Intel Owns Patent on Distributed Computing 176
GnrcMan writes "Now here is something frightening: This patent describes a method of using an ISP subscriber's CPU cycles to process the ISP's data." As if Distributed.net, SETI@home, and other, similar projects hadn't been doing essentially the same thing all along, eh?
Re:I claim first comment. (Score:1)
Greg
If jumping to conclusions were an Olympic sport... (Score:1)
Someone who submits the abstract of a patent, which has no legal effect or enforceability whatsoever, gets moderated up to a "2," whereas someone else who tries to start a discussion of the patent *claims* (which *do* have legal effect) and to show why the patent doesn't cover what some of you think it covers receives only a "1".
One might reasonably reach the conclusion that the moderators are trying to fan the flames just a little bit, especially after having featured the article about Corel and the GPL.
Come on, folks. Grow up. Act as smart as you think you are.
Re:busting bad patents - gnupatent.org lives (Score:1)
Is this something that could be done for all GNU software? Is this something that we *should* be doing? To say nothing about the *BSD peoples concerns.
What about me? (Score:1)
I call it WALKING.
Re:anti-linux trademark... (Score:1)
WTF? COM? CORBA? @HOME? GIMPS? Got brain? (Score:1)
Shit, any OS with any multithread/tasking at all has (minimally) message passing. What's the difference if the other thread/task is local or not.
Either that, or I'm missing something.
OBJ patent first post:
Oh, and by the way, I just patented semaphores, so, please, you all owe me money.
Even More details (Score:2)
First of, the basis is exactly as for SETI@home but for the one single line (1a): All of the 'for the good of the world' process distribution systems (like SETI@home) are initiated localy, at the discretion of the user. If I want to run my client overnight that's my decision, not the decision of SETI@home. But here Intel patents a method where the user has nothing to say about if s/he wants to give processor power away.
Moving on (16): Now this, at least to my untrained eyes, seems to say that any time the central computer wants it the remote must perform the task. Once again there is no question of 'wanting', it's simply stated that the central comp can order the remote to perform certain tasks. Of course there is the part about 'at a predetermined time'. But that doesn't say anything about what that time is. An ISP using this system might require it's customers computers to be available at any given 'predetermined time', even if this would be in the mid-afternoon when they're in use.
The idea is expanded in 21a: The first set of tasks, stored localy on the remote, might be nothing other than a program that listens for a request from the central server. Thus the servers might, on a predetermined schedule say once every minute, querry the remote. If the remote would then be available it would recieve new data (ie second set) that would contain the 'real' data to be computed. Brilliant, AOL can now enter our comps and request us to compute for them at any time.
Correlation 1(25): And 2 (27):
Ok, so now it's a quetsion of intention. In the patent Intel states its benign intentions, to lessen the server load of an ISP and/or sitehost. But, as with any patent, this could be seen more generaly as the ability to 'slave' a users computer on the whim of an outside agency. Thus it is up to the reader to decide whether they belive in the goodwill of the Intel Corp. or not.
On a side note, wasn't this form of distribution common already in the 70s with linked Microcomputers being assigned tasks from a central computer?
Security Problems Here (Score:2)
While it's true that this patent relates to distributed computing, that's not what this is about. What they've really got here is a patent on a method of stealing subscribers CPU cycles. Better check out the ToS when you sign up with an ISP cause, guess what? If this idea comes to be, you might just be required to do some of the processing for your ISP. Another quote to illustrate exactly what they're looking to process:
If a ISP did that you could launch some very nasty attacks against their network by sending them back bad data. Depending on exactly what the parallel processing does you could cripple an ISP. You would have to check the acuracy of all of the data sent back from your customers. That would take up lots of cpu time. Probably more time then you would gain. It just isn't practical unless you can trust your customers. I don't think many ISP's do.
Secondly this isn't compatible with Linux. :)
Re: (Score:1)
time pre-determined.... (Score:1)
A lawyer could argue that it is 'predetermined time frame' that after getting the new packet, work is started.
And another lawyer could argue that SETI is done for personal reasons, and try to invoke the 'built for personal use' clause.
(being a realist, Intel won't bother with SETI/Distributed.net/etc. They want things that generate cash, and therefore SETI etc la don't matter.)
I even doubt they will get a $ on this patent....instead it will be part of the 'trade' when companies swap patents. (thus erecting barriers for entry for any new company. Helps keep your competitors known and limited. It would be a bad thing to have a new gal come in upset the way business has always been done.) If it never costs you, you have no reason to challenge it.
As for prior art I submit CRON and tape backup scripts I've seen in the past on UseNet.
Suggestion: (Score:2)
Just my opinion, of course...
- A.P.
--
"One World, one Web, one Program" - Microsoft promotional ad
Re:Microsoft (Score:1)
-- ----------------------------------------------
Vive le logiciel... Libre!!!
What on earth are the US patent office doing? (Score:1)
What on earth are the US patent office doing?
I wonder too. In France, having a patent implies that a "recherche d'antériorité" has been done. It's a kind of research in archives to check if your idea haven't already been patented and/or discribed. This research is done by the administration, and ensures you your idea can be patented (actually it costs money).
In the US, it seems that the patent just proves that you had the idea at the time you get the patent. So in a court, you will be able to use this as a fact to defend you right for money, but it does not ensure you will win if someone else claims for anteriority.
In my opinion, the US system gives more strengh to the guys with a-lot-of-lawyers. You will never go to a court if you are alone with no money.
Fasion? (Score:1)
I think I'm going to patent going to the toilet.
No, that's not it. (Score:1)
I mean, you can't say that all software patents are invalid because someone had already invented assembly language. Or that no mechanical devices can be patented because they all make use of the classical machines (lever, wheel, whatever, I forget, I'm a software guy).
raph's [mailto] mention of Sprite [berkeley.edu] seems to be getting close...the "main" system determines what and when the "remote" computers execute.
--
Note the date (Score:2)
Wouldn't a CPU ID come in handy for this? (Score:1)
One word: UUX (Unix to Unix eXecution) (Score:1)
The remote computer was capable of determining when (if at all) it was going to accept execution requests (this was much more important when just LOADING emacs took a few seconds an a good-sized VAX). A computer's admin could also decide WHICH commands could be executed by the so-called central computer.
Of course it all fell together when UUCP over TCP/IP (inc. over a lan) was implemented. You now have the INTEL patent, as described, over TCP/IP and in the earliest days of the internet (read: early to mid-80's). If you ignore the need for it to be done under TCP/IP, then the prior art stretches back to the '70s.
How can they get through with this? (Score:2)
Of course this is a frightening situation because I start having this phantasy of getting letters from lawers saying I have to pay a license fee for using "their" software technology.
Can anyone familiar with patent law comment on what their chances are to get through with this?
This technology is documented, has been done, so acutally the patent should have been not accepted.
Cheers,
Chris
yea right (Score:1)
Intel is right (not in the way we think) (Score:1)
Abstract (Score:3)
A system executes computer processing tasks on a remote computer that communicates with a central computer. The remote computer receives a start message from the central computer. Based on the start message, the remote computer processes raw data to generate processed data, and stores the processed data. Finally, the remote computer sends a complete message to the central computer.
Re:It's a bit different (Score:1)
Just my
Re:What on earth are the US patent office doing? (Score:1)
I'm really scarred (Score:1)
The patent office has been patenting ideas like these for more than a decade, and you would think that they would learn by now, but they don't. The worst part is that distributed computing has been the thing over the past several years with SETI@Home, distributed.net, and the like. Even though Intel patented a specific form distributed computing, what does that say for the other forms? Does that mean Microsoft will patent distributed computing between office workstations and the corporate server?
I am currently working on a research project dealing with distributed computing with the idea remotely similar to what Intel patented. Does this mean I will have to stop research in order to keep the attorney away, or what? Gotta love corporate america.
On a personal note, my research project will be submitted to a science fair that is coincidentally sponsored by Intel. Weird.
First to file problems? (Score:1)
Re:How to make money: (Score:1)
Step 5: go ahead and try it. The state of patent law is as follows: the undermanned and undereducated patent office simply can't determine which ideas in technology are fresh and unique -- so when in doubt, they grant the patent.
However, if a patent is not legit (ie neither fresh nor unique), the courts will not hold up the patent... and the patent sluts know that.
IMNAL, and I am concerned about 'the government' making big incorrect decisions, but I have more faith in the judicial system than the legislative system...
So, at the very least, keep steps 1 and 6, but do not worry about the patent sluts, for they will be humbled.
- Tommy V
Why it might cause Feds to take a look at patents (Score:2)
A system executes computer processing tasks on a remote computer that communicates with a central computer. The remote computer receives a start message from the central computer. Based on the start message, the remote computer processes raw data to generate processed data, and stores the processed data. Finally, the remote computer sends a complete message to the central computer.
Ummmm... I think Microsoft might just initiate some action on this matter. That's DCOM. and probably CORBA as well, although I know nothing about it, an let's face it, whoever 'owns' CORBA isn't gonna get a nasty as M$ about it.
How is it not COM? A COM object in a 3-tier app exposes an interface, takes a request from the remote computer and returns data. Any M$ developer who attends there tech briefings knows Microsoft is COM-crazy these days. It's being billed as a cornerstone of every new development problem you need to solve. (OT: They are even taking about making the interface between objects XML.)
Anyway, I just think M$ is gonna take one look at that and say BULLSH*T!
Re:Breathing air (Score:1)
A system whereby oxygen (O) is transferred to cells in a liquid (blood) and other gasses are removed from said blood.
Inventors: God; The Almighty (Heaven)
Assignee: Most living creatures. (Exception: Al Gore)
Appl. No: 1
Filed: Way Back When
CLAIMS
1. A method of transferring usable and no usable gasses to and from the blood to facilite cell usage and growth. Whereby a repeating precess is utalized first filling containers (lungs) with raw material (air) and allowing it to pass close to the blood. The blood therby absorbs Oxygen and other needed gassesd and releases Carbon DiOxide and other unneeded gasses. Raw Exhaust is then expelled from the said containers, sometimes forcefully (cough).
BACKGROUND OF THE INVENTION
One day I was sitting around bored and said hey.. why don't I make something to take cure my boredom? YOu can read more about my inventions in the very first chapters of The Bible. Unless of course you are a scientist or an atheist. Then you can read some Darwin and see that it was all a big accident.
SUMMARY OF THE INVENTION
The present invention is for staying alive. Period. Stop using this invention and see what happens. We will be charging for its usage soon, that is if the HMO's don't start first.
Re:Security Problems Here (Score:1)
-TShrew
Re:It's a bit different (Score:1)
beowulf cluster? (Score:2)
Maybe some day when people try to enforce these patents, defendants will look to old
Re:Sun OS had fractal computation demo. (Score:1)
There was also a language called LINDA that ran on VMS and other idle machines at one of the government labs in the 1986-1989 time frame.
There were several VMS and Unix distributed make programs/wrappers that farmed work out to clusters based on the load average or other criteria of idleness. Again, this is in the late 1980's time frame.
As someone who read info-vax and comp.sources.unix during this time frame, I'm surprised that others haven't pointed out the archives of the early days of USENET.
Patent Examiners need many-eyes help (I repeat) (Score:1)
A parallel distributed search for prior art via public review is going to be a lot more effective than one or two examiners going through their search, IMHO.
Re:More details (Score:2)
Re:Dose Science Fiction qualify as prior art? (Score:1)
Bad cat... bad cat person...
Oh well.. thanks for da correction....
:)
Re:More details (Score:2)
Besides, yet another moronic patent is old news. To me, the idea of an ISP requiring subscribers to process data is more intriquing. Just how much mileage can you get out of, "Wow, that's a dumb obvious patent that has been done before". I think that whole angle is running out of steam. We all know that the patent system is broken right now and if I spent an hour looking through the database, I'd be able to find hundreds of examples to prove it.
--GnrcMan--
Learn about Patents (Score:3)
Re:How to make money: (Score:2)
This is a common misconception. The patent office does not have the duty to search for prior art. The patent applicant has the duty to notify the patent office of any and all material prior art. The patent office may find some on its own, but the job is not theirs, at least not under US patent law.
The reason behind this is twofold:
--
Re:How to make money: (Score:1)
--
Microsoft will fight this--maybe (Score:2)
What Intel is describing (in effect, the ability to push processing tasks to remote machines) is something clearly anticipated by Microsoft's COM technology. (In 1997 the marketing name du jour was OLE 2.0, but since "OLE" had come to be a synonym for "speed brake" they changed the name.)
This technique is really quite appealing to large corporations, particularly corporations with operations around the world. I have a client with 16,000 PCs in 34 countries (soon to be 25,000 PCs in 52 countries when a pending acquisition is completed). As I write this they have thousands of PCs sitting on the other side of the world, in the dark, doing nothing. Imagine being able to harness all that CPU time....
There are lots of CPU-intensive tasks that are particularly suited to processing on a plain old desktop PC (as opposed to a top-of-the-line smoking quad Xeon processor box). OCR, image manipulation, and image compression are three activities that spring to mind. If you have unused PCs sitting around, and the network bandwidth to get the data to those PCs, this kind of distributed computing can provide a tremendous benefit to a corporation for essentially zero additional capital.
Microsoft already makes this possible: create a DCOM agent on the remote PC. Ship an install set, and have the agent run the remote install. Ship the data, get the data back (repeat for the next eight hours). You can leave the installed app on that remote machine, or you could even get clever and remove it for security's sake. (This, incidentally, is precisely the technology in Internet Explorer that Microsoft is insisting must be considered a core part of Windows.)
Microsoft isn't going to let Intel prevent anybody from using DCOM--just isn't going to happen. But on the other hand, Microsoft may not fight very hard--Microsoft and Intel (and most of the major technology companies) have cross-licensing agreements that permit using patents of one another. (Sort of a free trade agreement among tech firms, if you will.) If this threatens the functionality of COM (which is the core of Microsoft's Distributed Network Architecture) then we'll see some fur fly. But I don't see that happening.
Zephyr? (Score:1)
Re:Beowulfs? (Score:1)
Looks like they were aiming for other things (Score:1)
I believe so because any person who has done serious work with distributed computing knows that hub's aren't supposed to originate transactions but to route them (As in receiving a start signal and sending the appropiate chunk of data or passing the request to another node).
BTW, SETI@Home is IMHO more a client-server oriented application than a real distributed network because all data is still centralized.
ZoeSch
Re:Fasion? (Score:1)
A Quote (Score:1)
Hmm... this concept sounds intriguing. I wonder who owns the patent on "a display unit"... I'm going to have to get my but over to the patent office, something like "a display unit" could be worth millions.
World too big for patents (Score:1)
Anyone know of an organization of some sort that's working toward the end of patents? if you do, please E-Mail me.
japanese company patents curry (Score:1)
Perhaps Intel is being too modest - maybe they should just patent computing? (hey, that's a joke, not a troll)
This patent covers all remote job entry! (Score:2)
Hear, hear (Score:1)
I've learned to examine what a patent actually claims before I decide whether it's ridiculous. The blurb that you read in a news article is never a description of what is actually patented.
--
Re:No, looking ahead (Score:1)
--
Prior art up the wazoo .... (Score:2)
It doesn't matter about the abstract and description sais - the claims are important - here there are 4 (plus dependants):
Note the 'predetermined times' clause, I think this seems to be what they think is different from prior art, though they seem to miss most of what's been done. Sadly the patent system only encourages the citing of prior patents - if someone did something and didn't patent it it's not likely to end up in the cites and the examiner will miss it
Here's prior art I can think of
Re:Intel is right (not in the way we think) (Score:2)
Allow me to introduce a concept that most slashdotters appear to have no concept of: fiduciary duty. The officers and directors have a legally binding and enforceable duty to their shareholders - and NOBODY ELSE - to use the assets of the corporation to maximize the value of the corporation itself. If the corporation has an asset and fails to take best advantage of it, then the shareholders can sue and the officers and directors can be held personally responsible for the lost value.
This principle means that a corporation that has a patentable invention has a duty to its shareholders to patent it, and a corporation that holds a patent has a duty to exploit it to the best advantage of the shareholders.
I don't like the way Unisys is handling the LZW patent either, but they're hardly evil for doing so. It's quite likely that they are only doing what the law requires of them.
--
Re:busting bad patents (Score:1)
May I suggest that along with a strictly factual database of prior art citations (as a service to the USPTO and anyone who has to defend against the patent pirates), the web site also include a collection of philosophical writings that discuss the common intellectual property of mankind and corporate attempts to sieze it for private gain. It might also be useful to have a Ridicule section that exposes the unscrupulous, the selfish, and the sly in colorful language that will draw the attention of popular media.
Re:anti-linux trademark... (Score:1)
True but not in the way you think (Score:2)
If two individuals or companies apply for patents on the same invention within a short time period neither of them deserves the patent.
If two or more companies apply for a patent of the same invention within a short time isn't it a proof that it was obvious enough that many different people could invent it, therefore making it non patentable.
I think this would be the logic reasoning and would have the same result with a small nuance, in the first case both believe they should have a patent on it, and they are angry not to have one. In the second case it prove that the thing is obvious, then denying any right to patent it AND following the rules of patents.
Not to worry (Score:1)
I as well as thousands/millions others who has written networked applications has prior art for this patent.
Besides, I more and more get the feeling that patents should be neglected in the field of computer science. At least the patents which everyone with basic knowledge in the field can see are merely another formulation of some well known problem or application of technology/theory.
I'm not going to check the patent office when I write apps. I'm convinced that I'm infringing a handfull of patents every time I write something, and thats good enough for me. This madness can be stopped in one way; not by changing the rules (because too many people with too much influence seem to like the rules), but by ignoring the rules.
It is not in general a viable solution to a bad rule, to just ignore it. Society can't build on a strategy like that. But in this case, I think it's justified. This could be thought of as a ``virtual civil war'', where the oppressed fight back but noone gets hurt.
I'm not afraid, and you shouldn't be either. If we live in a world where people can get arrested for using common knowledge in a way that helps society, then so be it. Go ahead and make my day.
You should sue the government, he stole your IP (Score:2)
"changes in the patent system in the US cannot be made retroactive in any way that would result in doing away with existing patents. That would be a government taking of private property"
I this isn't possible then when the copyright period was extended from 50 years from the author's death to 70 years to the author's death the then existing copyright shouldn't have retroactively been affected by it, meaning that Mickey and co would be public domain by 20003 or something like that.
I we go one step further, the government took YOUR property (that is, the twenty more years given to the then existing copyrighted material) and should refund you for the amount they stole you, that is for each dollar somebody gave to Disney that was related to Mickey.
Your government owe a VERY big debt to it's citizen, I think you should sue him, you could make a LOT of money (and make the State go bankrupt BTW)
'Inside' view from the patent office (Score:4)
What they are thinking about doint is adopting the (European?) system of posting patent applications on the web to get peer review (i.e. us) for prior art etc.
Bottom line -- the P.O. is overwhelmed with applications not of their own choosing and knows they are up the creek. They could do with help, not flames -- right your congressperson.
Tnx
Alanl
Sun OS had fractal computation demo. (Score:1)
Sinan
Analysis of this patent (Score:3)
When you analyze a patent, the most important thing to look at is the claims. The abstract has little or no legal force, it's just there to help people searching (it's on the front page of paper patents, which was important back in the days when people searched through stacks of them
That said, this is another example of the US Patent and Trademark Office screwing up royally. Claim 1 covers a pretty generic computation load-sharing system - a central computer keeping track of a bunch of tasks, sending a start message to a remote computer to ask it to take the task, and the remote computer sending a complete message back to the server.
The problem here is that there is prior art up the wazoo. I'm sure experts in the field could come up with more, but just about any operating system with process migration, such as Sprite [berkeley.edu] should do. The main work on this project was all done in the early '90s, plenty of time to serve as prior art.
The disclosure doesn't help the case. It talks about the types of tasks to be distributed in extraordinarily vague terms, so much so as to not make much sense. Compressing MPEG's remotely? Are they on crack? Even a 320x200 at 30fps is over 5 megabytes per second of raw data. And of course serving up web pages is fraught with problems, such as latency, security, admin costs, reliability, and so on. The patent does not so much as mention these problems, let alone propose a reasonable method of dealing with them.
So what do you do when you have such a crappy patent? I think a reasonable thing for free software authors to do is ignore it. In theory, Intel could bring suit against a free software project for violating this patent. However, in that case it seems likely to me that we'd be able to get a good pro bono legal team together, and the patent would almost certainly be overturned.
In the meantime, I think our best option is to keep well informed about patents in general, and about specific patents that may be relevant. Shallow, "golly gee-whiz, look at the patent they just got on breathing" stories don't help much.
Re:'Inside' view from the patent office (Score:1)
protest and civil disobedience that every other citizen is entitled to. They could show up on the
doorstep of the justice department with pitchforks, plowshares, and torches, if they "really" don't like the status quo. But I guess they don't dislike it *that* much now do they?
They *do* make the law. They make it stupid, by blindly following it.
But your patent examiner friend doesn't see it that way; rather, he chooses to be "powerless".
Re:Another idea... (Score:1)
--GnrcMan--
Re:Another idea... (Score:1)
Agree completely (Score:2)
Apparently he decided to play patent interpeter and make the (incorrect) assumption that this patented all of distributed computing. My comment [slashdot.org] explains what I believe to be the real discussion point. And I say discussion point because, hey, Intel hasn't made any indication that they will ever even do anything with this patent. There are lots of patents on things that never even get implemented.
--GnrcMan--
Software patents were not ruled fair game (Score:2)
As for peer review. too late. the cat is out of the bag. There are thousands of bogus software patents. The only way the PTO could again gain respect in my eyes would be to invalidate all software patents and start the approval process over again.
Plus I don't pity the patent examiners... just doing their job. oh please. A big part of their budget is based on how many patents they approve. So they have an incorrect bias in approving patents. It is a blantent case of conflict of interest. Much like how a judge can't preside on a case where he has an interest in the outcome (like he would make a mil on his stock in the defense won for example)
Re:Fasion? (Score:1)
Not useful anytime soon (Score:2)
But, when cable modems and all the rest really catch on, look out!
This patent is a good thing. (Score:1)
Intel obviously wants to sell more of their big server chips to ISPs, be it Xeons, Itaniums, or PPros back when they took out this patent. This basically forces the ISPs to buy their own computers.
Further, Intel does not want the burden of this distributed computing placed onto the user's computers. The ISPs want a reasonable turnaround, and probably aren't going to settle for idle time only. Most likely, this processing WILL cut into the user's experience. Intel has lots of other uses for that power in mind.
We can all guess what happens to the high end markets. They suffer a reduction in CPU performance that they will likely miss and possibly some jerky performance a few seconds at a time. The high end market is likely not going to settle for an ISP like this, but might get stuck paying much more for their internet connection as the typical web surfer.
The typical web surfer's processor power is also something that shouldn't be wasted. Intel is embracing cheap PCs. Software modems are perfect for intel here. When built into the motherboard, they only cost $10, and get all their processing power from the CPU. Next comes sound. The number of chips in a computer can be greatly reduced if it's capability is just handled by the CPU. Of course, this comes at a major CPU performance price. By doing this, computer prices are falling, and even low end users have a use for all that processing power. Again, the ISP's processor usage would cut into the experience of the user.
If users percieve a decrease in performance, computers all of a sudden become a worse deal, and people are less likely to buy them. Remember, there are still lots of people without computers. If people catch wind that ISPs are only bothering stealing processor power from the fastest processors, all of a sudden a low end machine and a high end machine look a lot more similar in terms of performance.
The way I see it, Intel just wants everyone to buy their own processor an to get what they paid for.
More things that are not patented (Score:1)
Each user had to activate the system when they left work, so the "predetermined time" was determined by the remote PC. It probably did react to "a start message from the central computer", but I'm not sure.
(*) If I told you the name of the company, I would lose all my moderation points and my karma would drop to zero. :-)
--
Prior art (Score:2)
PVM is a good cantidate for prior art. To quote the relevant section of the header comment:
* Revision 1.1 1996/09/23 23:44:26 pvmsrc* Initial revision
*
*
*/
Distributed computing in general is much older still. I was using it on '386es over Lantastic in the mid-80's. As far as I knew then, it was uncommon, but not exactly unheard of. Intel is nowhere near being the inventor of any of this.
The patent office was created to encourage inventors to publish in exchange for a brief (for the time) monopoly on the idea. That was fair compensation since they could just as easily keep it a trade secret and have it last nearly as long. It's primary purpose was to encourage progress by giving inventors a way to safely find an investor and keep inventions from being lost.
Today, patents like this one and the MANY other examples show that the USPTO is having EXACTLY the opposite effect on society. Small inventors can loose their patents by having large corperations steal it, and then throw money and lawyers at them until they go bankrupt (and loose by default) or sign a deal, and inventions that are already common knowledge are being bottled up under similar threats until they are well past their useful life.
Its absolutely shameful, and it has to end.
I'm through ranting now.
Not their idea... (Score:2)
I dunno, it seems like their implementation is pretty obvious to anybody who's read the article. Any patent experts care to educate on how this works?
Interestingly, I read the hotwired article when it came out, and now, nearly 3 years later, I've been thinking of doing some x-mas break hacking to throw together a generic implementation of this very thing! It would have been GPL'ed. Anyone know how different the implementation would have to be to avoid hassles?
This is a perfect example of the "chilling effect" of software patents. As a grad student, I definitely don't have the resources to defend myself against a legal attack from Intel!
J.
Re:Intel is right (not in the way we think) (Score:1)
--
Re:Intel is right (not in the way we think) (Score:1)
I don't expect the Europeans to understand this concept, but it's very real, and very important, to us Americans.
--
The abstract.. (Score:1)
A system executes computer processing tasks on a remote computer that communicates with a central computer. The remote computer receives a start message from the central computer. Based on the start message, the remote computer processes raw data to generate processed data, and stores the processed data. Finally, the remote computer sends a complete message to the central computer.
What the hell is this? I can actually understand what the abstact is talking about.. ;-) I thought all patents were ununderstandble jibberish that a normal person could not comprehend. Just read any Transmeta patent and you know exactly what I mean.
Re:Even More details (Score:1)
peter
Re:It's a defensive patent (Score:1)
I agree, but I'm not so sure Intel does much with their microprocessors anymore at all. It seems to me that other manufacturers make better x86 chips than they do (eg., AMD's Athalon vs. Intel's PIII). Moreover, I think the x86 architecture itself should have been dead and gone a decade ago; it's about damn time everyone started using a processor with a streamlined instruction set that runs cool enough not to need a fan. Processors with better architectures run at lower clock speeds, use less energy, and do the same job faster. That's the reality of microprocessor design now, and Intel engineers know it.
The problem is that Intel's current success is based on millions of lines of legacy code attributed to thousands of developers who write for the hundreds of millions of users who are still using obsolete technology. The operating system monopoly for IA processors basically controls what Intel can and can't do, and it always has. Why is there *any* reason for a Pentium to think it's an 8088 before you tell it otherwise? Because the predominant operating system of the day of the i386, MS-DOS, was written for the 8088, and porting it would have screwed up lots of software for MS, hence they would lose money. So Intel compensated to save their business.
No matter what Intel does with the IA64 or Merced, they will still be subject to the IA OS monopoly. I don't think they plan to do much real engineering. They just need to up clock speeds every once in a while and advertise to the masses. This patent is an attempt to sidestep distributed computing en masse as it would probably first occur; the VNA itself is old news, and Intel knows it. They want to hang on to all the dead weight that is IA for as long as possible and maximize profit. That's what this is about.
Re:Read the book... The Adolescence of P1 (Score:1)
Prior art: (Score:2)
(It's just a job queueing program, no need to slashdot these guys, but it does everything I saw in that patent minus the not-quite-so-willing participant bit.)
2. NQS/DQS/... (similar programs).
3. GNU Queue
So, let's see (it takes a while before we
even need the queue programs):
(a) receiving a start message at the remote computer from the central computer at a time that the central computer has predetermined the remote computer will be available for processing one of said first set of processing tasks;
crond plus rsh
(b) processing raw data related to said first set of processing tasks to generate processed data;
(c) storing the processed data; and
(d) sending a complete message to the central computer.
covered by rsh.
2. The method of claim 1, wherein the remote computer communicates with the central computer over a Transmission Control Protocol/Internet Protocol based network.
3. The method of claim 1, wherein the remote computer communicates with the central computer over a local area network.
Still rsh territory.
4. The method of claim 1, wherein the start message comprises a task and a storage location of the raw data, further comprising the step of retrieving the raw data.
An rsh command with an argument referring
to a file served over NFS/AFS?
5. The method of claim 1, wherein the start message comprises a task and the raw data, and the raw data is processed in accordance with the task.
An rsh command where the data are in ARGV?
(data isn't dammit, data are)
6. The method of claim 1, wherein the complete message comprises a storage location of the processed data.
So the program spits an output filename to STDOUT.
7. The method of claim 1, wherein the complete message comprises the processed data.
So the program spits its output to STDOUT,
which is linked over rsh to the server.
8. The method of claim 1, wherein a plurality of server computers communicate with the remote computer and the central computer, further comprising the step of retrieving the raw data from one of the plurality of server computers.
Oooh, now we need rsh and rcp (or NFS. Where I work we use a queue that calls ssh together with NFS all the time, using several servers).
9. The method of claim 1, wherein a plurality of server computers communicate with the remote computer and the central computer, step (c) comprising the step of: (c-1) storing the processed data on one of the plurality of server computers.
And we write to those NFS servers.
10. [snipped for brevity]
Putting it all together is easy with PBS.
So finally we come to need it.
A worm is a better prior example (Score:1)
Didn't Robert Morris provide a demonstration of this in 1988 or so? I don't suppose he had a chance to patent his techniques before they copped him.
One Word (Score:1)
How to make money: (Score:3)
2) Find a good idea (hint: watch the little numbers...good ideas have 4's and 5's!)
3) Run, don't walk, down to the patent office and file for a patent on that idea (be sure to use key patent terms like "process", "method" and "system" to make it as broad as possible).
4) Wait six to ten months for the patent office to stare cluelessly at the patent and eventually give up on "seeking prior art" (patent employees do not read SlashDot...lucky for you!)
5) Take fresh patent and go beat several companies over the head with it. As long as the amount you demand is less than the cost of defense, you'll get paid.
6) Repeat as necessary.
- JoeShmoe
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Re:Abstract . . one more thing (Score:1)
Beowulfs? (Score:1)
I.e. does anyone have the patent on Beowulf clustering? Theoretically, NASA should, since they built the first one, and ergo should be public domain... but one never knows, do they?
More details (Score:5)
I'll start with a quote from the patent itself:
Service providers, such as American Online.TM. ("AOL") and Compuserve increasingly must buy more powerful computers to service the additional members and the new content that is
constantly being updated. These service providers could save on computer costs if some of the computational requirements of their system could be serviced by remote personal computers
owned by private individuals and other independent entities who subscribe to the Internet provider services.
While it's true that this patent relates to distributed computing, that's not what this is about. What they've really got here is a patent on a method of stealing subscribers CPU cycles. Better check out the ToS when you sign up with an ISP cause, guess what? If this idea comes to be, you might just be required to do some of the processing for your ISP. Another quote to illustrate exactly what they're looking to process:
Computers, including
servers, must perform many tasks including the providing and generating of WEB pages, compressing Moving Pictures Expert Group ("MPEG") and Joint Photographic Expert Group
("JPEG") images, and serving up WEB pages.
Everyone who wants to serve AOL's web pages for them, raise your hand. Everyone who feels comfortable letting AOL initiate processes, raise your hand.
--GnrcMan--
Re:Beowulfs? (Score:1)
Business practice... (Score:1)
However... (Score:2)
--
Doesn't this sound like a networked cron? (Score:1)
Tivoli has been doing this since at least 1995. Since IBM owns Tivoli, this could get very interesting.
Re:Intel is right (not in the way we think) (Score:3)
In this situation the governemt arbitrarily picks one of the inventors as the sole beneficiary of the invention. Regardless of how the winner is chosen, it is inherently unfair. One of the goals of the patent system is to encourage inventors to publish their techniques so that the benefits of the invention can be shared by the whole world after a time. But in this case there is no need since it would have already been demonstrated that the invention can be independently reproduced.
Re:Can the subscriber control/prevent this? (Score:1)
And I will add... suppose someone else is doing
something that may be illegal - like downloading
child porn - and its getting decompressed on
*your* machine - with traces of it stored there...
ditto for copyright violations with music and
movie video content...
Another idea... (Score:2)
*********
Using networked remote computers to execute computer processing tasks,such tasks either done without charge, or at negotiated fee, at times signaled byeither the remote computer or the central computer.
Abstract
A system executes computer processing tasks on a remote computer that communicates with a central computer. The remote computer sends a message of availability to the central computer, or the central computer may send out a signal requesting processing time. Following the initial contact message, the central computer may negotiate a fee with the remote computer for processing time. If negotiation is successful, the central computer sends a task and possibly the associated raw data to the remote computer. The remote computer generates processed data, and stores the processed data. Finally, the remote computer sends a complete message to the central computer, and the processed data is returned to the central computer if that is required.
**********
This differs from both the Intel patent and the current implementation of distributed.net and Seti@Home. The premise here is that the processor time of my system will have value to some other organization, let's say my ISP, but that value is variable based upon both the loading on my system, my preferences and the load on the ISP. Should my system be busy serving content directly when the ISP needs my processor's services, a higher micro-payment will be negotiated. When, on the other hand, my computer signals availability, and the ISP's load is relatively low, a lower micro-payment would be appropriate.
Clearly, this is not just a LAN implementation, although it could be. It would imply persistent connections (such as this DSL line), and could possibly be a method of funding part of the expense of a persistent connection. It does allow me to offer my computer's services for free to some effort like Seti@Home, while selling my processor time to an ISP or other content provider, as appropriate.
Re:Prior art (Score:2)
I hadn't run across Hydra, but I'm not surprised to hear of it.
busting bad patents (Score:3)
This can be done still avoiding legal problems; I've been thinking about this and working on it for some time. If anyone is interested in helping, on the legal or technical end, contact me at the address above.
Re:Intel is right (not in the way we think) (Score:3)
That's all well and good until they try to pull a Unisys and claim royalties on a firmly entrenched standard. That is why frivolous patents are worrisome.
--GnrcMan--
It's a bit different (Score:2)
The remote computer receives a start message from the central computer.
But with Seti@home and the others, it's the remote computer that sends 'a start message' to the central computer, asking for data to be processed.
receiving a start message (...) at a time that the central computer has predetermined the remote computer will be available
Seti etc. don't run at predetermined times 'that the central computer has predetermined', but whenever the computer has some spare cycles.
I don't see this affecting any current efforts like those mentioned, but IANAL, so who knows how it could be twisted to mean if the lawyers/Intel so wanted
Re:Security Problems Here (Score:2)
Well, you could have other customers verify each other. The same problem occurs in networked games, people can forge calculation so client-side calculations have to verified by the central server or by another client. If the client base is large enough then it's difficult to do collaborative "cheating".
The other problem occurs is if the ISP's data is confidential AND the customer has to have access to it to process it. I did some work on this last year ago using "crypographic computing" to solve this problem. Basically you use the ideas of permutation and rotation to encrypt both the executing program and the required data. The client machine runs a VM that is capable of executing the encrypted instructions. Actually each instruction becomes a tree of instructions that calculate the same result. Trees from neighboring instructions can be woven together as long as data-dependencies are observed. Because you are using encrypted data, operations like "add/mul" won't work in their native form. They would have to decrypt 2 numbers before they could be multiplied together. This would give away the data, so instead you have to develop a cryptographic add/multiply function than can work with numbers in their encrypted form.
Advantages:
- the client doesn't know what he/she is executing
- the client doesn't know what data he/she is processing
- both the program and data can have MACs to prevent tampering of either.
Disadvantages:
- Memory usage can grow exponentially with the number of instructions executed
- Encryption / Decryption of program data is much more expensive than processing itself. Decryption time can grow by the number of instructions executed, because some operation cannot operate on encrypted data.
Although it's an interesting academic problem, it has little real world use because of speed and memory issues. But, it has lead me to my current project which is very useful.
It's a defensive patent (Score:2)
1) Make processors fast
2) Make processors faster
These two things are very important to the business model Intel currently enjoys. Deciding *when* #2 takes place is very important the bottom line.
Imagine, if you will, Microsoft adding code into the operating system that uses the distributed approach described in the patent. This, effectively puts the control over how fast a users computer appears to run for the user, in another companies hands (thus breaking rule #2). Plus, Intel has no interest on thousands of Merceds biting into the slice of server sales. So this is defensive. Take out the patent, sit on it, and defend it in court.
Re:busting bad patents - gnupatent.org lives (Score:3)
Re:Intel is right (not in the way we think) (Score:2)
Yes, there are some bad faith patent enforcements out there, with LZW & GIF being probably the best known example. But the real problem is surely not that they've got the patent, but that the legal system made the patent possible and so effectively required them to at least try for their own protection?
Standard disclaimer in this sort of thing - I'm British and have no idea what the patent situation is over here. And I accept that I'm discussing matters in a jurisdiction which doesn't affect me, to stop potential flames.
Greg