The Real Reason For Microsoft's TomTom Lawsuit

Glyn Moody writes "We now know that Microsoft's lawsuit isn't just against TomTom, but against Linux too: but what exactly is Microsoft hoping to achieve? Samba's Jeremy Allison has a fascinating theory: 'What people are missing about this is the either/or choice that Microsoft is giving Tom Tom. It isn't a case of cross-license and everything is ok. If Tom Tom or any other company cross licenses patents then by section 7 of GPLv2 (for the Linux kernel) they lose the rights to redistribute the kernel *at all*. Make no mistake, this is intended to force Tom Tom to violate the GPL, or change to Microsoft embedded software.' Maybe embedded Linux is starting to get too popular."

Comments

Say It Ain't So

[alain94040]

Ok, I'll play devil's advocate for a second. Here are the relevant parts of section 7 of the GPLv2:

If, for any reason, conditions are imposed on you that contradict the conditions of this License, they do not excuse you from the conditions of this License. If you cannot distribute so as to satisfy simultaneously your obligations under this License and any other pertinent obligations, then as a consequence you may not distribute the Program at all.

Here's an example. The US government probably prevents you from selling your open source software to Cuba or Iran. If I read section 7 correctly, that counts as a "condition imposed on you". So really you lose all rights to using that code?

You got to be careful with literal interpretation of legalese... sometimes you can push the arguments too far.

I hope the same applies to this theory that Microsoft is forcing people to violate the GPL and therefore lose their rights to the code.

Re:Say It Ain't So

[FrYGuY101]

So really you lose all rights to using that code?

You lose all rights to DISTRIBUTE that code. You can still use that code in perpetuity, though.

Re:Say It Ain't So

[Hognoxious]

So really you lose all rights to using that code?

Only in so much as using means the same as distributing, or more specifically distributing to those countries.

So in other words, no.

Re:Say It Ain't So

[Anonymous Coward]

Since they have to distribute the code so people can use their devices they could just switch to a free FS. As I understand (IANAL) this lawsuit mostly concerns their use of FAT for their memory cards. If they used EXT2 and bundled EXT2IFS with their windows app they might well be able to avoid a lot of the hassles from Microsoft regarding this.

Re:Say It Ain't So

[dgatwood]

And would make updating the thing nearly impossible. They do firmware updates by mounting it as a USB mass storage device. Without a hardware redesign to emulate FAT (which would probably also violate M$'s patents), they're pretty much stuck here.

This is why I've been arguing for nearly a decade that file and volume formats should not be patentable, nor the means used to read and write those formats. Free and open access to data formats is fundamentally crucial to the interoperability of all hardware and software, and as such, statutes should very clearly define those as part of a class explicitly excluded from patent protection. As soon as the courts allow even one patent like this to stand, they are pretty much saying "f**k you" to the entire computing industry and depriving consumers of their fundamental right to have access to data of their own creation. That data isn't Microsoft's. It belongs to the users, and it is a violation of the most fundamental rights of the users to deny them access to content that they create merely because they choose to not use a particular software product, regardless of whether that product is made by Microsoft, AutoCAD, or anybody else.

Locking down user content is fundamentally wrong and unjust, and any laws that allow a company to do so are also fundamentally wrong and unjust. Therefore, it is our right, nay, our duty to users everywhere to violate those bits of intellectual property at every possible opportunity until it becomes such a legal nightmare for these companies that they are forced to back down. Anything less would be uncivilized. I know this is no Rosa Parks moment, but it still very much necessary for the long-term viability of computing as we know it. Just say no to data format patents.

Re:Say It Ain't So

[morgan_greywolf]

Forget ext2, they can use UDF, which is already supported on Windows, Linux and Mac OS X and is not patent encumbered.

Re:Say It Ain't So

[Rob Riggs]

I don't know why companies just don't get *BSD working for them instead Linux. It would save them a lot of headaches.

Because the primary reason for the success of Linux is that it forces everyone to share their improvements. You get an exponential return on investment. The best you can ever hope for with BSD is an incremental return.

Re:Say It Ain't So

[rbanffy]

You don't need to emulate the FAT long file names in order to do that. Only the long file name hack is covered by their patents. I see no good reason to even use long file names in applications like GPS or cameras, since you don't see the files most of the time.

Re:Say It Ain't So

[Jimmy_B]

Here's an example. The US government probably prevents you from selling your open source software to Cuba or Iran. If I read section 7 correctly, that counts as a "condition imposed on you". So really you lose all rights to using that code?

You have misread this section. Having a condition imposed on you which prevents you from to distributing to a specific party does not prevent you from fulfilling the conditions of the license, because the license does not obligate you to distribute the program to anyone; rather, the GPL gives the conditions you must follow when you do distribute the program. Since US export restrictions do not prevent you from fulfilling the terms of the GPL when you export to a non-restricted country, the fact that there are parties which you can't distribute the program to is irrelevant.

Note, however, that only a government can enforce export restrictions; the GPL forbids you from taking on that responsibility yourself. So if you send a GPL'ed program to someone in Europe, they could legally send that program to someone in Cuba, and the GPL would forbid you from interfering. If the US were to pass a law which said that you couldn't export something that could possibly be re-exported to a sanctioned country, then that would be a problem for the GPL, but to my knowledge no such law exists.

Disclaimer: I am not a lawyer and this is not legal advice.

Re:Say It Ain't So

[morgan_greywolf]

UDF is an ISO/IEC standard, so the format itself is not patent encumbered. Yes, ways of implementing it are, but the important thing is that Microsoft only has 3 patents that even mention UDF, and only one of those is specific to UDF. Also, IBM seems to be one of the largest patent holders on UDF implementations. I'm guessing they'd willing to launch a patent salvo against Microsoft should Microsoft try to sue someone over Linux' UDF implementation.

Re:Say It Ain't So

[dgatwood]

The problem is that these extensions aren't just used for long file names. IIRC, extra directory records in the FAT filesystem are also overloaded for other purposes like permissions, without which Linux et al would be unbootable off of FAT. I'd imagine that many of those uses would run afoul of the patent, but I could be wrong.

More to the point, if it only applies to its use for naming purposes and not to the concept of storing additional data about a file in additional directory entries with reserved type codes that older OSes ignore, then the invention should have been unpatentable anyway, as there's nothing particularly original about taking the Rock Ridge extension set from ISO-9660 and applying the exact same concept to FAT except insofar as it uses additional directory entries. That's literally all they did here. Instead of an additional entry in the system use area of the variable-length directory record, they use additional directory entries with a different type code, but that's basically caused by differences in how the filesystem describes a file....

Re:Say It Ain't So

[Anonymous Coward]

I don't know why companies just don't get *BSD working for them instead Linux. It would save them a lot of headaches.

Maybe because companies like Microsoft have a history of stealing BSD code, making minor changes, and then patenting their implementation. This is why Ted Tso' said he would use the GPL for Kerberos instead of the BSD license if he had it to do over again.

Re:Say It Ain't So

[JerkBoB]

Right, but then everybody gets your improvements. Where's the competition in that?

With BSD you get a solid base for your product and it's not infected with the GPL.

This old clinker again... Closed-source competition is so 1980s.

Software is a commodity. Any intelligent company now is not trying to make money on their proprietary code. They're making money on hardware to run their code, or services to support their code, or data to feed their code.

Sure, for certain niche markets, closed source can give a company a competitive advantage for a while... But if the market is "hot" enough, Open Source will eventually be there to eat its lunch. This has been happening over and over again for the past two decades. Were you asleep?

Let me ask you this: Have you ever written code and released it to the public? Was it used? I have. As a developer making contributions to public projects, I am much more inclined to contribute under the GPL than other licenses. Most of the world feels the same, hence the popularity of the GPL (and similar "viral" licenses) over the BSD-style licenses.

Re:Say It Ain't So

[Anachragnome]

"Therefore, it is our right, nay, our duty to users everywhere to violate those bits of intellectual property at every possible opportunity until it becomes such a legal nightmare for these companies that they are forced to back down. Anything less would be uncivilized. I know this is no Rosa Parks moment, but it still very much necessary for the long-term viability of computing as we know it. Just say no to data format patents."

This is precisely the tactic I encourage everyone I know to use.

These people are no longer playing fair, WHY SHOULD WE?

In this day and age, corporations are, quite simply put, walking right over common sense. There is no more "customer service", but rather corporations simply see us all as resources to be mined.

When these people no longer see reason, no longer work to provide a product without stifling the competition, then "Intellectual Disobedience" is the ONLY route left to address the situation.

Speak with your mind, voice and dollar, in that order.

Re:Say It Ain't So

[Anonymous Coward]

I don't know why companies just don't get *BSD working for them instead Linux. It would save them a lot of headaches.

Because the primary reason for the success of Linux is that it forces everyone to share their improvements. You get an exponential return on investment. The best you can ever hope for with BSD is an incremental return.

The primary reason for success of Linux is a combination of timing and luck. When the BSD project's viability was in serious question due to the AT&T lawsuit, Linux hit its stride.

Moreover, your response demonstrates a common fallacy: the idea that, if not forced to share improvements, companies will not do so.

The reality -- as usual -- is much more complex:

• The cost-benefit of analysis of failing to share re-usable improvements to BSD licensed code is fairly simple -- it costs quite a lot to maintain sizable private changes. Merging your code into the primary code base decreases maintenance costs and pushes the effort to the larger development community
• Many corporations will not invest in GPL'd software if it can be avoided, as they often won't be able to use it. How many contributions does GPL software *lose*?

I would argue that Linux has succeeded *in spite of* the GPL. Speaking with those in the embedded hardware industry, it's clear that Linux has a foothold because of its brand name, not because of significant technical differentiation. Many embedded hardware vendors -- even the major ones -- don't even understand the implications of the GPL! (see the Linksys GPL violations)

If you observe the BSD projects, you'll note that features, bug fixes, and developer time is often provided by corporations that leverage the BSD licensed code in otherwise closed source products.

To quote Juniper's VP of Foundation Technologies, Naren Prabhu:

Juniper benefits from the powerful collaboration between leading universities, individuals, and commercial organizations developing FreeBSD to advance the operating system functionality. The FreeBSD release system provides Juniper with a roadmap for features and a stable base for our code, while its practical licensing enables Juniper to develop intellectual property for advancing high-performance networking. Juniper employs many active FreeBSD developers that continually contribute to the FreeBSD project to further its development as a leading operating system.

I wouldn't be so quick to point to the GPL as a recipe for success. When you force individuals to choose between your way or the highway, many of them will choose the highway. The BSD project's more pragmatic stance has allowed corporations to contribute as they are able to do so, and has resulted in an extremely productive, ongoing relationship with commercial software vendors.

Re:Say It Ain't So

[Nimey]

Wrong. FAT does not support permissions of any kind. VFAT adds long file names, which is done by using half of a given directory's entries for long filenames and the remainder as the standard 8.3 short names.

If you were talking about FAT attributes (archive, read-only, hidden, system, volume label), those are done with flag bits.

Re:Say It Ain't So

[Man On Pink Corner]

UDF is an ISO/IEC standard, so the format itself is not patent encumbered.

Um, that doesn't mean it's not patent encumbered. MP3 is patent-encumbered all to hell, and it's an ISO/IEC standard.

Re:Say It Ain't So

[arkhan_jg]

Why not got one further, and simply ban software patents of all kinds, like most other countries in the world?

Software is covered by copyright; there's no need to grant patents on it too. Filesystems, like many other software concepts are relatively easy to come up with, on a broad basis - and the same general patent covers many, many approaches and denies them to competitive forces for way too long. The clever, hard work is in actually implementing the idea and making it work.

Patents were originally intended so that someone could come along, read the patent, and easily replicate the entire product, once the monopoly period was over, and the inventor compensated for putting his work into the public domain. The equivalent for software, direct instructions on how it works - is the code and documentation itself, not some vague patent intentionally designed to cover as many possible variations upon the idea as possible.

Software patents are solely about preventing competition and raising the bar to entry for smaller, innovative firms. When engineers are specifically instructed to never, ever read patents in case they get sued for subsequently implementing anything that vaguely touches on a similar area; when this happens, patents are actively harmful to innovation and endeavour. Get rid of them, as pretty much every other country in the world has done.

Re:Say It Ain't So

[Patch86]

The GPL and BSD licenses fulfil completely different purposes, and attract different contributors because of it.

You can't declare either license the "winner" without first deciding what it is they're competing in. If you want the biggest, most widely used cross-platform free software, GPL wins- Linux's desktop share beats any of the *BSDs, and it has massive penetration in embedded devices, webservers and cross-platform applications. If you want the biggest, most widely used platform full stop, BSD wins- MacOSX has a colossally larger desktop share than Linux (but is of course mostly non-free), and BSD code can be found in almost any project you care to mention.

For people and big companies who want their free contributions to remain free, the GPL is far more attractive (just ask all those Linux vendors). For people who just want to see technology get out there and be used, BSD is the drawer.

The 2 camps can't both be pleased at the same time, and that's why the 2 licenses can exist side-by-side. Acting like the GPL is detrimental to Linux (or the same with BSD) is just plain ignoring the reality of the situation.

Re:Say It Ain't So

[idlemachine]

These people are no longer playing fair, WHY SHOULD WE?

"He who fights with monsters should look to it that he himself does not become a monster."

Or how about: "Two wrongs don't make a right."

Because how you win is as important, if not more so, than winning itself. I agree that there are times that we need to fight but we also need to be sure we don't lose ourselves in the battle.

Really an attack on using Microsoft tech in Linux

[guruevi]

What TomTom (and others) need to do is start using EXT2/3 on their external cards and then distribute Fuse with their software. This will force FAT and with that Microsoft tech slowly but surely out of the market.

What do you think will happen when all external media starts using alternative formatting?

Re:Or Maybe...

[mlwmohawk]

maybe they are just trying to protect their intellectual property

I'm not saying that "intellectual property" is a pointless concept, but what is currently implemented is frighteningly Philip K. Dick.

Software is particularly problematic. An invention does not always come from the intellect and work of the inventor. More often than not it is merely an observation and augmentation of the work and intellect of others.

Software is nothing more than building on that which was built by others, which was built on the work of people before that, and before that, ad infinitem. Even the implementor of a statistical analysis system owes credit to the creators of the programming language used to write it, the creators of the math system, etc.

Intellectual property my ass, it is a land-grab of an environment created by two generations of engineers that worked and published without patent protection. Now college drop-out Bill Gates, sues for trivial implementations of theoretical models created by men far better than him.

Re:Misleading title.

[ichthus]

As a Samba developer, I would think he's especially familiar with Microsoft's strategies.

Re:Patents, not code

[shentino]

Actually, the GPL forbids restricting other people's freedom.

It is, and should be, a GPL violation to secretly get a patent license. If TomTom were to cave to MS without getting slapped with a GPL violation, then anyone who uses tomtom's work would be opening themselves up to patent infringement suits.

Please RTFM and actually read the GPL. A good grep would be "they do not excuse you". Search the GPL for that text and you'll zero in on what I think is a very critical "failsafe" in the GPL.

Re:Classic GPL

[iplayfast]

You seem to have a misunderstanding of what the GPL is. It is a method of payment. TomTom using GPL'd software must abide by the payment method. That is, by providing the source for use or change by the people who they are selling to. Instead of dealing with dollars they deal with IP.

The problem at hand is that MS feels that TomTom should be paying for their patents, which would violate the GPL payment method.

You should be marked troll as you are obviously trolling for this response.

Linux Action Show says no

[pembo13]

The guys over at the Linux Action show (in their last episode) seem to adamantly think that this lawsuit has nothing to do with Linux. Jeremy Allison is probably a pretty jaded individual at this point (and rightly so), so having the view of someone else more familiar with these legal quagmires may be helpful.

Good work Stallman...

[cyphercell]

"There is no avoiding war; it can only be postponed to the advantage of others." Niccolo Machiavelli The escalation is not all Microsoft's. GPLv2 is proof that the FSF was looking forward to this scenario.

fascinating explanation

[viralMeme]

'then points to another blog that promotes a "fascinating theory"'

Actually that's .. points to a fascinating explanation by Jeremy Allison of the Samba project...

Excuse Me?

[ratboy666]

Gates (and Allen) developed MBASIC, and DISK BASIC. DISK BASIC used the "FAT" system to control free space.

CP/M did NOT use the same scheme. Instead, CP/M built up free space maps by scanning the directory. It also did not use a linked list. Personally, I thought FAT was weak then, and still is....

But the "industry" adopted it. It was allowed; we had a (at least) minimal common system for file systems. Enhanced to support directories and sub-directories.

Then, Microsoft designed a long filename system on top of it, that was back-compatible with the old method. THAT was patented. And, no, it wasn't even the "obvious" solution -- that would have been a mapping file.

What does this mean? It means that the long filename code SHOULD be ripped out. 8.3, baby! You want longname mapping? Linux has UMSDOS on top of FAT -- same result, no patent violation. Or, just use short names. Or, build a program that reindexes MS FAT longnames into UMSDOS (for read compatibility). Just don't write that format. It can be argued (I would try) that ANY longnames in MS FAT format that were found on a FAT filesystem then MUST have come from an MS patent licensee (because our proposed system wouldn't generate the MS FAT longname format). So, there are solutions. Maybe UMSDOS is too "crufty" to be resurrected, but it strikes me that something like posixovl.fuse could be used (with modifications).

Microsoft was creative with the MS FAT longname solution. Either deal with it, or get the patent overturned.

Re:Excuse Me?

[mlwmohawk]

Gates (and Allen) developed MBASIC, and DISK BASIC. DISK BASIC used the "FAT" system to control free space.

Gates and company copied basic from other sources.

The FAT system is nothing more than a fixed size array allocation system, in use in many systems of the time.

CP/M did NOT use the same scheme. Instead, CP/M built up free space maps by scanning the directory. It also did not use a linked list. Personally, I thought FAT was weak then, and still is....

CP/M was better, yes.

But the "industry" adopted it.

One has to wonder about the anti-trust issues of patent usage. Yes the industry adopted it, but could it have, in any practical sense, adopted anything else?

Then, Microsoft designed a long filename system on top of it, that was back-compatible with the old method. THAT was patented. And, no, it wasn't even the "obvious" solution -- that would have been a mapping file.

The word "obvious" is subjective. LFN in FAT is implemented simply using the existing directory mechanisms. Is the only way to do something "non obvious?"

Microsoft was creative with the MS FAT longname solution. Either deal with it, or get the patent overturned.

"Creative" is not what makes something worth a patent. Was it obvious? The only answer to that is yes, as someone skilled in the art (someone with file system experience, particularly FAT) their method was the only way to do it.

Re:Really an attack on using Microsoft tech in Lin

[Anonymous Coward]

What TomTom (and others) need to do is start using EXT2/3 on their external cards and then distribute Fuse with their software.

No, the easy way out here is to not use long filenames. The patents is not about FAT, but about long filenames on FAT. If they need long filenames, distribute it as an tar file or loop back-file system file with a short name.

Re:Misleading title.

[geekboy642]

one of the worst configuration file formats ever

Wrong.
I hold up for example: XML, the Windows Registry, and sendmail.cf

Re:Misleading title.

[Jeremy Allison - Sam]

Thanks dude. I guess I should take my complements where I can :-).

We are moving to a registry based config in later versions, but I'm not sure you would think that an improvement :-).

You have to remember Samba is 17 years old, and you can still parse original smb.conf files from the first version. These things do tend to acrete over time, and it's hard to break existing configs. Not an excuse, but...... :-).

Jeremy.

Re:Misleading title.

[FictionPimp]

Just get me release quality AD support via 2003 server and we will call it even.

Collateral Estoppel / Issue Preclusion and YOU!

[MarkvW]

If Microsoft wins, it sucks for Tom Tom and it creates FUD. That's bad, but not too bad. Microsoft still has to sue everybody violating its software patent.

But if Microsoft loses because the Court rejects the concept of software patents a'la Bilski, then Microsoft is royally screwed because if it sues anybody else over a software patent, that defendant can argue that Microsoft can't argue software patents anymore because they already fully fairly and finally litigated the issue against Tom Tom and they lost. They don't get to relitigate the same issue all over again.

You can see why this is HUGE for Microsoft. If they win, they get some money from Tom Tom and they put a scare into the Linux community. If they lose because their software patents are no good, then Microsoft's whole software patent edifice is gravely jeopardized. Microsoft will really fight this hard.

Tom Tom is really vulnerable because the GPS market is slammed in this economy. I suspect that Microsoft is betting that they'll give up. The Linux community ought to prop up Tom Tom with legal and technical support--at least on the software patent theory.

Microsoft's invasion should be defended at the beaches. They should be thrown back into the sea before they create more FUD!

 

Re:Misleading title.

[Sponge Bath]

sendmail.cf

Gah! Noooooo!
You have invoked the name of ultimate evil, dooming us all!

You should edit sendmail.mc (evil light) instead and process it with m4.
Nothing could be more intuitive.

Re:Classic GPL

[Jeremy Allison - Sam]

This :

"The problem here is that the GPL sets up restrictions which no intermediate vendor can realistically comply with" should get you marked down immediately as -1 Troll".

Or at least as -1 clueless. Do you know how many intermediate vendors ship GPL code, both v2 and v3 ? It's a *lot*. You can even get patent cross licenses for all your other code so long as the patents you are licensing don't cover the GPLed code. Please post your ignorant long diatribes elsewhere.

Jeremy.

Those that cannot compete, litigate.

[QuietLagoon]

Wasn't that what Microsoft said about the Windows anti-trust lawsuits?

Remember GIF

[fritsd]

I'm wondering if a patent can become so diluted that it is unenforceable. FAT is ubiquitous. It is used just about everywhere in every industry in innumerable devices. If Microsoft were to fight to enforce their patent they'd essentially be taking on the entire IT space. I doubt the courts would allow that to happen.

IIRC, that's exactly what happened to GIF: After it was well established and used, Unisys decided to see dollar signs.

Re:I disagree!

[Lord Bitman]

Software is covered by patents, there is no need for it to be covered by copyright, too.

Software is often distributed in binary form: a form which cannot be derived from. The protection, for a limited time, of original works, is meant to allow them to be developed so that people in the future can create derivative works based on them.

  - NO protection without source code.
  - NO copyright on compiled software (makes as much sense as copyrighting a hammer)
  - Patent protections on binaries, contingent on the full source being provided.
  - NO obvious patents.

Software patents aren't bad, they just have a bad name because stupid ones have been granted.