Slashdot Mirror
BT Sues Google Over Android
phonewebcam writes "British Telecom is claiming billions of dollars of damages from Google in a lawsuit filed in the U.S. which says that the Android mobile operating system infringes a number of the telecoms company's key patents. The lawsuit, filed in the state of Delaware in the U.S., relates to six patents which BT says are infringed by the Google Maps, Google Music, location-based advertising and Android Market products on Android. If successful, the suit could mean that Google or mobile handset makers will have to pay BT royalties on each Android handset in use and which they produce."
The patent cited in the article as an example of BT's amazingly valuable innovations:
Deciding whether to stream music based on whether the phone is using a wifi or cellular connection.
Why the fuck is that patentable? Seriously, I just looked up "obvious" in the dictionary and it gave that idea as an example. That's not an invention or a technology. What the fuck, America?
Here's my poor non-lawyer summary of the patents for those too lazy to look them up.
Busioc granted in 2000 seems to be a troll patent on anything that reacts to detected network characteristics. TCP window size control since the 80s seems to be prior art, although anycast root DNS servers from the 90s would appear to be a close second.
Mannings1 granted in 1994 seems to be a troll patent on anything navigational that relies on a base and mobile part. Like LORAN from the 60s, or any of the moon shots from the 60s where the capsule relied on the IBM 7094 mainframe to run the calcs back home.
Titmuss1 granted in 2002 seems to be a troll patent where the the contents of a list depend on the location of the user. Like my Garmin GPS-12XL "nearest waypoint list" from the 90s, or any brick and mortar website with a "find the closest store" functionality.
Gittins granted 2003 seems to be a troll patent where you have a database server accessed over the network that has user based permissions. Like any mysql installation. It seems to be a pretty good description of the DB2 IBM mainframe server I was tangentially involved with about 20 years ago (%^&# source route bridging SDLC by mac addresses still gives me nightmares)
Mannings2 granted 2003 seems to be a troll patent where you have a Mannings1 system plus the result depends on the type of vehicle. Apparently providing different "walk" vs "drive" route results is safe because my shoes are not a vehicle, but providing "car" vs "boat" results would be a direct violation of this patent.
Titmuss2 granted 2004 seems to be a troll patent where a distributed architecture and network is used to store location information. Basically, any computing infrastructure storing location information that does not have an obvious single point of failure; The CLR/DLR circuit layout system from my previous telecom employer would seem to be a pretty good example of an infringing product; of course that was from the 1980s, and Ma Bell had much older networked location aware systems. Remember ma bell's weird V+H coordinate system? I do.
I believe this is a pretty accurate non-lawyer summary of the patents involved.
"Science flies us to the moon. Religion flies us into buildings." - Victor Stenger
I don't think they are patent trolls, somehow.
http://en.wikipedia.org/wiki/BT_Group
Depends on your point of view. This is the company that sued (and lost) claiming a patent on the hyperlink.
http://www.zdnet.co.uk/news/networking/2000/12/18/bt-sues-over-hyperlink-claim-2083266/
Carmack posted a better message about software patents right here on /. : http://games.slashdot.org/comments.pl?sid=151312&cid=12701745
Dilbert RSS feed
To the extent developers feel otherwise, I haven't seen a calm, and rationale counter-argument.
The obvious problem if you're expecting citations to caselaw or statutes is that developers don't know anything about that stuff. They see the results, they don't like them, they complain about it.
But let me see if I can help you understand the frustration:
Let's put software patent holders into three classes: Larger practicing entities (e.g. Microsoft), smaller practicing entities (start-ups) and non-practicing entities (trolls). The problem is this: Consider what happens in the interaction between members of each class.
It sounds like you already agree that the trolls are a problem, so l won't expound upon that other than to say that it's a serious problem when actual innovators can be stopped in their tracks by an army invincible trolls that never have to defend and so can keep attacking until they win.
So let's consider the interactions between the two remaining classes of software patent holders. The least problematic is interactions between two smaller companies, because there basically aren't any; neither is worth the other's time. But neither do these non-interactions provide anyone any benefit. They just don't happen at any scale. Next least problematic is the interactions between two larger companies, because they generally just cross-license. This isn't the most serious problem, but it does have the tendency to create a wealth transfer from smaller large companies to larger large companies, since the latter naturally tend to have more patents, and that can be seen as undesirable. (The largest companies have enough advantages, thank you.) Moreover, it creates a large inefficiency, because most patents are just used in cross-licensing negotiations to cancel out the other guy's patents, so that the whole exercise becomes a giant waste of resources because each side would be better off if neither had any patents (because they didn't exist) than to have to file, prosecute and negotiate only to came to the same result: a cross-license that cancels out the respective efforts of each company with those of the other.
But the most serious problem is the last: Interactions between major companies and start-ups. This is where the patent system is supposed to make up for itself, by preventing a large company from copying the little guy's product with impunity. The problem is that because of the way software patents work, and the way software works, unlike in other industries you just can't make a significant product that doesn't infringe major companies' software patents. They have too many that are too broad, and even if you think you can win you certainly can't afford to litigate it as a small company. So you the patent system is worse than nothing for the little guy, because it allows larger companies to crush them with patent litigation at will. If the larger company wants to infringe their patents, all they have to do is hang the threat of litigation over them before the larger company starts infringing, which puts the little guy over a barrel and allows the larger company to extract a license. Or just force them to accept a buy out for millions instead of billions. The only way a smaller company can defend itself is by discontinuing its product and becoming a patent troll, which as far as I can tell is the single most abominable result possible.
You raise some fair points, although I disagree that smaller companies interacting doesn't benefit anyone; they may form a partnership based on a mutual use of their respective technologies and establish a market foothold that neither could individually.
Sure, they can form a partnership if their technologies are complementary, but couldn't they do that regardless of software patents?
Second, and I realize this is a little unfair to raise it at this stage, but I don't see how software patents are special or should be treated with any differently than any other technology. None of your examples hinge on the technology being software - those scenarios apply in the biology world or semiconsiductor world as well. Why is it that software patents specifically should be abolished? Or are you arguing that the entire patent system should go?
Software is a different animal because it has effectively zero reproduction costs. The consequence is that you get the kitchen sink with everything. There is no feature that must be left out because it costs too much to manufacture.
It means that everything is infringing everything because everything includes everything. If you invent a new brake disc for a car, you don't include a free car with every brake disc you sell. In software, you do, because including the rest of the car has no cost. You can include every kind of car and truck ever made, every different kind of tire, the road, a garage full of lifts and tools and a variety of petroleum refineries and electrical generating stations to provide fuel for the various cars. All by just including a few libraries. And users expect these things. Your competitors have them. Which makes the surface area for possible infringement expand to completely unmanageable proportions.
It breaks an assumption the patent system is built on, which is that non-obvious combinations of prior art stuff are unlikely to be created at random. When you can combine everything with anything for free, getting those otherwise unlikely combinations without even trying suddenly becomes extremely likely. Independent invention becomes the rule rather than the exception. Which leads to all of these problems.