Slashdot Mirror
Microsoft Patents Sudo's Behavior
Foofoobar writes "Just when you thought all was safe on the crazy patent front, Microsoft has come out of the obvious patent closet to file patent number 7617530, which basically duplicates the functionality of 'sudo' which is found in all Linux systems. PJ over at groklaw has a wonderful writeup on the entire fiasco."
It's US patent system's fault, not Microsoft. They have to file these to cover their own ass. And actually I haven't ever seen MS patent trolling, they've even gave their patents to organizations which purpose is to keep them open. Even the TomTom vs. Microsoft case was because TomTom attacked MS first and they had to counter.
Patent system is the one to blame.
Patenting sudo is a slight legitimate error?
Damn. I want some of that anti-guilt thing you are taking.
NO SIG
Oh no, I've gone cross-eyed.
According to patent law, the above example of murder-by-verbiage is supposed to help third-parties implement the invention described, but the language employed is clearly designed to accomplish the exact opposite. I think it's time to put the patent system out of its misery.
"In prison you just have to shut your eyes and take it. Here you have to shut your eyes and give it."
Where's your analysis of the degree to which this "isn't exactly sudo"? It's pretty damn close. If it comes down to the degree of "exactly," please provide some examples from patent case law that show that the degree of difference here is sufficient for the two programs not to be close enough to the same that sudo, had it been invented after this patent, wouldn't violate said patent.
I'm nothing like a patent attorney. But my understanding is that if someone invents a special right-angle shovel, and patents it, you're going to be in trouble even if your shovel head is only at an 80 degree angle rather than 90 degrees. If not at 80, certainly at 89.
Besides, this patent ends with language claiming that the method of implementation is only the preferred one, while the patent covers other methods of implementation of the same underlying concept. And in which sense is the underlying concept even a few degrees different from what sudo does? Your analysis?
"with their freedom lost all virtue lose" - Milton
Since when do programmers need to be patent lawyers? Patents are written in fluent legalese, not plain $HUMANLANGUAGEOFYOURCHOICE.
Understanding the scope of the problem is the first step on the path to true panic.
Adding a GUI is no more "creative" and "non-obvious" than adding "on the Internet".
Then again, it might be non-obvious to Microsoft. Does anyone remember if Microsoft XENIX had a sudo equivalent? It would be nice to use something from them from a quarter-century ago as prior art.
macos x has been doing this since its inception.
gksudo has been around for a long time as well.
this is NOT new.
You've said this in at least two different posts, yet failed to indicate what those do that this patent covers. For example, OSX doesn't present an interface with a "selectable help graphic", the selection of which causes display of other accounts that have a right to permit the task, based on frequency of use, association with the user, and an identified higher-rights account that can permit the task. And that's just three of the limitations of claim 1. I doubt gksudo does them either.
Thanks for telling us that those claims are too complicated for you to read. Please make sure to put that on your resume, because if I was a potential employer looking to hire you for anything even remotely technical, I'd want to know that you give up whenever a discussion gets remotely above the complexity of "M$ sux0rz."
That's not a technical description: it's legalese. I've done my share of technical writing, ranging from scientific journals articles to user and developer documentation, but I'd never be able to get away with producing such incomprehensible gibberish.
copyright doesn't protect against duplicating functionality - only copying the exact binaries/source code. If I want to write my own sudo replica, copyright doesn't stop me... but a patent would.
That is one of those statements where both sides shout "EXACTLY", and then stare at each other.
If you have a copy of his source code and duplicate product in another language, you'll get your clock cleaned in court. To effectively copy the other guys software, the best defense would be to have no knowledge whatsoever of his code.
Historically everyone in software has been copying everyone else all along. Things were fine before patents became all the rage. Imagine is Apple had patented the GUI in 1984. The windows GUI couldn't have been developed patent free until 1999. It's an absurd idea, no matter how much I currently dislike windows dominance. And, yes, I do realize Apple stole the GUI from Xerox...
Do we actually want to prevent duplication of functionality?
Also, isn't patent still supposed to allow duplicated functionality if the implementation is different enough?
And actually I haven't ever seen MS patent trolling,
Their shakedown of camera vendors and threats to OS implementors over the VFAT patents are a classic case of patent trolling.
The technology covered by the patents no longer has any intrinsic value, because nobody uses OSes that don't support long filenames. The only reason to use the long/short filename conversion in VFAT is purely circular: to ensure compatibility with VFAT itself.
Thus, these patents only remaining purpose in life is to create a barrier to entry in the markets that Microsoft operates in. The technology covered by them is is providing no end-user benefit, and consumers are paying royalties and getting nothing in return other than a less competitive market.
This patent was filed more than four years ago, in April of 2005. This filing predates Red Hat's announcement of PolicyKit by about a year. And PolicyKit probably wouldn't cover this even if it predated the Microsoft concept because it doesn't meet the "automatic" criteria, AFAIK.
And for anyone thinking that this is a patent on sudo, it is not. It also is not a patent on Apple's AuthorizationExecuteWithPrivileges, though it is much closer to that. It differs from the Mac OS X design in that it:
It further differs from sudo in that it presents a GUI (in addition to the two ways above).
Regarding launching a GUI window when a privilege violation occurs, this is precisely why Windows got the "Allow or Deny" reputation it got. You really don't want to authorize every little action. Further, when it comes to a typical desktop environment, a rights system should not be so complex that there are more than about two classes of users anyway---those who have the rights to modify system files and those who are limited to their own files. Therefore, something like sudo, PolicyKit, AuthorizationExecuteWithPrivileges, etc. is generally a much better design because it puts the application in control of the experience and allows you to run a series of actions with elevated privileges, forcing apps to be designed with proper privilege separation, and reserving elevated privileges for only the minimum portion of the code necessary. The Windows "automatically throw up a GUI when you get a permission denied" design has a significant risk of creating user indifference towards important security notifications, which results in a significantly less secure system in the long run.
Also, I'm under the impression (based on the patent) that Windows is temporarily elevating the privileges of the application itself, which means that you now have a much larger chunk of code that must be checked for security holes, lest malicious individuals co-opt the application for nefarious purposes. Such a design also makes it very hard to adequately use code signing to ensure the authenticity of the code running with elevated privileges, thus allowing security holes in the app to readily be exploited and turned into the equivalent of root holes just by the user clicking "Allow".
In short, it's a terrible security design filled with myriad fundamental design flaws, all codified in a patent filing for all to mock. I certainly won't lose sleep over this patent getting approved. No one should reasonably want to implement the sort of security architecture that would violate this patent.
Check out my sci-fi/humor trilogy at PatriotsBooks.
I always thought a sudo-intellectual was someone who thinks they are smart enough to be given the root password...
If we can put a man on the moon, why can't we shoot people for Apollo-related non-sequiturs?
Math equations can be brilliant and stunningly innovative yet they can't be patented. Why should software be any different?
Law is the programming language for the system of society. The problem is, rather than doing exactly what you told it to do, regardless of whether that's what you wanted it to do, the system makes every possible effort to interpret the code in such a way so that it doesn't have to do what you instructed it to do.
upon the advice of my lawyer, i have no sig at this time
I presented this argument to someone just the other day, but here it is again: Mathematicians develop insanely difficult and complex algorithms all the time, and must share their work in the public domain because you can't copyright or patent mathematics. Not a formula, I'm talking about full algorithms, logical procedures, proofs and so on. Algorithms which have changed the world by such orders of magnitude that no matter how novel and amazing some little piece of code looks to the programmer, compared to the work of mathematicians it is almost always will come up looking pretty much completely trivial.
Imagine if someone had patented the fast fourier transform? Or any number of a virtually infinite set of unique and groundbreaking algorithms that have literally changed the course of science. Technology and science would be weaker for it, you might not even have a job with a computer in front of you.
Now why is it that sequences of logical steps, algorithms, when developed by mathematicians are anybodies game, and yet when a programmer or a software company comes up with an algorithm, a sequence of logical steps no different to the ones in the field of mathematics, it is suddenly different and needs monopoly rights granted to the author? Do you honestly think that novel method 3.57a to make database requests in a unique way is as important to the world as something like the fft? Or the Kalman filter?
Get over yourselves programmers, your code is not special, logic is logic, patenting a logical procedure is about as wrong as it gets in my books. If you develop code and it is useful, you are going to be the foremost expert in your new system. You will make money without a patent. The problem is this isn't about putting food on the table, this is about geeks who fancy themselves Knuth thinking they ought to be millionaires.
There are multiple issues getting mixed up in the Groklaw article and the discussion on Slashdot.
A patent application has three hoops to jump through to be patentable:
35 U.S.C. 101 - the claims must be patentable subject matter. The question of "is software patentable" is what the Supreme Court is deciding in In Re Bilski. This is the largest issue most of the Slashdot community seems concerned about, and it's obviously a big issue right now. These claims, as written, may be patentable subject matter under current 101 criteria. This is why there were written with all the "computer readable media" language.
35 U.S.C. 102 - the claims must be "novel" subject matter. This is what people object to when they yell "BUT I DID THIS BACK IN 1990!"
35 U.S.C. 103 - the claims must be non-obvious subject matter. This is what most people appear to be objecting to in the present discussion....if sudo existed before this patent, then laying down Microsoft's GUI idea on top may be obvious. (This is NOT a Section 102 issue). This is the part where the patent office (and examiner) screwed up. Even if the examiner couldn't find a reference that taught exactly what Microsoft claimed, he/she should have at least rejected the now-issued claims as obvious. Maybe he did, but half-assed the rejection...who knows.
The Groklaw article points out an "obvious" patent and yells that is shouldn't be patentable subject matter. Those are two separate issues. Yes, it's probably obvious. Depending on your view of software patents, it should or should not be patentable subject matter. That fact that it's an "obvious" idea will NOT in any way be affected by the Supreme Court's decision in Bilski (that case is about patentable subject matter under Section 101).