Slashdot Mirror
Altnet Threatens P2P Companies Over File Hash Patents
devil_doll writes "I saw over on p2pnet that Altnet is trying to 'mug' a number of P2P companies with seemingly bogus patents. One of them is titled 'Data processing system using substantially unique identifiers to identify data items, whereby identical data items have the same identifiers,' and appears to be nothing more than a simple hash table."
Next they'll be patenting making dumb patents.
When it comes to software isn't this just tautology?
"Nothing can shake my belief that this world is the fruit of a dark god whose shadow I extend." - Emil Michel Cioran
But that hash table is patented. It's a hell of a fight to get around a government-granted monopoly.
Then again, this is P2P we're talking about, so it's not like we're expecting them to close up shop because they are violating some ambiguous law.
P2P is here to stay. It's doubtful that this company will win in the long term because the technology is already out there used by millions of users. The genie, so to speak, is out of the bottle.
I did RTFA, and to be quite honest, I never knew that there were commercial p2p companies after the fall of Napster and co...
It seems that PiXPO, the company referred by the article is selling a p2p application that lets its users share photos (not trolling, but I don't see the point) - does any Slashdotter use commercial p2p products? If so, for what reason?
Any old databse basicly gets data and hashed it to generate an internal index key, ie not the real data but a unique identifyer to said data. Now given that and this approach was even taught in my days at school and were talking 20+ years ago. So just counter sue for extortion/intimidation/blackmailing and stuff these IP wannabe's. People who try to enforce silly patents are worse than organised crime, because the law dont see them for what they are; Well at least for now. Things change, just need bigger loo paper to handle it.
If it's not a real patent, wouldn't they get accused of fraud and fraud with the intent to extort, or whatever?
Or do you mean it's a real patent, but one that should be indefensible? That's a different matter.
If this is the case, maybe they're doing the standard trick of going after people too small to challenge the patent in court, who will settle quickly.
You need a special table to make hash?
Here is s simpler way -
1) Powder dry herb
2) Place in a jar of 90% isopropyl alcohol
3) Shake vigorously for 2 minutes
4) Strain, filter
5) Evaporate on a plate over a source of steam
6) Scrape up the goodies
7) Profit
...isn't a hash table just something to stop your hash dropping on the floor? I know the USPTO has made some weird decisions, but still...
Did he inhale?
What, from my reading, the patented technology does, is find dupes, and reassign the "truename." to the dupes, whether remotely or locally.
For example, you have foo.txt. Someone copies foo.txt to bar.txt, without changing any of the data contained within foo.txt (it's some pretty piece of ascii art, just to keep you amused for a moment....).
This thing would keep tables on the files, and when run, would go back and rename bar.txt to foo.txt if wanted, or could delete bar.txt if the user requested.
But still, it's pretty obtuse. Even as someone with legal training, and a computing background, I had a hard time making out exactly what they were patenting.
A link to the Washington Post article mentioned in the p2pnet article would be nice, too, if someone can find it...?
Yet more evidence that reporters are idiots.
:)
When I did my own patent last year, part of the process involved the patent lawyer explaining how to read patent'ese. It's just like a programming language.
Claim 1 - hashed files
Claim 2 == Claim 1 && something else
Claim 3 == Claim 2 && something else
etc
etc
So claim 1 probably has no chance of being enforced whatsoever.
However, claim 25 may be enforcable.
"some incredibibly specific thing in the context of some bigger thing in the context of some bigger thing... etc... in the context of a bunch of hashed files"
If they wrote it all in one claim, then it would only take the most minute difference to invalidate the whole thing.
So they do this 1 && 2 && 3 etc etc thing so that they get real coverage.
Nobody expects claim 1 to be upheld.
Think of it as a giant complex regular expression on the field of computing.
That said, it does appear like it's an attempt to create a blanket patent of the entire field of manipulation and distribution of hashed files, and so it's probably still qualifiable as a mugging
But it's not an attempt to patent the hash table.
...why patent it? ...while the lawsuit will be thrown out of court as soon as the P2P company will show: "We use MD5 which expired even before this patent was granted, and this patent covers exactly the same thing as MD5 only without technical details how to accomplish the task." And even if not, sooner or later some company WILL start a lawsuit, and once the obvious result makes the patent invalid, all companies that actually paid, may counter-sue for damages.
I mean, they HAD to know the patent is bogus. They hoped it will pass through USPTO, and they hoped right. But how can they hope anyone will agree to pay them money for that?
If I wanted to sell Eiffel Tower, I don't think I'd avoid jail. Why people who try to sell (force!) idea they don't own could go free?
Anagram("United States of America") == "Dine out, taste a Mac, fries"
When I was much much younger, I used to purchase a magazine - Micro User for my BBC Micro. For about 10+ years, this published code listings ever month. You typed in 500ish lines of code, and were rewarded with a game or a useful little utility.
It was very frustrating to enter all of the code and not have the program run. Therefore, they introduced a checksum program. This ran on the code and gave you a string of digits back, which you could compare with the digits issued in the magazine. This was active from 1984 onwards, and most likely even before that.
Prior art?
Sunday you're Thinking Different, Monday you're a huge tool, paying too much and waiting to think like everyone else.
But still, it's pretty obtuse. Even as someone with legal training, and a computing background, I had a hard time making out exactly what they were patenting.
Patents were supposed to be monopolies granted for disclosing a discovery - that's what sets them apart from trade secrets.
In reality, most patents can barely be understood by anyone but other patent lawyers. They are designed to be obtuse, complex, uninformative and in legalese. Why? So the patent will have no actual value, it is basicly a free monopoly.
At least we in Europe got away from the abomination called software patents, this round. But there'll be more...
Kjella
Live today, because you never know what tomorrow brings
I saw david Boies last night on Fox news talking about the fact that our justice system was broke. Even though he still didn't get it, he had to admit that it had serious problems. His scenario for effective jurisprudence was when large entities bought expensive lawyers or when poor people couldnt afford lawyers at all. He just didn't get the fact that large companies to wasting capitol on lawsuits is very bad for the whole economy and having poor people accepting rough justice is very bad for society.
It doesn't matter wheather the patent is right or wrong, it doesn't matter how rediculous the tort, what matter is if it will generate collectible fees for a lawyer. If you are upset about rediculous government granted monopolies get upset about the monopoly on justice granted to lawyers. The fact that one of the most common tactics employed by large companies to eliminate competition is litigation to death should be enough for anyone to realize its time to do something.
I started a thread on the P2P-Hackers mailing list abuot this, and a number of people have responded with examples of prior art and other relevant information. You can find the post that starts this thread here.
Unique association of identification string ("the patent number") with the content of the patent is obviously patented under patent# 5,978,791 for which USPTO most obviously doesn't have patent rights. So most obviously they are in violation of a patent they have granted.
Anagram("United States of America") == "Dine out, taste a Mac, fries"
Nice. Your disustingly obscene fees are just giving the fatcats yet another statistic - "We had to spent xxx thousands of dollars per hour to defend against these criminals!"
Maybe you should try actually talking to your esteemed clients, and suggesting to them that if, quoting the music industry as an example, they didn't want to make $8 of PURE PROFIT on a CD, whilst giving artists a 50 cents share that they may be able to reduce prices, and make the cost of a CD lower to the point that it becomes less attractive to download music than to buy an inexpensive plastic disc. But then you might be actually making a difference and changing things for the better, and I guess if you wanted to do that, you wouldn't be an attorney.
Sunday you're Thinking Different, Monday you're a huge tool, paying too much and waiting to think like everyone else.
There are certainly hundreds of cases of prior art, and Tripwire is probably one of them. It computes and maintains a database of hashes for all the files on a file system to check for intrusions and corruption. The wiki entry says it first surfaced in 1992.
It seems that my purchase of An Introduction to Algorithms has borne fruit. I actually understood the patent application, although it's probably the worst description of a hash table ever. It's worth reading the patent, BTW, it has an unintentionally silly background history for its case.
Insofar as it's a specialized implementation of a hash table, how altnet thinks it has a case is beyond me. Code containing the word TrueName would be a dead giveaway, otherwise this is just harrassment litigation.
insecurity asks the wrong question irritation gives the wrong answer
This is a common thing in US patent law. A patent in the US isn't really granted until it is upheld by a court. Often, one company will target a smaller, weaker company with a similar product for patent infringement on the hopes that they can beat the company in court, have their patent upheld thereby, and then go after bigger fish with requests for royalties.
I was once involved in a patent case where a medium sized company was suing a smaller company over their use of "laser etched checking fixtures" (an industry standard item). The larger company's patent documentation was so vague that part of it actually seemed to be claiming the invention of the Cartesian coordinate system and the idea of graph paper.
It was laughed out of court but just think who they could have sued if that patent would have been upheld.
I suspect we would never have given software patents a second thought, were it not for the countless abuses that were foisted on the world. In other words, the people getting the patents brought our rage down on themselves by being total asses about it. One-click patent indeed...
I'd love to see a list of top-ten "good software patents". In other words, patents that meet (at least) the following criteria:
- The patent is on software (duh).
- The patent covers something not entirely obvious to an experienced programmer (the "five minute test": given the problem, could an experienced problem come up with a solution in less than five minutes?).
- The patent represents an innovation, rather than a restating of previous known techniques (as this one appears to be).
- The patent describes something that actually exists, as opposed to wishful thinking (like patents on artificial intelligence)
And since everyone who is in favor of software patents mentions that the poor inventor spent so much of his time and resources, I'll also add:
- The patent protects significant investment.
To me the "five minute test" is the most important: any problem that can be solved in that time isn't worthy of a patent, and any patents in that category will only hamper development of the field as a whole. Maybe the patent office should have panels of experienced programmers who get five minutes to reproduce each patent, immediately invalidating it if they do? That would certainly cut down on a lot of crap...
incidentally, any "all patents must be abolished" responders need not bother. go visit economic history 101 instead.
Is that the one where you learn that the USA became an industrial and economic powerhouse by shamelessly stealing every invention they could from Europe during its formative years, i.e. before it acknowledged any so-called intellectual property from other places in the world?
The Article
With some patented hydraulic invention, I am still free to come up with a better way of doing the same thing.
With these software patents, I'm prohibited from making anything that accomplishes X, even if I have a novel method, because company Y has a patent on software that does that.Free Mac Mini Yeah, it's
I am utterly certain this will be _the_ decisive blow in the great saga of MPAA/RIAA vs P2P. The P2P companies have nothing but the utmost respect for all Intellectual Property, Copyright in particular. I'm sure they revere and respect Patents equally and will hasten to ensure that they are free of all infringement immediately.
Not necessarily :-)
The "counting argument" can be used to show that no compression algorithm can hope to reduce the size of every file. Hence most (all?) compression programs will, from time to time, have to store an uncompressed version of the input file, and will thus have to have some mechanism for signalling to the decompressor that they have done so. Even if this "flag" is one bit long it will still represent an increase in size compared to the input file.
Real-life compression programs will of course be storing other information in the output file (e.g. original filename, size/checksum) so in practice we are talking about several bytes increase rather than one bit.
Note it's still possible to create a hash that is unique (like the above) but retrieving the content of the file from it (decompressing) is impossible. Just take your .bz2 "hash" and encrypt it, using its own MD5 as password.
Not actually impossible, just decidedly time-consuming. As mentioned above, a .bz2 file will have a recognizable internal structure. It's then just a matter of brute-forcing the decryption until a file of the appropriate format appears. With 128 bits of MD5 "key" this will take quite a while, but it's theoretically possible.
"Maybe you should try actually talking to your esteemed clients, and suggesting to them that if, quoting the music industry as an example, they didn't want to make $8 of PURE PROFIT on a CD"
I'm not sure I understand you. Are you talking Australian dollars? In the US, CDs are sold into the channel for around $8 and are sold to the consumer for about $12.95. That $5 delta is used to pay for employees, keeping the lights on, etc. shrinkage and so on and is not "pure profit." More importantly, it's money the record company never sees.
Of the $8 that the record company gets for the CD, about two bucks goes to cost of manufacturing, a buck goes to royalties, and some goes to marketing, accural for returns (the record company eats the cost when the retailer returns it) and various other realities of selling something on the retail market. Even if sales, marketing and shipping were free, and no CDs ever got returned or damaged, you're still looking at a gross of about four or five bucks, which is, percentage-wise, about what Logitech makes on mice.
I'm guessing you're confusing gross profit with net profit, but as people who work in retail know, all that matters is the actual money you have left after paying for all the expenses. The term "pure profit" would more typically be applied to the net, and not the gross. Overall, the record industry gets by with pretty shitty net margins compared to most of the stuff that you buy. It's good to argue against unfair IP laws, but claiming that "record companies make too much selling CDs" is the wrong approach.
Sitting in my day care, the art is decopainted.