Microsoft aiming IBM-scale patent program at Linux? (Register) [LWN.net]

The Register speculates on Microsoft's plans for patent enforcement. "Microsoft's actions so far don't constitute a full frontal attack on free software. It's often been rumored that Microsoft has a number of patents - the number varies - on the Linux kernel itself. But it has chosen not to pursue such an inflammatory tactic, just yet, and may not even need to at all in order to succeed."

to post comments

Microsoft aiming IBM-scale patent program at Linux? (Register)

Posted Dec 8, 2003 16:11 UTC (Mon) by vblum (guest, #1151) [Link] (4 responses)

It seems that the FAT implementation in the Linux kernel would have to fall under MS's patents - or is there some more subtle detail here?

Anyway, this is bad news - they are starting with legacy applications first, but this is a warning shot only.

Microsoft aiming IBM-scale patent program at Linux? (Register)

Posted Dec 8, 2003 16:36 UTC (Mon) by stumbles (guest, #8796) [Link]

Yes, I think your right this is only the beginning with Microsoft. The more open source
is adopted the more Microsoft will attempt to throw up walls between their software and
anyone else that does not jump on their boat. That's ok. If they continue on that track it
will further isolate them from the demands of business. While Microsoft is looking only
at their bottom line, so are all the other businesses. On that point Microsoft thinks a
quarter is no big deal. But that quarter multiplied by hundreds of thousand units of
whatever is a big deal to companies that would rather have that quarter.

It's not really surprising they have/will do this. After all, in general every monoploy has
used similar tactics against their competitors and customers. Any thing to lock them in.
Often the opposite effect is to alienate themselves.

In a way I see them trying to put themselves into a similar situation as public utilities
where your choices about who you do business with is extremely limited or there is no
choice at all.

Microsoft aiming IBM-scale patent program at Linux? (Register)

Posted Dec 8, 2003 18:51 UTC (Mon) by JoeBuck (subscriber, #2330) [Link] (2 responses)

First, the "FAT patent" covers the encoding of long filenames, not FAT itself.

Just because someone has a patent does not mean that the patent is valid. In fact, when a patent fight reaches court, it is quite common for the patent to be overturned.

The USPTO, I've read, give patent examiners an average of 25 hours to do the entire review for a patent, which is why so many questionable patents are granted.

actual patents

Posted Dec 8, 2003 19:04 UTC (Mon) by vblum (guest, #1151) [Link] (1 responses)

http://www.microsoft.com/mscorp/ip/tech/fat.asp

lists four patents regarding FAT (plus unspecified further ones). Do all these cover file name encoding?

Yes, patents may be overturned. But the amount of legal FUD that can be launched with all this is potentially huge. And, some patents may actually be enforceable.

Anyway, I was curious whether the actual patents cover the implementation in the Linux kernel. Is that so?

actual patents

Posted Dec 8, 2003 19:40 UTC (Mon) by drathos (guest, #6454) [Link]

I just looked up all four on the USPTO site. The earliest of them is dated Nov 26, 1996 (filed April 4, 1995) and, based on the summaries, they all relate to maintaining both long and short filenames in the filesystem.

Microsoft aiming IBM-scale patent program at Linux? (Register)

Posted Dec 8, 2003 16:33 UTC (Mon) by ccchips (subscriber, #3222) [Link] (6 responses)

Microsoft is at risk on the security front. If enough people get fed up with security problems in Microsoft products, they might look seriously at alternatives such as Linux. This would badly impact their revenue stream.

In the past, I had the impression that Microsoft people weren't all that crazy about software patents, and were operating mainly in a defensive and/or exchange mode. But what would happen if the company started to really sink?

I have really cut back on my need to transfer files between Windows and Linux partitions on the same machine, but I'm starting to feel as if it might be a really good idea for GNU and other free-software supporters to get those Windows-side EXT2 and EXT3 file systems working...

...or for IBM to start working out some deals with Microsoft....

Microsoft aiming IBM-scale patent program at Linux? (Register)

Posted Dec 8, 2003 16:38 UTC (Mon) by stumbles (guest, #8796) [Link] (3 responses)

I would rather see a driver developed of Windows that would allow ext2 or some other
non-windows filesystem to be used on such devices.

While I am glad IBM has been such a great contributor to Linux, I would not want them
to cut some deal with Microsoft.

Microsoft aiming IBM-scale patent program at Linux? (Register)

Posted Dec 8, 2003 16:47 UTC (Mon) by proski (guest, #104) [Link] (2 responses)

It doesn't matter if such driver exists or not. Users of digital cameras expect that they connect the camera to the PC and the contents of the flash card is shown as a drive. They don't have to install any software now. Installing software on Windows is still a barrier for many users, and the manufacturers of digital cameras and other devices don't want to lose those users.

Another filesystem supported by Windows is iso9660 (used on CD). NTFS is not supported by Windows 98 and Windows ME. But I'm not sure it will work out-of-box like FAT on non-CD devices, and it's working out-of-box that matters here.

Microsoft aiming IBM-scale patent program at Linux? (Register)

Posted Dec 8, 2003 18:08 UTC (Mon) by yohahn (guest, #4107) [Link] (1 responses)

Actually, how many of these devices use long filenames. If what I have read elsewhere is correct, all of microsoft's patents have to do with storing both a long filename and a short filename for each file. Why not just create a FAT compatible system which does not do this? In place of each long name (or each short name) put nothing.

Not ideal, but would mostly work.

(N.B. I have not read the patents and even if I did, I'm not a patent lawyer, so I really have no idea if this would cya or not. It SEEMS to me that it would)

Microsoft aiming IBM-scale patent program at Linux? (Register)

Posted Dec 8, 2003 19:19 UTC (Mon) by pjs (guest, #10927) [Link]

The 32 byte FAT directory entries that are used to store the long filename do NOT store critically important information about the file, specifically the starting cluster, size, attribute bits, and modification time. So, it is quite easy to have only short 8.3 filenames, but there really is no option to have long filenames without their 8.3 counterparts.... at least within the scope of FAT, which is implemented in every almost operating system and countless portable devices.

Microsoft aiming IBM-scale patent program at Linux? (Register)

Posted Dec 9, 2003 11:10 UTC (Tue) by amikins (guest, #451) [Link]

There's already EXT2 drivers out there for various WinOSes, with (optional) write support. It's not exactly robust as of yet, but I've made use of it on a few occasions. Some also can do (read only) access of EXT3. (mostly so that my gaming Win2k partition can access my MP3 collection; Even when I'm doing read-only, I don't like using non-journaling filesystems)
Though, I don't think it'd be even remotely trivial to get Windows to /boot/ from an EXT2/EXT3 partition on any version of Windows.

Sadly, at this moment I don't have access to my links list for what I've used. Fortunately, someone who wants more info can always use google. =>

Microsoft aiming IBM-scale patent program at Linux? (Register)

Posted Dec 9, 2003 23:02 UTC (Tue) by emak (guest, #488) [Link]

There is a driver for Ext2 & Ext3 :

"What is EXT2IFS?

EXT2IFS is an Installable File System Driver (IFS) for Microsoft Windows NT 4.0 and Microsoft Windows 2000 (NT 5.0).

The driver can read the Second Extended File System (EXT2) and Third Extended
File System (EXT3)"

http://uranus.it.swin.edu.au/~jn/linux/ext2ifs.htm

FAT not the problem

Posted Dec 8, 2003 18:05 UTC (Mon) by ncm (guest, #165) [Link] (2 responses)

It will be easy to work around the FAT patents, for most uses. They only apply to creating (or renaming) files with long names. To copy files between MSWindows and our systems, we can just avoid the long names. Cameras don't create files with long names, they can easily stick to 8.3 names. Flash memory manufacturers can format the devices with no files on them at all, so they also are not affected. However, the vfat driver in the kernel might need to be pared down.

The bigger issue is what other patents MS has. Probably if IBM cuts a deal with MS, it will be that MS promises not to enforce anything against Linux, in any form. MS is likely to cheat by "encouraging" nominally-unrelated companies to attack Linux, just as they have bankrolled SCO.

FAT not the problem

Posted Dec 8, 2003 18:52 UTC (Mon) by JoeBuck (subscriber, #2330) [Link] (1 responses)

You're assuming that the patent is valid; first look for prior art.

FAT not the problem

Posted Dec 9, 2003 4:17 UTC (Tue) by larryblunk (guest, #12587) [Link]

Patent 5,579,517 mentions a couple of articles from PC Magazine
on "long file names and extended attributes" that were published
in April and May of 1990. They were written by one of the patent
authors, Ray Duncan. However, the initial patent application
was not filed until April 1, 1993. I was under the impression
you only had a year to file from the date of the initial publication
of the details of your patent claims. Perhaps, the extensions
mentioned in the PC Magazine articles are substantially different
from what's in the patent claims. However, it would seem to be
worthwhile to look up these articles. I wonder if the patent
examiner even bothered to do this?

Waiting for European software patents?

Posted Dec 8, 2003 18:31 UTC (Mon) by hazelsct (guest, #3659) [Link]

I've suspected for years that Microsoft has been waiting for the EU to harmonize software patents, hopefully in their favor, before doing something like this. Then the EU, like the US, would be forced to either pay Microsoft to run all software (Windows, Linux, or anything else), or else reverse their previous legislation on patents, the latter of which is very hard to do. In the meantime, if they were to start enforcing first, then that would likely tip the European Parliament against them.

Perhaps they're starting to give up on the EU harmonizing in their favor, or have some dirty tricks to pull in the background to swing it the other way, or else (as the article suggests) they're so confident that they think people don't care.

In any case, this presents an opportunity to remind MEPs of the looming potential danger of such patents...

UMSDOS and ZipSlack

Posted Dec 8, 2003 18:48 UTC (Mon) by busterb (subscriber, #560) [Link] (3 responses)

Does anyone know if this patent would affect UMSDOS, a Linux FS that uses 8.3 FAT names and maps them to long, unix-compatible filenames along with unix file permissions using extra files to hold the meta-data. I know that ZipSlack and several other Linux-on-a-dos-partition distros use this method to install Linux without reparitioning. I don't know if UMSDOS uses any stuff from MS's long filename support.

UMSDOS and ZipSlack

Posted Dec 8, 2003 21:04 UTC (Mon) by proski (guest, #104) [Link] (2 responses)

/* 
 *  linux/fs/umsdos/inode.c 
 * 
 *      Written 1993 by Jacques Gelinas 
 *      Inspired from linux/fs/msdos/... by Werner Almesberger 
 */

VFAT first appeared in Windows 95 released two years later.

UMSDOS and ZipSlack

Posted Dec 9, 2003 10:07 UTC (Tue) by forthy (guest, #1525) [Link] (1 responses)

UMSDOS is either prior art or not affected by the patent, because it's implementation predates the filing of the patent. Other prior art of having long and short filenames together in one directory are the Rock Ridge extensions of the ISO 9660 filesystem (NM record for alternative name). The question still is if the way Microsoft does it is novel and different enough to qualify as innovation. IMHO Rock Ridge is closer to vfat than UMSDOS, which keeps the additional metadata in a file, not within the directory itself.

UMSDOS and ZipSlack

Posted Dec 9, 2003 12:27 UTC (Tue) by IkeTo (subscriber, #2122) [Link]

The idea of UMSDOS is to keep a separate file to tell which 8.3 filename "maps" to the long file name, and to store the uid, gid and umask which is also needed for a Unix-style directory. I don't think it can serve as a prior art of VFAT, which uses directory entries which are interpreted as "deleted" in the old filesystem implementation to hold the long filenames. The former would leave an extra file visible by the old 8.3 filesystem implementation, the latter would not. The former is generic and can be done to any filesystem, the latter is specific to FAT and won't apply to anything else.

I think the patent is probably valid (unless the objection that it is filed too late would hold some water). On the other hand, I don't think it really make sense for MS to ripe on it, since it would essentially kill the file format: there are dozens of better filesystem that (1) are free of patent issues, (2) have free implementations, and (3) have better performance than VFAT. Once MS start collecting patent money, vendors will start ditching the format (or fallback to 8.3).

Incomplete Analysis of Patent Problems

Posted Dec 8, 2003 19:49 UTC (Mon) by josh_stern (guest, #4868) [Link]

As software developers, we frequently get outraged at the admin of
software patents by the USPTO
in part, because what is patented often seems like the sort of
trivial solutions to a given problem that developers would come to on
their own in their everyday work.

This is related to a general (and theoretically recognized) concern
about patents that they often do not strike a desired balance
between encouraging innovation (intended consequence of patent law) and
restricting competition (considered to be an unintended or necessary
evil part of the law).

But the importance of network externalities in modern technology
fields is related to a whole different type of theoretical problem
with patents that hasn't received enough attention. That is that
the value of a granted patent will often lie in its ability to
act as a gatekeeper for a technology standard rather than in any
intrinsic technological capability of what is patented. This
case of VFAT would appear to be a classic example of this. FAT
itself was a technologically backward filesystem even at the time
of its development in the 1980s. It is only significant because
of widespread use. Similarly, VFAT/FAT32 are only signficant
because of their standard use. Only Microsoft had the marketplace
capability to make any extension to FAT a standard. So allowing
a patent on this type of thing is simply a covert mechanism to
restrict competition and/or collect monopoly rents without any
element of encouraging technological innovation (arguably the patent
capability offers both incentive and mechanism to do the reverse).

Some links for those who didn't see the patents

Posted Dec 8, 2003 20:27 UTC (Mon) by proski (guest, #104) [Link] (4 responses)

Since so many users here are asking about the patents in question and the the exact claims as voiced by Microsoft, let me just post this information. Maybe other users should do their research and post links rather than specualte and ask questions that can be answered by simply looking at the patents.

Microsoft: FAT File System Technology and Patent License
U.S. Patent #5,579,517
U.S. Patent #5,745,902
U.S. Patent #5,758,352
U.S. Patent #6,286,013

The first of those patents was granted November 26, 1996, years after UMSDOS was implemented in Linux.

Some links for those who didn't see the patents

Posted Dec 8, 2003 21:53 UTC (Mon) by Ross (guest, #4065) [Link] (2 responses)

And don't forget the ISO9660 extensions for Unix (RockRidge) which add long filenames, ownership, permissions, devices, and symbolic links. Without that ISO9660 only supports uppercase filenames under 30 characters (and Windows actually restricts it to 8.3).

Some links for those who didn't see the patents

Posted Dec 9, 2003 1:14 UTC (Tue) by khim (subscriber, #9252) [Link] (1 responses)

DOS restricts it to 8.3 (by obvious reason). Windows with 32bit driver (most Windows systems today) are just fine.

Some links for those who didn't see the patents

Posted Dec 11, 2003 9:13 UTC (Thu) by Ross (guest, #4065) [Link]

Ah, but Microsoft has its own ISO9660 extensions (they had to reinvent
the wheel) called Joliet and something else which I can't remember at
the moment :)

At least NT4.0 had problems with 30 character filenames on a CD I burned
about 4 years ago...

Some links for those who didn't see the patents

Posted Dec 9, 2003 20:43 UTC (Tue) by edgewood (subscriber, #1123) [Link]

The first of those patents was granted November 26, 1996, years after UMSDOS was implemented in Linux.

The important date for prior art, as I understand it (IANAL), is the patent application date, not the date it's granted.

Microsoft aiming IBM-scale patent program at Linux? (Register)

Posted Dec 9, 2003 1:02 UTC (Tue) by foo@share-foo.com (guest, #7940) [Link]

Hopefully this will encourage the long needed legal reform here in the states. Or perhaps it will encourage a real monopoly busting court case that should have occurred half a decade ago.

Microsoft aiming IBM-scale patent program at Linux? (Register)

Posted Dec 9, 2003 8:58 UTC (Tue) by Zelatrix (guest, #5163) [Link]

I thought that patent licensing was the responsibility of the user, not the developer (although usually developers obtain the patent licenses on the user's behalf). Wasn't that how users of Microsoft SQL Server - not Microsoft - got into difficulties when some aspect of SQL Server turned out to be in the scope of someone else's patent?

Given that anyone using the Linux FAT implementation will be doing so in order to interoperate with Windows, they already have a licence to use any patents in FAT. No?

I'm not a lawyer by the way.

FAT patents in europe

Posted Dec 9, 2003 10:49 UTC (Tue) by dps (guest, #5725) [Link] (2 responses)

IANAL but have heard something about patents and copyright.
- Patents cover ideas and claiming you came up with it independlty out of pure woodsmoke are not defence.
- Copytights cover expressions of ideas. If I can prove I wrote my word for word indepentical novel independently on pure woodsmoke, you have no case. (In such an extreme exmaple really solid evidence of pure woodsmoke might be advisable).

Indeed you can copyright almost anything, for example some bits of german countryside (albeit bits that were being passed off as many places nowhere near germany).

Almost everywhere except the US *everything* patented and otherwise published is "obvious" and therefore can not be patented. AS such the 1990 articles would make the FAT patents impossible in EU even if you worked around the "not software" requirement---putting the algorithm or data format in a device or on an IC, for example (courts have held a licnce is required for software implementations of the alogrihtm on your patented IC).

There is a least one case of the US government and a large biotechnology company being denied a patent in the EU due to prior art. Somehow the prior art, which predated the US filing date by many years, failed to sink the US patent AFAIK.

In the US you can restrict the scope of "obvious" to things not published by yourself (but not exclude other people's comments on what you publish). This is why RSA was only ever patented in the US. The IDEA people did not publish prior to filing and did get both european and US patents.

FAT patents in europe

Posted Dec 9, 2003 12:27 UTC (Tue) by Wol (subscriber, #4433) [Link]

Incidentally, RSA was a UK trade secret for some five or so years before the patent was filed ...

That's another problem with patents. If X files a patent, and Y was using the exact same idea as a trade secret, I think X can demand royalties from Y or even stop them using the idea! Even if they can prove their trade secret is many years old!

Cheers,
Wol

FAT patents in europe

Posted Dec 9, 2003 13:50 UTC (Tue) by sammythesnake (guest, #17693) [Link]

(Posted Dec 9, 2003 10:49 UTC (Tue) by dps) (Post reply)

Copytights

indepentical

exmaple

Are you George Dubya Bush, by any chance? ;)

Cheers & God bless
Sam "SammyTheSnake" Penny