FAT licensing

2003-12-07

By now, you've probably all seen the article on Slashdot (and posted elsewhere, I'm sure) that Microsoft will be licensing the FAT filesystem to embedded systems/devices designers. This sparked a large volume of email on the fd-dev mailing list, with some suggesting that FreeDOS drop the FAT filesystem altogether and re-implement our own FAT32-like filesystem.

Lucho writes:

I see a strange discussion on fd-dev about a non-standard FAT32 filesystem. But one of the main goals of FreeDOS is compatibility, and this proposal contradicts with it! Why is all this worry? The problem is not ours! They ask the manufacturers of embedded systems or devices to pay for using FAT, not us or Linux! As as we don't link with or use any code of theirs, we're perfectly legal GPL-wise! So let's not worry and keep fixing kernel bugs and develop. Below is a refreshing quote from www.neuro-tech.net (note they mention FreeDOS).

FAT licensing.

Not licenses to get fat. I don't like parroting Slashdot, but has everyone seen the announcement about Microsoft now charging for use of the FAT filesystem? Great googa-mooga.

They're charging on either of two methods: Per chunk of media, like for SD/CompactFlash/floppy manufacturers, and per device (like a digital camera or MP3 player).

Now the expected thing is that both objects, when manufactured, incur licensing fees of 0.25USD. But the weird thing is each manufacturer is capped at 250,000USD.

This strange cap makes no sense at first glance. Then I started thinking that it's a huge PR move on Microsoft's part, and they're not really going to enforce any of it. Then I started thinking maybe not - maybe they will enforce it, and the two hundred and fifty grand is the limit on litigation costs.

It does seem though that they're not going to prosecute software that uses the FAT filesytem, which - I hope - leaves projects like FreeDOS, and oh, I don't know, Linux in the clear.

Or - and this is a big or - maybe they're just licensing their specific implementation of the patents in question. Aimless speculation is fun, isn't it?

Lucho also points us to this discovery on Slashdot:

(http://www.microsoft.com/whdc/hwdev/download/hardware/fatgen103.pdf) Just discovered this link in a comment over at Groklaw. Section 1.e. of this document would seem to indicate that MS has already granted the right to use FAT for hardware and operating systems:

(e) Each of the license and the covenant not to sue described above shall not extend to your use of any portion of the [FAT 32/VFAT] Specification for any purpose other than (a) to create portions of an operating system (i) only as necessary to adapt such operating system so that it can directly interact with a firmware implementation of the Extensible Firmware Initiative Specification v. 1.0 ("EFI Specificaation"); (ii) only as necessaary to emulate an implementation of the EFI Specification; and (b) to create firmware, applications, utilities, and/or drivers that will be used and/or licensed for only the following purposes: (i) to install, repair, and maintain hardware, firmware, and portions of operating system software which are utilized in the boot process; (ii) to provide to an operating system software runtime services that specified in the EFI Specification; (iii) to diagnose and correct failures in the hardware, firmware, or operating system software; (iv) to query for identification of a computer system (whether by serial numbers, asset tags, user or otherwise); (v) to perform inventory of a computer system; and (vi) to manufacture, install and setup any hardware, firmware or operating system software.

It doesn't seem like they could actually sue anyone for using FAT under this covenant, which is copyrighted 2000.

Finally, Lucho points to this interesting post on Slashdot:

I thought there were similar filesystems, and besides FAT is so simple, a cleanroom implementation would not take long, hence no need to licence

This is certainly not true. With copyright law, it's illegal to copy code. With patent law, ideas are patented. Wheter it's implemented in a 'clean room' or not, that really doesn't matter. THAT's the reason why we detest software patents in the first place!

To be able to bring out preformatted FAT flash devices without paying the Microsoft license, one would have to claim rights to 'prior art'. In contrary with copyright law, however, it's the responsability of the IP holder to come down on the infridger (so as long as you don't get a letter from MS, you aren't obligated to take action).

Yet IANAL but in my past businesses talked about these issues alot with lawyers. Regarding the question wrt European manufacturers usage of the FAT filesystem. First needs to be seen if these patents are also valid in Europe or not. After initial issuing a patent in Europe, US or Japan it's automatically valid for 3 years in all of these regions. After this period it needs to be registered in the specific region. As I presume these are quiet old patents, one should look into this.

However, there still is controversy regarding software patents and its enforcebility in Europe. European software patents should also have a hardware part. This license has a hardware part, but the patents themselves not. You might want to consult a patent lawyer to verify this, but I would bet that it's unenforceable in Europe. However, I wouldn't bet on this for 250k USD

A lot of smaller device vendors will probably sell the unformatted version after they receive letters from MS (which is a pity as FAT is readable/writable by Win/Mac/Linux).

A lot of users will now unknowingly format their cards using NTFS making it harder to exchange data with non-Windows users...

Regarding the FAT driver in Linux; as this MS license only speaks of preformatting digital media in the FAT filesystem, this is not an issue today. Could Microsoft ask money for inclusion of the FAT driver in the Linux kernel? Remember, patents are about ideas, not about the actual implementation or even in which language certain algoritms are written (it's about what is accomplished, not about how it's actually done). So as the FAT filesystem is patented technology, they could theoretically take action. However, the action needs to be taken by them first. If 'prior art' can prove that the Linux implementation is based upon technology very simular than the patents issued, a case in court might prove the patents to be not really valid.

Such a thing would also destroy all possible revenues from licensing programs such as these (it's higly unlikely that device manufacturers will try to prove they had access to prior art, the long-bearded fs developers in the OpenSource community are probably less easy to convince - especially since the patents where only filed in 1995).

If the outcome of a legal case would be different, chances are higher that distributions would just drop the filesystem driver instead of paying money to Microsoft.

So, to me it seems that Microsoft would have more to loose than to gain from going after the FAT driver in Linux.