Slashdot Mirror
Linked List Patented in 2006
An anonymous reader writes "Congratulations are in order to Ming-Jen Wang of LSI Logic Corporation who, in patent #10260471 managed to invent the linked list. From the abstract, "A computerized list is provided with auxiliary pointers for traversing the list in different sequences. One or more auxiliary pointers enable a fast, sequential traversal of the list with a minimum of computational time. Such lists may be used in any application where lists may be reordered for various purposes." Good-bye doubly linked list. We should also give praise to the extensive patent review performed by Cochran Freund & Young LLP."
People are immediatly shouting "prior art!" I don't care about prior art as much as I care about the fact that it's another software patent and unworthy of being patented.
OK, so where do I go to submit prior art?
I've got personal programming I've done that uses linked lists. I've got an instructor who's been teaching them in a 200-level C++ course for god knows how long. Hell, Herb Schildt's "C++: The Complete Reference" was published before this patent was filed in 2002.
tasks(723) drafts(105) languages(484) examples(29106)
How exactly is backwards traversal more complicated than multiple sort orders? Backwards is just another sort order.
617B3B7F7E7C7D7F00EOF
The same patent was mentioned in this Slashdot article. What is new since then?
This is completely trivial, and any programmer could easily come up with this idea, if the application he is working on happens to needs to store objects in several sequences. I would not be surprised if thousands of computer programs use this structure but don't actually have a name for it. It's just a bunch of pointers after all.
Luckely, the data structure descibed by the patent has very limited use. A more generic version would allow any number of pointers per item, and is typically called a graph. Still it is a terrible shame that the US patent office grants such patents.
assignment != equality != identity
Patenting y=mx+b as a method of drawing a line is wrong. Patenting a device that uses this to draw a line, that is different. Algorithms that describe basic ways of doing things are not novel. The biggest difference though is that with a device, if someone discovers some neat effect they can patent a certain way of doing it, but others can come an innovate on that device, inventing something that has the same effect but does it differently and is therefore not covered under the original patent. In software they are patenting the effect, here we have a mutli-way linked linked list, something that CS students have to do in freshman level classes. But now, we can't use it in any way without having to worry about defending ourselves. As someone who writes software for a living I would rather give up on ever getting a patent than having to worry for the rest of my life about implementing something someone else came up with and patented.
The there is the issue that computer science moves along at such a clip that one company having a monopoly on an idea for the standard patent term could seriously stifle innovation in the field, which is directly against what the entire point of patents are.
Well, if your read the patent, it's for triply-linked lists, the purpose for which I can only fathom. Possibly, it's easier to sequence them for block encryption or somesuch. Though, I also imagine that the patent will get killed for its obviousness; adding a tertiary pointer when you need to access a list via another sequence is pretty damned obvious.
110100 1101000 1101000 1100110 0 1101111 1101000 1100011 1
Texture mapping.
How about slocate?
Or rsync?
Oh and data muxing.
Fantastic!
Someday, I'll have a real sig.
This is actually great and should server as a reminder to us all.
We should continue to submit rediculous patents just to show how counter productive software patents in general actually are. Anyone still have their first year CS course notes? There should be plenty of material to patent!
I don't mind they patent a hammer but not how it's actually used.
Views expressed do not necessarily reflect those of the author.
First, when discussing patents (or copyright), everyone must agree that this is an artificial concept developed to try and achieve a purpose. The purpose in the U.S. is (arguably) to promote innovation. There's no natural or inherent quality of such concepts beyond an inconsistent and ill-defined sense of possessiveness.
Second, with that in mind, if we look at your question, there are a few ways to approach it. "more patentable" could mean a few different things. I'll argue the discussion about "should be more patentable" i.e. I'll be arguing achieving the purpose, not the current letter of the law legality, of which I'm not fully clear (patents aren't copyright, and that's where the bulk of my limited legal knowledge is).
When a device is patented, there are a few areas of benefit:
1) The industry benefits, because the patent is (theoretically) detailed and recorded, meaning that the knowledge will not be lost, and that knowledge will eventually become public domain.
2) The inventor benefits, because his/her device is recorded, and for a time, no one can use that device or claim credit for it without getting permission.
3) Licensees benefit, because they can develop new inventions that improve upon or use or otherwise rely upon the patented device.
4) Competitors benefit, because there is new idea introduced, and though they can't copy that implementation, they can try their own ways of achieving that idea. (knock-offs) In some cases, these prove to have some benefit over the original (a better method, cheaper production, etc)
I would argue that #4 is why algorithms, business processes, and the like shouldn't be patented. Patenting Babbage's engine doesn't prevent the next guy from trying his luck at building a better mousetrap, so long as he doesn't actually copy Babbage. Patenting the algorithm (effectively the concept) does. Less competition means less benefit, and if I build a better mousetrap that doesn't rely on anything from your mousetrap, I'm not violating your work.
That's a "should" argument from my personal views. It could well be that patents expressly try to protect ideas and concepts, though I suspect otherwise.
its nice that drug companies each invested $100m and more drug research, got their patents, and they are now selling the drugs for reasonable prices to public and drugs are accessible anywhere in the world. not only that, but they are also making discounts and providing easy access, and even giving out drugs in most poor and needy countries so that things like aids or other dangerous diseases do not spread out.
oh wait - none of these are valid in this universe - all these are happening in alternate universes.
Read radical news here
Well, if your read the patent, it's for triply-linked lists
True. So it is just slightly less ridiculous than the headline makes it out to be. For crying out loud, I implemented various sorting methods on my linked lists by adding multiple pointers to them two decades ago as a teenager, and I don't believe for a second that I was doing anything remarkable at all. Once you have heard of linked lists then doubly-linked lists, triply-linked lists or whatever-linked lists are pretty much obvious.
Anyhow, what is really missing in all of this discussion is a response from the patent submitter or the persons in charge of accepting the patent; we never get this on Slashdot nor the stories referred to. Since the patent appears to be so unbelievable, I am very curious as to what their official response would be. Perhaps some IT journalist can get one?
Finding compounds that are potentially active against some disease, especially compounds with relatively new mechanisms of action, is considered "basic research" - academia does a good job in this area, because it makes for good papers, PhD dissertations and academic plaudits.
Taking those compounds and ensuring they are safe to administer in humans (rather than just animal models), and that they are as efficacious or more efficacious than other existing treatments is a costly, time-consuming process that requires managing a huge staff, coordinating clinics and hospitals, managing information systems, etc. This is not something universities or most research labs are set up to do properly.
If you eliminated patents, you clearly wouldn't stop the scientists, but you'd put a huge damper on industry and financiers wanting to back the latter part of this process. This would result in far fewer drugs getting through the FDA approval process.
It may be the case that there is a more societally efficient way to do this than the current system, but I'm not sure what it is. One problem with the current system is that one effect of it is that the US effectively subsidizes other countries drug availability, because drug companies expect to earn a large portion of their profits here, and have to deal with centrally negotiated pricing and other issues in foreign markets. But I don't see how you'd think that eliminating patents entirely would help the situation.
Sure, the R&D bill needs to be paid. Right now, a ton of it is going into researching new lifestyle medicines that can be marketed via tv adds to affluent aging baby-boomers. That's where the big money is, not curing diseases. Academics are less inclined to pour $100M into researching a new arthritis medication simply because their old (though highly effective) one has gone out of patent.
You have a good point. I would add a second situation that for me is the worst with algorithms:
A patent shouldn't be granted if there's a good chance that someone will come across the same exact solution with no knowledge from the patent. By "good chance" I mean probability... in Information Theory terms this would mean that an algorithm should yield some amount of information before being patentable.
IMHO the n-linked lists don't meet the "amount of information" quota because it's very likely someone comes across this solution on his own when faced with certain types of problems. The same way that many people with no knowledge on sorting come across their own versions of bubble sort... or the same way you can "build" (i.e. emerge) quicksort out of some sorting theory.
As a Slashdot discussion grows longer, the probability of an analogy involving cars approaches one.