Which Software Patents Are Worthwhile?

Dan Jagnow asks: "The USPTO issuing brain-dead patents is a recurring theme on Slashdot, as is the question of whether software patents ought to be issued at all. To turn the question around, what are some software innovations (historical or potential) that Slashdot readers feel merit patents?" I feel it quite reasonable to ask this question. If we must have software patents, and currently the USPTO feels this is so, then I would like to know which software patents you all feel are worthy of the name.

There are a number I believe to be good...

[Anonymous Coward]

While it was a pain in the ass, I felt that the RSA patent was valid and legitimate. The algorithm was very strictly defined in the patent, and the patent itself only covered the algorithm, nothing abstract.

The Diffie-Hellman patent is a little more problematic. While the Diffie-Hellman algorithm *itself* seems to be legitimate, I really don't agree with the idea of the D-H patent covering the entire concept of public-key cryptography (this caused problems with licensing the RSA patent alone, eventually leading to the creation of PKP, in which both patent-holders attempted to mutually rape each other...).

I guess I think that methods for doing something should be patentable (i.e., RSA and Diffie-Hellman). But the actual results of that process shouldn't be patentable. That is, a patent on a cryptographic algorithm doesn't mean you have a monopoly on using keys in cryptosystems.

None. Really.

[His name cannot be s]

It really does not benefit anyone to allow software to be patented (other than lawyers...).

Really, this is quite the flamebait question to start with...

The problems with software patents relate to USPTO's inablilty to determine (a) what is new, and (b) what is really innovative. The fact that the USPTO cannot verify either of these qualifications invalidates the entire system. In today's vast vast vast society, there are literally millions of people working out similar problems. The Patent system is geared solely to those who have $bucks$ and to those who can work the system. If a bright poor student in a far-off corner of the country comes out with something brilliant, and at the same time(or even after), some other rich company/individual figures out the same thing, the rich guy wins. Why? because the rich guy files the patent, and therefore, in order that the poor student to actually profit from his work, has to win the court case. Law #1 of the USA: No Money, No Way.(*see OJ Simpson.)

Let innovation stand, get rid of the USPTO.

G

Diffie-Hellman key exchange

[harryseldon]

Novel and non-obvious to one skilled in the art.

(If you think it's obvious, why did it take so long for someone to publish?)

Another nice property of this patent: it has expired.

Re:The lack of postings here means?

[MrPeach]

Or maybe it means that this was not on the main Slashdot page and so get's much less exposure to the Slashdot public?

It only gets seen by those that look at this section of Slashdot or who look in 'older stuff'.

Re:Algorithms

[fozzmeister]

what would not be able to be done if the b*tree algorithm was patented, forget lzw this'd be loads of programs, filesystems (reiser) etc. patents really need to be specific to a particular product, if valid at all

No MORE software Patents, or any patents!

[haplo21112]

The govenment of the US and other countries world wide needs to stop this non-sense. The idea of patents WAS(a big was) to protect inventors and allow them to reap profit and recognition for their work. The intent being they would create a new work, and then owning the patent beable to sell that work to some company for mass production. The system has become very preverted from this intent. Now, every company in the world seems to think that they have the right to patent every little idea, and then apply it very(VERY) broadly to discourage competition, or even shut down competition. The most fustrating cases of this are companies that turn around and get patents recently, and then try to apply them to long standing (or common sense solution) tech. The one I loved most is BT claiming they had the hyperlinking patent(if anyone deserves it I think XEROX, had the first use on this concept in the 60's pretty much when they invented the mouse). At the very least the rules need to change so that a patent cannot be enforced retroactively, or against already existing projects. Additionally if you fail to enforce your patent for a certian amount of time, I say one year, after vilolating tech comes into existance you loose the right.(See the .gif mess, that happened a few years back of an example where they should never have been able to enforce it because they ignored it for so long). Additionally patents should be norrowly enforced/granted on very specific details, ideally against working tech "This is what I did, it works like this", not "I have an idea it would work like this". Similar problems demand similar answers, in biology there is a theory called co-vergent evolution, that states that if life arose in two different places(on Earth, or else where in the universe) under similar circumstances, and conditions, that the lifeforms would be very much similar. Ie. We have sharks here, if you went to another earth type world out there someplace, it would probably have creatures, that looked, acted, and behaved very much like sharks.(A.C. Clarke expalins this very well in his book 2061). The same is true, of software works, my boss could come to me tomorrow, and tell me that he needs a system to do X on the Web site I develop. It is quite possible that another company in our same business, could have already have been doing X, in a very similar way to my design to the solution. I'd never know until one or both of us tried to claim we though it up first, where really we were doing it at the same time. The problem is similar the solution is going to be similar, we didn't use each others works or ideas though. Just my view.

It's all good

[scott1853]

Everybody seems to be thinking along technical guidelines as far as patents go. Especially everybody on /. that works in a high tech field. We usually like our answers in x = y + 1 format. Well that can't exist when you're judging the obviousness of something new. There's no reference point.

Everybody seems to be attempting to find a single flaw in this idea and implementation of a patent system but there are many flaws. Patents have existing for over a century and the process has really not been updated since creation. We have a hundred year old process being used to regulate an environment that changes daily. We also have corporations that abuse the flaws in this outdated system and lawyers that encourage them.

The idea of a patent is really something that will protect an idea, which in itself is a vague summary of a process that has not been implemented. For a patent to be granted, you should be required to provide an implementation to demonstrate the actual end result. Then the end results should be included in the evaluation of its worthiness of a patent as well as the enforcement of the patent. You can't evaluate a process if you can't see how that process will and can be used.

For demonstration purposes, read the following patent claim [delphion.com]. Take a step back and pretend you are not a computer guru. Would you be able to evaluate this as obvious. If I put myself in "impartial mode", I wouldn't know what the hell they were talking about. If I come back to reality, what they are talking about could be something like a complex encryption mechanism that will only allow certain machines to access certain parts of a file. Or, it could be something as simple as an XML file with tag attributes indicating whether or not end users should be provided with the enclosed information. If there were an actual demonstration of their intentions and implementation of this, then it's validity could easily be judged.

No software patents are worthwhile

[bdhall1313]

Any reasonably skilled programmer needing to solve the same problem that is covered by a patent is likely to come up with a very similar algorithm to solve the same problem without ever having seen or heard about the patented algorithm.

Just because one person encountered the problem first should not entitle them to prevent others from solving the same problem in a similar fashion.

Re:Diffie-Hellman key exchange

[pheph]

How can one train a minimum-wage mind to recognize 'obvious'?

Way too subjective

Re:

[account_deleted]

Comment removed based on user account deletion

Cryptography patents

[acidblood]

In general, any cryptographic algorithm patent is a good patent. It takes a lot of research to come up with these algorithms.

I really couldn't come up with anything else, but in general, you can spot a valid patent by excluding all others that don't meet these requirements:

-it must be non-obvious;
-it must be valid by itself and not only for being something done on a computer or the Web;
-there is no prior art.

Requirements #1 and #3 require some expertise from the patent issuers, and the USPTO is clearly lacking such expertise (although for some patents even common sense would do, but they seem to lack it as well.) Requirement #2 is a mere consequence of the other two.

We need some sort of peer review of patents. Before being issued, it must go public for people to comment on it, clue the examiners on whether there is prior art, or even if a previous patent exists, but it didn't contain the word "Web." And a paradigm shift would be nice as well: instead of granting all patents except a few, grant only a patent once the author proves it deserves it, kinda like when you defend a Ph.D. thesis. If I wasn't clear, think of it as whether you are guilty until you prove yourself innocent, or innocent until you are proven guilty.

It just occured to me...

[cmowire]

I just had an idea... ;)

How about we do patents sorta like Jurry duty. The Patent Office goes through the reputable jurried journals in a variety of different areas and makes a list. Then they pick randomly from the list a team of PhD's in the different areas. Each PhD would need to review X patents and give opinions. Put Y PhD per patent to ensure that there is enough difference in opinion.

The problem, of course, is that every quack-infested medical journal and every whacko PhD out there would insist that their publications would be included in the search. Which would then mean that they would need legal standards -- perhaps codifying that certain orginizations like the AMA, ACM, IEEE, etc. would need to approve of the journal, and that the PhD would have to be from a university with accreditation.

Algorithms

[wolfman3000]

I think that algorithms should be able to be patented. They require a LOT of work to create, and are often worthy of patents.

Our current Patents

[aol]

Here at America Online we believe that patents are the lifeblood of any company. We are granted mutiple patents every month. Here is a list of what is up for our April patent list.
Email
Emoticons
Online Chat
Software Plugins, (We are fairly sure winamp plugins were first, and we own winamp!)
The web browser
The html file format.
Check back in May when AOLTimeWarnerFordTri-StarUniversal will be patenting food. (Because no one else ever did!!)