|
|
Subscribe / Log in / New account

A new VFAT patent avoidance patch

Andrew Tridgell has posted a new patch intended to enable the kernel to work around the VFAT patents. Unlike the previous version (covered on LWN in May), this patch preserves the long filename functionality which is at the core of the patent. There's an associated FAQ which describes the patch and the approach which has been taken in its development and posting. "The claims of both of the VFAT patents involve the creation (or storing) of both a long filename and a short filename for a file. The 2nd patch only creates/stores either a short filename or a long filename for a file, but never both. The 11 bytes created by vfat_build_dummy_83_buffer() to pad the field for short filenames cannot be used to access the file, and contain bytes which are invalid in FAT and VFAT filenames, and therefore are not filenames as that term is and has been used in the technical community."

(Log in to post comments)

A new VFAT patent avoidance patch

Posted Jun 27, 2009 15:38 UTC (Sat) by jordanb (guest, #45668) [Link]

So now that Microsoft has Linux intentionally breaking itself to wriggle around their bogus patent claims, all they have to do is make sure that the next service pack of Windows 7 bluescreens when it sees any bogus 8.3 filename at all.

A new VFAT patent avoidance patch

Posted Jun 27, 2009 16:23 UTC (Sat) by pr1268 (subscriber, #24648) [Link]

According to the FAQ linked in the summary, that "bug" (feature?) already exists in Win XP, but not Vista nor Windows 7. I'm referring to the random bytes filled in the 11-byte field used to store the short (8.3) filename.

A new VFAT patent avoidance patch

Posted Jun 27, 2009 16:40 UTC (Sat) by jeff@uclinux.org (guest, #8024) [Link]

I should think that if MS did that, it would raise Antitrust issues.

But I also think a problem will be encountered with some embedded devices,
where the 8.3 filename might be all they knows how to access... like a camera
or media player using a tiny custom RTOS. Yes, you can use msdos fs...

A new VFAT patent avoidance patch

Posted Jun 27, 2009 18:14 UTC (Sat) by tao (subscriber, #17563) [Link]

It's fairly uncommon that you create files (with long filenames) under linux and then use them for something on a camera though...

A new VFAT patent avoidance patch

Posted Jun 27, 2009 18:51 UTC (Sat) by hawk (subscriber, #3195) [Link]

Agreed, files are mostly copied off a camera, not onto it.
For a media player, however, it can definitely be an issue...

A new VFAT patent avoidance patch

Posted Jun 27, 2009 20:36 UTC (Sat) by tao (subscriber, #17563) [Link]

Yes, but then again, most (all?) media players support long file names, since there'd only be space for a 5 letter title otherwise (first 3 used for 01- or similar).

A new VFAT patent avoidance patch

Posted Jun 29, 2009 1:38 UTC (Mon) by jamesh (guest, #1159) [Link]

If the media player does not support long file names, then use short filenames when writing to it. The description of the patch indicates that you get valid FAT 8.3 file names if you use a valid 8.3 file name to start with.

A new VFAT patent avoidance patch

Posted Jun 29, 2009 14:29 UTC (Mon) by anselm (subscriber, #2796) [Link]

I think this is a dangerous assumption.

My digital camera connects to a TV set via a special »AV cable« with an RCA plug at the TV end. If I want to show a bunch of digital pictures to my parents, who don't happen to own one of those newfangled TV sets that take USB sticks, I put the selection on an SD card and use the digital camera to display the pictures on the TV via the AV connection. It's maybe not the greatest quality, but it works — and it is much nicer to sit in the living room than to crowd around the computer in the den.

Of course this involves assembling the collection of pictures to be shown on the computer (Digikam is useful for that). I don't know whether my camera will insist on/tolerate 8+3/long file names/both but I would certainly hate that functionality to go away, in particular because AFAIK the VFAT patents aren't actually valid here in Germany.

(Also, think »digital picture frames«.)

A new VFAT patent avoidance patch

Posted Jun 27, 2009 20:55 UTC (Sat) by drag (guest, #31333) [Link]

> So now that Microsoft has Linux intentionally breaking itself to wriggle around their bogus patent claims, all they have to do is make sure that the next service pack of Windows 7 bluescreens when it sees any bogus 8.3 filename at all.

Remove the word 'Microsoft' and replace it with 'The United States Government' and your statement would actually be accurate.

Lets be very aware of the actual group of people who are doing the oppressing here. Microsoft is one of thousands of different companies that can possibly leverage patent claims against Linux users. Blaming Microsoft with trying to deal with a system were they are much more frequently a victim then a aggressor isn't going to address the real problem here. (although, in this case, MS is certainly being a asshat)

Same thing with things like the DMCA, codec patents, RIAA going after grandmas, "tivo-isation", etc etc etc. These are symptoms of a government out of order.

A new VFAT patent avoidance patch

Posted Jun 27, 2009 22:28 UTC (Sat) by sbergman27 (guest, #10767) [Link]

"""
Lets be very aware of the actual group of people who are doing the oppressing here.
"""

Let's call that the "People don't kill people. Guns kill people." view.

In my view, the laws make it unnecessarily easy for people to use guns to kill people. (But that doesn't roll off the tongue quite as nicely.) The person pulling the trigger bears some responsibility for any killing they do. And MS did just recently pull the trigger against Tom Tom.

Sorry for the substitution, BTW. I'm all out of car analogies. ;-)

A new VFAT patent avoidance patch

Posted Jun 28, 2009 1:02 UTC (Sun) by drag (guest, #31333) [Link]

> In my view, the laws make it unnecessarily easy for people to use guns to kill people. (But that doesn't roll off the tongue quite as nicely.) The person pulling the trigger bears some responsibility for any killing they do. And MS did just recently pull the trigger against Tom Tom.

It sort a works like that, but sort of doesn't.

The law requires that people protect their patents or otherwise they will lose them. Microsoft needs their patents in order to protect themselves from other companies that wouldn't hesitate go after them. Microsoft has been the subject to numerous lawsuits regarding patents in the past. They've been sued under MP3 related patents, been sued for browser plugins, etc etc. All over the place.

In other words... the less you use your patents the more you are at risk of patent lawsuits.

Guns, in a armed society, operate mostly on a deterant level. I cities were there is a large amount of in-house gun ownership there tends to be much less home invasion and burglarly. I don't need to own a gun personally to benefit from that sort of thing.. the statistically likelihood that a violent attacker is met with deadly force is mearly enough to protect me. It's in towns with high levels of private gun ownership laws have the highest rates of that sort of violence because it's unlikely that private, law-abiding citizens are going to be able to protect themselves. (while gun laws have very little effect on the ability for criminals to arm themselves... the Washington DC area is a perfect example of that.)

It's a analogy that does not really work.

(this is why if you look at rates of gun ownership in the USA they do not correlate with rates of gun violence. Areas high in violence have the most restrictive laws and relatively low per-capita gun ownership.. were places (mostly rural) everybody and their mom owns a gun but the rate of gun deaths are low. If you want a correlation between gun violence and something.. that "something" is rates of illicit drug use and drug trafficing. People addicted to "dangerous" drugs do extreme things in their desperation and black-market gun trade attracts violate criminal elements. If you want to solve gun violence in the USA the best first step will be to semi-legalize drugs and remove the capitalistic incentives for the underground drug trade.)

A new VFAT patent avoidance patch

Posted Jun 28, 2009 1:24 UTC (Sun) by dlang (guest, #313) [Link]

if people lost their patents by not enforcing them there would be no such thing as a 'submarine' patent

there have been too many cases where something has been in use for years and then the patent owner starts to demand licensing for things to work the way you say they do in the real world.

I think that it's trademarks that have to be defended or you loose them.

A new VFAT patent avoidance patch

Posted Jun 28, 2009 2:22 UTC (Sun) by tao (subscriber, #17563) [Link]

+1 Insightful.

You're totally correct. Patents can be enforced or not on a totally arbitrary basis. Copyrights likewise. Trademarks have to be enforced strictly though, or they risk being considered void.

If patents had to be enforced non-arbitrarily not to lose validity, a lot less patent trolling would take place, since few if any trolls would dare go up against IBM or Microsoft in a patent court. And with its multitude of patents, IBM would not be able to do any sales/development, they'd be too busy spending time in court.

Mild straw man

Posted Jun 29, 2009 12:10 UTC (Mon) by k3ninho (subscriber, #50375) [Link]

To achieve the goal you have set up, you would need to change patents from being a granted monopoly with the right to exclude others from doing what you claim monopoly in. What you'd change them to, I can only guess, and it looks like this: a person applies for a patent for a novel and inventive device or method, which is near-automatically granted based on search results. Then if litigation is to occur, the apparent rights-holder has to have the patent examined as to novelty and non-obvious inventive concept.

This would have your must-be-enforced criterion met. As it stands, international law has a qualification that a patent holder may have their patent revoked if they aren't in business and making use of their patented device or method -- and if also the challenger can show that such behaviour has harmed their business, which is unheard-of in my experience of patent practice.

However, there's a downside to this approach: with a granted but unexamined patent (or a few hundred of them) I can threaten you much more cheaply than with present full-monopoly patents. So I think that the change would make the situation worse.

K3n.

no, patents are not "enforce or lose"

Posted Jun 28, 2009 3:35 UTC (Sun) by JoeBuck (subscriber, #2330) [Link]

It is a requirement that if you have a trademark, you have to enforce it or you risk losing it. There is no such requirement connected with patents. Microsoft was under no obligation to sue anyone for infringing a patent.

If there were such a requirement, the legal system would be completely deadlocked, because every software product on the market, plus most open source programs, infringes patents. If drag were right, everyone would be required to sue. But a lawsuit over a patent is always a choice.

A new VFAT patent avoidance patch

Posted Jun 28, 2009 3:40 UTC (Sun) by sbergman27 (guest, #10767) [Link]

"""
The law requires that people protect their patents or otherwise they will lose them.
"""

A very common misconception. Others have commented upon that. My first thought when I read it was about how convoluted our laws governing various forms of IP are, and how even people paying attention can lose track of the ball. But then it occurred to me that in this case, it makes a great deal of sense for trademarks to be handled in this way, and differently from patents and copyrights. If you aren't defending your old trademark... it's not really your trademark anymore.

Once government gets into the business of handing out monopolies (and we've actually moved out of the realm of "limited monopolies", haven't we?) things get complicated fast, even without the influence of special interests.

No wonder the wise folks at the Philadelphia Convention tried to be conservative in that vein. Unfortunately, they seem to have had more confidence in the Congressional branch they were creating than turned out to be warranted.

There's a science fiction short story (the author's name eludes me. Perhaps it was Asimov) in which William Shakespeare is thrown forward in time and ends up in a contemporary college course on the interpretation of the works of Shakespeare... and flunks it.

I sometimes wonder what the framers of the U.S. Constitution would have to say about our current condition, and legal interpretations of their document.

A new VFAT patent avoidance patch

Posted Jun 28, 2009 11:14 UTC (Sun) by nix (subscriber, #2304) [Link]

It is Asimov. _The Immortal Bard_. One of the works from his 1940s
efflorescence of creativity.

A new VFAT patent avoidance patch

Posted Jun 28, 2009 19:01 UTC (Sun) by sbergman27 (guest, #10767) [Link]

"""
It is Asimov. _The Immortal Bard_. One of the works from his 1940s
efflorescence of creativity.
"""

*sigh*

He's been gone 17 years... and I still think about him every day, I suspect. I never met him in person. Never corresponded with him. And yet, when he died, I felt as if I had lost a personal friend. An amazing man.

A new VFAT patent avoidance patch

Posted Jun 29, 2009 17:01 UTC (Mon) by davide.del.vento (guest, #59196) [Link]

Tomorrow I'll patent the verb "to be" and you will need to pay me royalties for almost every sentence you'll write. You are warned!

A new VFAT patent avoidance patch

Posted Jun 28, 2009 8:57 UTC (Sun) by zenaan (guest, #3778) [Link]

1. "Microsoft is one of thousands of different companies that can possibly leverage patent claims against Linux users."

True fact. However:

I can accuse you of being a pedophile, and you will have to defend yourself in court, against my accusation (to the extent that I choose to be legally and/ or publicly aggressive). You will have to prove that you are not a pedophile in a system that favors the accuser.

Whilst the onus of proof does lie upon the accuser, it is irrelevant, because the 'winner' in all these battles is the lawyers. Tom Tom is well aware of this.

Whilst the accusation and claim by Microsoft, most likely has no substantive ground and little or no prospect of success against a deep pocketed member of our community, it is nonetheless the case that Microsoft in its belligerent aggression against our community is willing to try any avenue, not the least of which is direct aggression by way of litigation.

Those in the know in the community are confident that it is only a matter of time before these particular patents fall. Unfortunately, the 'time to correct' and 'cost to correct' factors are an indictment against the US patent and legal systems.

2. "Blaming Microsoft with trying to deal with a system where they are much more frequently a victim then a aggressor"

False presumption.

Explanation:
An implied presumption you have put here is that 'because Microsoft is more frequently a victim, it's OK that they be an aggressor'. More strictly, you are saying that because Microsoft is more often the victim, we should _not_ blame them for being the aggressor (implied context, is _this_ particular case of Microsoft's aggression).

But if they're being an aggressor then they _are_ actually being an aggressor. And to blame or rather _name_ them as being an aggressor has benefits including helping others in our community to identify Microsoft as an aggressor against our community, at the very least in this particular case of actual, demonstrated aggression.

I say it is OK to name Microsoft as an aggressor when Microsoft acts aggressively towards those in our community!

3. Microsoft aggressive? Look at aggressive hardware deals (the win-modem fiasco is yet to be fully told), operating-system license pricing lock-in with pc vendors, the endless SCO saga (see groklaw.org), recycled "Get the facts" campaigns, anti-competitive practices, netbook 'sales figures', and on and on it goes...

Don't try to convince me Microsoft is unworthy of the "aggressive against the community" label! Anyone abused of such a notion, ought to promptly disabuse themselves of such a notion.

From the desk of Zen

A new VFAT patent avoidance patch

Posted Jun 27, 2009 17:20 UTC (Sat) by endecotp (guest, #36428) [Link]

I have a GPS that appears as a vfat USB storage device. It uses long filenames like 20090627095959.log. I had a script that reads these files and then renamed them to have .bak on the end.

Recently I disabled vfat and I've been trying to access these files via their 8.3 names, i.e. 200906~1.log. A snag is that I cannot append .bak but I don't mind instead replacing .log with .bak. Appart from this it seems to work. I am a bit nervous that because the long filenames are not being changed during the rename, something bad might happen when the GPS looks itself at the directory. So far this does not seem to have been a problem.

So I'm wondering what will happen with this new patch. Presumably I'll end up with valid long filenames but invalid 8.3 names after renaming, and I'll have to see whether the GPS minds that. My concern is that the GPS may dislike the "corrupt" entries and fail in some non-immediate way, e.g. creating corrupt files that I don't discover until much later.

I guess it comes down to which seems least likely to confuse devices like this, having no long filenames or having corrupt short filenames.

A new VFAT patent avoidance patch

Posted Jun 27, 2009 17:57 UTC (Sat) by ncm (guest, #165) [Link]

Or just turn off the patch.

A new VFAT patent avoidance patch

Posted Jun 30, 2009 1:59 UTC (Tue) by jlokier (guest, #52227) [Link]

That will be difficult when distros adopt the patch, if he/she's not already running a custom kernel, as it's not a run-time switch. Building a custom kernel isn't hard but it's not easy either if you're not used to it.


Copyright © 2009, Eklektix, Inc.
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds