Guest post by As seen in this TechFlash article, Microsoft has launched a patent suit against TomTom, a seller of (Linux-based) navigation devices. “It’s believed to be the first time Microsoft has filed a patent suit over Linux, after claiming for years that elements of the open-source operating system violate its patents. However, Microsoft says open-source software is not the intended focal point of the action.” The lawsuit focusses on several navigation techniques, and includes a few patents related to Linux. LWN lists the individual patents in question.
I read the discussion on Slashdot: it’s not really related to Linux in any way. One of the patents is related to FAT32. TomTom uses FAT32. If TomTom would have used a different OS and would still have used FAT32 TomTom would still be sued over the patent violation.
Well, all Linux kernels since .. like forever .. ship with FAT32 support. How is this not related??
Just because they don’t ship ON a FAT32 formatted device?
While the linux related aspects of this case seem to be circuitous, it represents a new (at least for outsiders like me) quality of relationships between Microsoft and their competitors on several levels.
It may be difficult to find a larger embedded vendor that does not use Linux in one of its products, but they specifically targeted a player that is known of its involvment in Linux (on the device site, operating their devices from outside MS Windows on the user side of things is still a big PITA) even outside of the circles usually interested in embedded operating systems. Everytime the discussion deals with Microsoft and patents, folks on the Microsoft side of fence tend to point out that MS never did weighted their patent portfolio against a direct competitor and therefore all reservations about MS technologies and patents should be attributed to pseudo-religous zealotery of some FOSS whackos.
I don’t know about you, but the “stay clear of MS patented technology, they may decide to go after you one day” mantra suddenly looks a lot less like unfounded paranoia (not that I ever thought it was, but now I have a handy browser bookmark for future reference).
Then there is the issue of the “wait it’s no cross license for patents” agreements that MS did the last years. It may be easy for a datacenter or a 1000-cubicles office to buy peace of mind by using the “blessed” distrobutions instead of let’s say Debian or Slackware (to name two randomly choosen examples). But in embedded space, there is a lot more fragementation and a completly different landscape of providers, with a lot of smaller hardware shops doing heavy customisations of kernel, toolchain and userland themselves in order to accomodate different useage scenarios. At least, that’s pretty much the situation in those areas of embedded development where I work, where seemingly every small vendor has it’s own out-of-tree kernel modules and patches for BusyBox and uClibc. It is because of this customability that Linux (and F/OSS in general, as I agree that they could have used the vfat/fat16 patents more or less against every other player in the industry regardless of the OS) has seen such an uptake in embedded space and why for example WinCE is not the logical choice for a wide range of products.
Lawsuits take time to prepare and typically don’t happen overnight (at least not lawsuits on this scale, if experience is anything to go by). But the little paranoid tinfoil hat wearer in me finds it interesting that this coincides with the recent increased coverage of ARM and MIPS based devices in the main-tech press (e.g. the Marvell plug computer, ARM and MIPS based netbooks, etc.). It may be an attempt to send the message to the industry “even if you are not using x86, you *really* should be using our (=WinCE et al) platform to develop your products, you know”.
OTOH, it may be an attemt to test the waters for a future buisness model that uses “monetizing of IP” as a substantial source of income.
(Disclaimer: I’ve stopped reading the article(s) prior to point specific patents were mentioned. This may be too little to prevent getting “tainted”, but it may make the difference between willful infringement and mere casual infringement at some unlucky point in the future. Thanks in advance for considering this should you choose to reply to my comment.)
I agree. Maybe MS wouldn’t try to sue massively today -it wouldn’t be in its best interests-, but what about 10 years or 20 years from now? In a different environment, MS might decide to sue everyone over patents and take them down with it, like a collapsing star…
Edited 2009-02-26 12:09 UTC
I agree, especially when MS fires many employees, and needs to present good financial results. They may think it’s time to get their return on investment out of all these pattents (cash cows). Sueing all companies that sell devices that use MS patents would yield a fortune for MS. Imagine all MP3 players, all USB flash keys, mobile phones, memory sticks, digital video recorders, digital cameras, etc… They all use FAT32. Imagine all other patents that we don’t think about. MS has hundreds of thousands of patents in their sleeves. There are billions of electronic devices around, not to mention software patents and commercial applications sold every minute…This represents a lot of money Microsoft could suck in in a period of economic downturn.
Edited 2009-02-26 13:57 UTC
This very much depends on what you’re trying to do. In the end, the whole storage of the device is exposed through USB mass storage, so you’re free to change anything in the filesystem. In general the software is even able to cope with most users doing less intelligent things that way.
However, I agree that not having a Linux version of their TomTom HOME software is a PITA, but it is not because of portability of the code, as they do have a proper MacOS version out.
I guess the complication on the Linux side is the diversity, and the extra investment in their support teams for a diverse platform like Linux….
Hm, in retrospect PITA is probably a label that is a little bit too strong verbality wise, because the mass storage support nowadays seem to work more or less as expected, so thanks for reminding me of that. (It is funny that negative impressions from several years back stick better in memory than recent positive experiences, hm, talking about selective memory )
Wine support for HOME was always a little bit hit and miss in my experience (with a larger fraction of hits for HOME 1.x at the end of the wine 0.9.x cycle), I guess I’m just a little bit dissapointed that a company which bases its product around on alternative OS like Linux (and can if I’m not mistaken be considered a good member of the community with active contributions to upstream, like for example the ARM branch of the linux kernel) has a customer support that do not even know what “linux” or “free bsd” is and seemingly has no plans to at least contribute to the wine project in order to get a cross – platform program (a la EVE online or Google picasa) together.
The implications for linux are indirect, but there are broader implications for commercial OSS implementations in general.
The actual patents in question cover the garblygook for using long filenames and maintaining compatibility with the archaic 8.3 filenames.
In TomTom’s case, and most other linux-derived products, the functionality is provided via GPL related components (ie. the kernel, userland utilities etc.).
MS will happily license the related patents for FAT32 to any vendor (and many vendors do license them), and the fee is relatively cheap, *but* the GPL wouldn’t permit distribution of patent-licensed code unless the patent rights we transferable downstream. So TomTom can’t license the related patents for their products if they want to use the GPL toolchain they’re using to implement the functionality.
The alternative is to create a proprietary solution that is non-transferable, which is somewhat counter-intuitive for a strategy of leveraging OSS for product development.
It’s not a shot at linux directly, but it is a warning shot at the commercial OSS ecosphere. And that could have indirect implications for OSS in general.
Too soon to tell if the sky is really falling, but when you couple this action with Ballmer’s recent statement that linux is the largest competitive threat to Windows, it looks like he’s lifting the lid on Pandora’s Box to take a peek at what’s inside.
Microsoft ramps up its effort to monopolize the patent trolling business…
Yes, this is a PITA but… Does this have real effect on other countries outside the “US and A”?
This is a complete nonsense… Did anyone patent “patent trolling”? It should be fun…
Software Patents are stupid…
It will when the EU forces MS to start including alternative filesystems as an option in Windows…
Not really, but if it goes ahead, it will cripple the US for software production. they will be left behind by other countries and their systems.
Open source has thought for too long that IP infringment doesn’t effect them.
Maybe it’s time Open Source started playing by the same rules as everyone else, or better still abolish the crazy software patent system for everyone.
Instead of making the veiled threats it has in the past, Microsoft should go to the Linux Foundation with their concerns if they’re serious about these patents they’re claiming.
GedMurphy wrote:
–“Open source has thought for too long that IP infringment doesn’t effect them. Maybe it’s time Open Source started playing by the same rules as everyone else”
Last time I checked Reactos had FAT32 support (iirc it can only be installed on FAT32). Since you present yourself as a Reactos developer, shouldn’t you be working on getting it removed?
ReactOS isn’t sold at the moment by any companies, so there’s no requirment for a licence.
To answer the second part of your question, there will be many alternative file systems along in the foreseable future.
gedmurphy wrote:
–“ReactOS isn’t sold at the moment by any companies, so there’s no requirment for a licence.”
So you are allowed to distribute patented work as long as you do not sell it? IANAL but that sounds like you’re just being hypocritical. Unlike Linux, Reactos currently RELIES on Fat32 and the future use of ‘alternative’ (I guess it was hard for you to say ‘open’) file systems will likely do little to alleviate this due to Reactos being a stand-in for Windows and thus need Fat32 compability.
I’m also not a laywer but my understanding of this is that anyone selling a product which uses FAT must buy a licence for each unit sold from Microsoft. Something in the region of $0.29.
This is what TomTom are failing to do in this particular case.
One thing you can be sure of is that ReactOS will do everything in its power to act within the law and follow patent and IP guidelines, however stupid they may be.
If ReactOS ever becomes mainstream we are well aware that we will need the assistance of laywers to guide the project legally. Developers are notoriously bad at pretty much everything other than writing code
Ok, I’m another developer who – from time to time – does some armchair-lawer-posing, and on top of all that I’m not even very familiar with ReactOs, but :
According to your projects homepage, ReactOs is licensed under the terms of the GPL (presumely version 2?), at least according to the information available from
http://www.reactos.org/de/license.html
If your FAT/FAT32 drivers are licensed under the GPL v2 and/or the whole compound of kernel + fs drivers falls under this license, all potential recipients of the software, regardless of their “commercial” status, regardless of the point in time and supply chain they may be in when they use the source code in question, regardless on the nature of their useage, will have to have a valid patent license, or the code has to be removed. Inherent with the nature of the GPL, this patent license must therefore be of RAND-Z nature.
Therefore, given my assumptions are correct, ReactOs is exactly in the same situation as the linux kernel project: As long as nobody taps on your shoulder and accuses you of infringing a patent, it is – at least for individual F/OSS coders – considered to be good practice to avoid looking around for potential infringing patents, as you may get tainted. (Don’t take this as an legal advice, as I’m in no way qualified to give such, this is just what the legal department of my employer has told me when I asked about this issue some years ago).
Once a patent is discovered to potentially interfere with your GPL and/or LGPL licensed code, you *must* act, otherwise your are not entitled to distribute the code (under the terms of the GPL at least) at all (see the preamble of the GPLv2 and section 7 therein, GPLv3 the GPLv3 is famous for even stricter language wrt these matters). These actions may include trying to invalidate the patents in court, trying to negotiate a compatible patent license for all possible up- and downstreams or trying to work around the patent in question.
It wil be interesting to watch how this plays out for different linux vendors. Microsofts F/OSS “embassy” ( 451 group) has already started to downplay the implications of this on the whole linux ecosystem, e.g.
http://lwn.net/Articles/321105/
and Microsofts ability to collect damages may be severly limited by the Bilski decision, and the overturned decision wrt FAT32 patent from 2006 does not prevent others from trying to invalidate the patents in question with different arguments or examples of prior art, etc. . These seem to be “interesting times” for F/OSS developers and hobby lawers alike.
For now, let me reiterate this: AFAIK, there is no possiblity to “fly under the radar” or to negotiate a seperate peace treaty with a patent holder, once push comes to shove. You may be able to negotiate a patent license for your driver (e.g. for your piece of GPL code) which does not stretch to the relevant linux driver (e.g. a different patch of code licensed under the GPL), though. But in the end of the day, all GPL covered implementations of file system drivers dealing with FAT/FAT32 sit in the same boat.
The software patent system is a mess. I hope at least this is something we can all agree upon. Acting like this patent does not affect you just because your application has not gone “mainstream” (under the terms of the GPL you are mainstream the day you start to distribute the software) is not a very sensible position, in my humble opinion.
As always: I would be grateful for any corrections to may layman’s analysis of the situation.
Edited 2009-03-01 09:17 UTC
You’re missing the point. We have literally dozens of alternate filesystems now, most of which are so much better than FAT it’s not even funny. And yet, what is the universal filesystem? The one that every operating system can understand, read, and write without causing any oddities? It’s fat. It is the only fs that, no matter what os you use, you will certainly be able to utilize your media.
With all the alternative filesystems out now, which are much better (even ext2 trumps FAT for crying out loud), and further many of these filesystems are under a completely free license, how come none of them have become universal? I don’t, honestly, get it, unless this is another situation caused by the ubiquity of Windows. I don’t think so, however. Look at the open source oses, we don’t even have a universal filesystem there that can be equally used across all foss platforms, or even across all foss UNIX-like systems… except one, and guess what it is?
Sad thing is, with ntfs-3g and fuse being ported to quite a few different platforms out there, ntfs might end up the universal fs next. This would be equally bad, even though ntfs is a much better filesystem than FAT, we’d just have a repeat of this situation again.
I don’t see how the number of alternate filesystems in the future is going to help. We have dozens of filesystems now and don’t have one that is under a foss license, or at least an open standard, that we use even now. Having more alternatives is not a fix for this–we don’t lack for filesystems, we just lack one that is created equal everywhere. That is what we need.
IP infringement and Software Patents are two different things.
IP infringement is take seriously by producers of all software, open source or not, but in most countries software patents are not worth the paper they are written on.
First there was PNG, then came the OGG and now …
FAT is the new GIF! Let’s take the opportunity to deprecate FAT and create a better faster free replacement. Yes, it will take some effort but that’s what we need to preserve freedom!
There already exists several better, faster, free replacements. That’s the easy part. The problem getting people to use them.
The thing about FAT is that everybody and everything can read and write FAT. Sure I could format my USB memory with a different filesystem, and it may even perform better, but I wouldn’t be able to use that USB memory on all the computers I may come across. I can plug my USB stick, mp3 player or digital camera into basically any computer and that computer will be able to read it. Until a new file system has this sort of penetration FAT is here to stay.
And that’s why the GP is right to say that FAT is the new GIF: in both case, the developer allowed free access for a long time to ensure widespread usage before using patents to get money..
PNG didn’t manage to displace GIF, despite being superior from a technical point of view, so it’s quite likely that FAT won’t be displaced either I agree.
The only slight hope is that Microsoft could loose their patents if TomTom choose to fight..
It didn’t? I’m pretty sure JPEG and GIF has combined almost wiped out GIF. Who’s using GIF these days?
You’re kidding, right?
I saved the HTML source of an OSAlert page:
$ fgrep -i .gif 21043.htm | wc -l
11
$ fgrep -i .png 21043.htm | wc -l
3
[ jpg’s being compressed don’t count. ]
In the beginning of 2008 TomTom traded at 70 euro per stock on the Dutch stockmarket. Now they are worth 3 euro per stock.
Microsoft is hitting hard on a company that is already in the drain. How can TomTom afford to defend themselves after a disaster year like 2008 was for them? Microsoft has got money for sure, and with this economy everyone will be scared to step up against Microsoft. They are using the economic climate to bully the entire industry (remember TomTom is just one company, they used blackmail to extract license fee’s from many companies before this).
The “computer in the dashboard” patent of Microsoft’s is ridiculous, but unfortunately we probably won’t actually find any prior art to invalidate it.
The patent calls for the computer “running vehicle-related programs (such as navigation) and non vehicle-related programs (such as music playback)”, and while there were many car PCs before 1999, I’m not sure if any of them ran “vehicle-related programs”.
However, I don’t see Tomtom being really affected by this patent, as it also calls for wireless internet access. AFAIK no Tomtom devices can do this.
The FAT patent is a real concern. I’d venture a guess that most consumer electronics use FAT in some way or another; just look at MP3 players or cameras for instance! If this one stands up in court, Microsoft could sue most of the CE world.
That was argued and won, and then lost in appeals in 2006. MS does have the ability to sue anyone who doesn’t license FAT from them.
Incorrect. They cannot sue anyone outside of the US and I think Australia as they were dopey enough to support software patents.
People in other countries can safely ignore this crap.
The “computer in the Dashboard may have been deployed in the early 90’s as part of the TravTek field operational test in Orlando, Florida. See, for example, http://ieeexplore.ieee.org/Xplore/login.jsp?url=/iel2/4469/12662/00…
Do a search for TravTek and there are many online references, as it was a cooperative effort between government and private industry.
Wireless access is how TomTom and most GPS systems work. They get updated information from the satellites, which also happen to be on the internet. The latest maps and such are then downloaded from the servers they reside on and update the device. TomTom devices tend to update often, which is why they work so well.
My understanding is that GPS units download new data from the PC, when you connect them to your PC. Urgent updates to maps might be pushed out via the PC, but usually you have to purchase new maps each year and they don’t cost much less than a basic GPS unit anyway.
Up-to-the-minute traffic information, on models that support it, is sent “underneath” a regular FM radio station’s signal, much like teletext is sent along with a TV station’s signal. There is no two-way communication, only one-way.
The units themselves find their location by use of the satellites, but AFAIK they do not download maps or extra data from them. Unless GPS units in America are much more advanced than what we get in Australia.
I think this counts as “vehicle related programs”
“””
Most major technologies required for modern automobile navigation were already established when the microprocessor emerged in the 1970s to support their integration and enhancement by computer software. These technologies subsequently underwent extensive refinement, and a variety of system architectures had been explored by the time practical systems reached the market in the late 1980s. Among the other enhancements of the 1980s was the development of color displays for digital maps and of CD-ROMs for digital map storage.[1]
“””
http://en.wikipedia.org/wiki/Automotive_navigation_system
Back in 2k6, there was a big stink when MS won the appeal to patent FAT32. The situation for linux regarding has not changed at all since then. What I do not understand is why FAT32 is still in the kernel trunk, since it has been under patent for 3 years now, and by definition open source needs to be patent free.
(insert IANAL and all the other usual disclaimers here).
As far as I know, the FAT32 patent covers only the “long filename” features of FAT32. So technically, there is no need to remove the whole FAT32 support from the mainline kernel, as long as the long filename support is either done differently (e.g. there may be ways to provide the same structure on the mass storage device with a different mechanism) or completely left out.
Side remark:
This is incorrect, at least according to my knowledge about the FSF definition of Free Software and OSI’s definition of Open Source (and if I’m right, then I can only guess that this is an acciential misrepresentation on behalf of you, probably caused by the compactness of your comment, because so far I always thought you know better than that).
Both definitions rely soley on copyright law and don’t even mention software patents. This may be in part to their philosophy (e.g. copyright is deemed to be protection enough for software), because of historic context (copyleft predates the presumed validity of software patents, at least in the US, IIRC) or because of a desire to keep the definition generic.
Of course, licenses like the GPL mention software patents and it is probably a safe bet to say that the majority of participants in the F/OSS ecosystem would like to get rid of the whole idea of software patents rather sooner than later. But as long as the patent is licensed under RAND-Z terms for all participaints, there is afaik no problem with either the GPL or the LGPL. (Licenses that do not even mention patents have no problems with software patents pretty much by definition).
Edited 2009-02-26 17:29 UTC
Very simple:
Software developers are much better off, not reading patents. As infringement of software idea patents is unavoidable because most of them are overly broad and obvious, it is better to ignore patents altogether, and in case you get sued you can only be held liable for unintentional infringement, instead of willful infringement.
I cannot find TomTom’s source code on its web site.
Maybe it’s not an open source software.
All opensource software _used_ in the TomTom products (the kernel code, userland tools, toolchains, etc) can be found @ http://www.tomtom.com/gpl . However, indeed, the main software on these devices (the TomTom Navigator software) is non-free.