The language is very plain, you just have to skip to the sections which list the claimed infringements. One is for assigning unique identifiers to files (UUIDs, such as SHA1 sums in case of git), and another is about addressing files using those IDs. Basically they're attacking essential elements of git itself but because github is a company they sue them instead hoping to extract money.
There are more infringements but they all seem pretty trivial and variations of each other. They're related to data duplication and access permissions.
Sure, it isn't really that bad - mostly the same boilerplate repeated _ad nauseum_. They are being sued by two patent troll LLC's in the Eastern District of Texas. The patents (9) are very generic sounding:
* Data Processing System Using Substantially Unique Identifiers to Identify Data Items, Whereby Data Items Have the Same Identifiers.
* Identifying and Requesting Data in Network Using Identifiers Which Are Based On Contents of Data.
* Enforcement and Policing of Licensed Content Using Content-based Identifiers.
* Controlling Access to Data in a Data Processing System.
* Distributing and Accessing Data in a Data Processing System.
* Similarity-Based Access Control of Data in a Data Processing System.
* De-duplication of Data in a Data Processing System.
* Computer File System Using Content- Dependent File Identifiers.
* Accessing Data in a Data Processing System.
The rest of the document (linked above) just says that they want reparations for damages incurred from these 'patent infringements'.
After thoroughly reading patent 1, I must conclude that this patent is invalid as are all of the others as they're obvious applications of the first (some are even described within the first, in totality).
There was merit in the concept of file deduplication, especially in distributed file systems, as I'm unsure if this existed as prior art since I can't find any reasonable literature before 2000. I'm pretty sure someone had created a file deduplication script in UNIX at least 20-30 years prior, but I can't find references to that work, so I'll assume none exists and give this patent the benefit of the doubt. Though the 'uniq' command in UNIX could be trivially used to implement this if you consider the use of MD for file uniqueness checks, so even that credibility is questionable. For example, fill a file with contents such as:
<md hash> <filename>\n
For all files in the filesystem, then uniq on the first N characters where N is the length of the md hash, and use -d to display duplicates, then remove those files. This can also be used as the mechanism for "attaching a substantially unique id to a file" as well as half of the claims in the first patent, which to me makes a strong argument that it was obvious.
However, everything else about it describes an obvious application and implementation of a hash table using a "substantially unique" hashing method, which could easily be described by MD5 which was published in 1992. For that reason, at least half of the claims in this patent are covered by prior art, and so the patent is, IMO, far too broad to be considered valid.
The gist (ha) is that Github and Rackspace themselves use some techniques that are described in those patents. Because of that they're now being sued for patent infringement.
It's an 'obvious' patent troll, I doubt these patents would even hold up in court but 'till then...
How is it really possible that there are companies that just holds patents and are allowed to sue others who does things. No shame. What a world we are living at... This seems to be company who is suing Github and Rackspace. http://investing.businessweek.com/research/stocks/private/sn...
I think it's perfectly legitimate for a company to just hold patents and sue others based on those patents (or, more likely, license them). That's a core aspect of the patent system.
The problem here isn't patent protections. It's bad patents. We're still going to see submarine patents for the next ~16 years, and I'm sure patent trolls will exist beyond that (unless there is reform), but the patent system isn't the problem here. It's just that so many bad patents get through.
> I think it's perfectly legitimate for a company to just hold patents and sue others based on those patents (or, more likely, license them). That's a core aspect of the patent system.
I could understand "inevitable consequence", but I can't see where you get "core aspect". Explain?
The patent system is meant to reward inventors, including inventors who don't have the capital for commercial manufacture of their invention (i.e. by licensing it to someone who does).
Applying for a patent is a costly and tedious process, so it's not strange the there are companies ready to take the work from individual inventor's hands. If that weren't allowed, large companies would have an advantage since they have a legal department. In any case, making a law against patent trolls would be difficult because if you require a minimum amount of non-patent work from them, they would simply meet that amount and keep on trolling. I think the solution is not to make rules against them specifically, but to change the incentives. First make sure patent applications are non-trivial. Second place the initial burden of evidence on the plaintiff to prove that their patent is actually being infringed.
But you're right, these people have no shame. But it's just not the patent troll, it's also a host of lawyers and judges who have interest in this stuff going on.
