implementation, or the documentation provided by the patent for that matter, as there is in the case of clean room copyright circumvention.

          So the question for OpenBSD and similar projects then is whether you can implement FAT32 reading and writing capability without infringing on Microsoft's claims. I doubt their patent lawyers are that bad, and I wouldn't want to fight them in any event. If it becomes an issue, which it doesn't appear to be at the moment, strip out long filename support and live with 8.3. Hopefully that is the only thing they've managed to cover.

  4. By krh () on

    Their website says,

    "Microsoft's FAT file system license offers limited rights to issued and pending Microsoft patents on FAT file system technology, as well as rights to implement the Microsoft FAT file system specification."

    If their patent claims are valid, anyone who includes an implementation of FAT (such as OpenBSD) needs a license.

    Microsoft lists four US patents: 5,579,517, 5,745,902, 5,758,352, and 6,286,013.

    The first is about storing long filenames. This is the ugly hack that allows FAT to use more than 8.3 character filenames. I'm not aware of any prior art on this one. Most filesystems with long file names were built that way from the beginning.

    The second is about converting to and from long and short filenames.

    The third is about storing long and short filenames inside the same directory structure.

    The fourth is an API for long filenames in MS-DOS.

    Any implementation of FAT which includes long filenames uses patents one through three, because all that they do is describe long filenames in FAT. I don't think there's any way around them.

    The best hope to keep OpenBSD's FAT implementation is for someone to find prior art. I'm not a filesystems expert. Does anyone know of a filesystem that does something similar, i.e., stores long filenames inside short filename structures?

    If there is no prior art, the safest thing to do is remove long filenames from FAT in OpenBSD.

    1. By Anonymous Coward () on

      Wouldn't this mean that it's not the usage of FAT but the usage of the FAT32 extension that become problematic?

      If it were to affect OBSD, could they not just drop any support for long file names and circumvent any patnet issues?

      AFAIK anything that can read/write FAT can read/write FAT32, but is limited to the 8.3 truncated file names.

      1. By Anonymous Coward () on

        FAT32 has nothing to do with long filenames. FAT32 has to do with how large a filesystem can be. FAT12 is used on floppies. FAT16 was used by DOS and Windows up through 95, and could only handle 8 GB partitions. FAT32 originated with Win95OSR2 and allowed for larger than 8 GB partitions. FAT32x originated with Win98 and allows for larger than 32 GB partitions.

        vFAT (or virtual FAT) is what allows for long filesnames on FAT drives. vFAT originated in Windows for Workgroups 3.11, and was extended to use long filenames in Windows 95.

        As long as you don't use vFAT, you aren't stepping on any MS patents.

        1. By mirabile () on

          This is wrong, VFAT is just a driver for their
          protected-mode programmes.

          The file system is FAT, it comes as
          FAT12, FAT16 and FAT28 (commonly called FAT32)
          which don't differ except for the bootsector
          layout and FAT entry size.

          The original MS-DOS FAT driver, VFAT and our
          msdosfs are just implementations of FAT.

    2. By hopfgartner () on

      There should be prior art by Apple. When the Macs had to write their files on a FAT-fomatted volume, an extra file (IIRC in an invisible folder called desktop) was created. This file contained the map for converting the Macintosh 32 byte long filenames to FATs 8.3 long filenames (also for mapping other then ASCII characters, upper/lower casse etc.).

      I used the Mac in the early '90, so it's before the MS patents.

      1. By Sloppy () on

        OS/2's WPS did something similar; it created an extra file to hold the fake 'extended attributes' so when I looked at a FAT formatted disk with WPS, I would see longer names.

        The problem is that these approaches are just prior art for the GENERAL case of having long filenames on a filesystem that only supports 8.3 names. They don't cover the actual WAY that Microsoft did it for themselves, later.

        Microsoft's approach was much more integrated into the filesystem, and actually changed the mechanics for how the directories themselves are stored, rather than adding a file.

    3. By Anonymous Coward () on

      4dos' descript.ion file maybe? 4dos was/is a powerful replacement for of

    4. By Anonymous Coward () on

      right? Have you seen those 'TRANS.TBL' files that appear on some ISO9660 CDs? There are the RockRidge and Joliet extensions used for representing long filenames and adequate (Unix and Windows-like) permissions on top of a ISO9660 FS.

  5. By Anonymous Coward () on

    Pardon my ignorance, but do flash-type devices _only_ work using FAT? Couln't they be formatted with some other filesystem (e.g. ffs)? I'm thinking specifically of devices similar to the Soekris boards.

    I know devices like the above are in the minority. I would think, though, that if enough manufacturers are displeased at having to pay MS this paltry sum (let's face it 250K for even a small-to-medium sized manufacturer is a small cost of doing business) they could come up with their own filesystem with drivers for all client platforms for reading and writing. This just seems to be another way for MS to grab a few more dollars.

    1. By Anonymous Coward () on

      ffs works on compactflash

    2. By krh () on

      Unfortunately, FFS is much more complex than FAT. If you're making an MP3 player, you don't need much more than FAT.

      A comparable solution would be to invent a filesystem that was a simple as FAT, but had real long filename support. However, then it'd be more difficult to connect your device to others, because everyone else would need a filesystem driver for your new system. Even if such a driver would be easy to write, it would take a long time for that new filesystem to be widespread.

  6. By Anonymous Coward () on

    Projects like OpenBSD having access to FAT doesn't hurt Microsoft in the least.

    What they are trying to do is capitalize on the removable media market (flash cards, memory sticks, etc). The manufacturers in these markets will surely pay $250K to use the FAT file system - otherwise they will need drivers bundled with their products to recognize the new filesystem (as opposed to just plugging it in and it opens up - on Windows XP at least). Combine the costs of the development, distribution overhead, and a complete lack of compatibility - and it's obvious where this is going.

    Looks like Sony, Kodak, etc. owe Microsoft some pocket change.

    1. By SH () on

      For them it's pocket change, but I'm pretty sure they're wondering what's next from Microsoft.

  7. By Anonymous Coward () on

    Since OpenBSD implement FAT, we are now illegal.

    Did you read the first paragraph that you obviously copied and pasted into the submission box? Please, go back and re-read it now. You're so wrong it's laughable.

  8. By mirabile () on

    All comments I saw until now are so lame.
    You're assuming the OpenBSD msdosfs will be
    illegal now (it belongs dead because it can
    cause data corruption and kernel panic, but
    that's a different thing). It isn't.

    They only cite USA patents; OpenBSD is Canada
    based -> no problem (or at least, I don't see
    any). Same for me here in Germany (okay it is
    off topic).

    Don't forget the "strong crypto" stuff.

    1. By Anonymous Coward () on

      I'm sure Microsoft has also an international patent which would include Canada unless they are not smart enough to do that which is unlikely. Though not only patents must be filed but you have to look at the Microsoft licences though a microscope to be safe in any country which implies Canada also.

      just wanted to comment on OBSD canada based comment you made.

      1. By erik () on

        There are no such things as international software patents. As a matter of fact US software patents are not valid in Europe. (yet?)

        1. By Anonymous Coward () on

          sure are. There are international patents and doesn't matter if its software of poop patents. The countries enrolled in that process are where the patents apply

        2. By Anonymous Coward () on

          US offers international patents etc.
          Checking your source before you speek is always a good idea.....

          1. By mirabile () on

            Some *.gov site can't govern me, it would have to
            be a *.de site.

          2. By Anonymous Coward () on

            They offer international searches for patents and help to apply for patents in all countrys. But software patents is not legal in europe, so they will not apply here, and are rejected.

            Even if that will soon end.

            Why do i get the feeling that US think US Law apply everywhere else in the world.

      2. By mirabile () on

        If you do a clean-room reimplementation of the
        spec, you aren't bound to licences at all.
        And a mere licence can't extend a USA patent
        beyond their country's borders.

        1. By Anonymous Coward () on

          its just a link to a .gov site for info only you moron.
          the Paris treaty covers all countries with international patents including .de .....
          Your government doesn't have control in the international courts for such disputs either.

          A licence is not a patent and BTW microsoft has international patents filed meaning if they do have something for there filesystem it means wether you are in .de doesn't mean shit.

  9. By Fábio Olivé leite () olive at on mailto:olive at

    Heck, not even the boot floppies use the FAT format. If they did they could be considered "formatted in FAT by the manufacturer", but then they are not solid state storage devices.

    If later they come to demand licenses for any implementation, it's simply a matter of removing the offending code from CVS. It's not so important anyway.

  10. By Xenotrope () on

    I don't think anyone is seriously entertaining the idea that Microsoft is going to start suing people who use the FAT filesystem. The licensing issue at hand is largely one of manufacturers of preformatted devices. It remains to be seen if reading and writing to a FAT filesystem will become the next big lawsuit or not.

    What concerns me now, as it did when I first got wind of this a few days ago, is this nagging question in my mind: "Isn't this going to kill the floppy diskette?"

    Nearly all floppy disks sold in the U.S. (Maybe abroad, too. I can't say.) are preformatted, meaning Mom and Dad and Grandma can all buy a box of floppies and use one on their Windows machine without a second thought. It "just works".

    Think about it. People still rely on floppy disks for a lot of things, even though the technology is very old and very unreliable. I use them because it's a lot faster to write a boot floppy half a dozen times than it is to burn a CD-R that I can never use for anything else.

    Floppies are an important tool that should be there for when you really need one. If floppy disk manufacturers have to pay exorbitant amounts of cash to Microsoft for the privilege of selling something with Microsoft's proprietary FAT filesystem on it, they're likely to discontinue the product.

    Plenty of people don't know how to format a floppy disk, and if they lack that knowledge, not having preformatted floppies is going to hurt them. If they don't buy them, the manufacturers won't make them. That hurts everyone, even the people that don't use them with FAT.

    1. By blackc () on

      CD-RW is the answer here. While CD burners are
      not as cheap as floppy drives, CD media is much
      more reliable and I doubt that writing a floppy
      and booting off of it is faster than doing the
      same with CD-RW.

      1. By Vince () on

        I might be wrong but most systems without a decent CD Writer and the same software ie InCD etc... cannot read the same CDRW at all. Even those CDRW reading drivers mostly does not work with basic CD-ROM drives. so CDRW is not the solution. USB Thumb drives would probably be the next step!

Latest Articles


Copyright © - Daniel Hartmeier. All rights reserved. Articles and comments are copyright their respective authors, submission implies license to publish on this web site. Contents of the archive prior to as well as images and HTML templates were copied from the fabulous original with Jose's and Jim's kind permission. This journal runs as CGI with httpd(8) on OpenBSD, the source code is BSD licensed. undeadly \Un*dead"ly\, a. Not subject to death; immortal. [Obs.]