Weekly Edition
Return to the Press page
| |||||||||||||
|
| |||||||||||||
Patent problems plague Linux (ZDNet)
Dan Ravicher justifies his Linux kernel patent survey in this ZDNet column.
"A study that quantifies the potential risk eliminates the guessing game by supplying users with specific information they can use to determine whether they are sufficiently prepared. Studying a threat does not create the risk; it only makes that risk easier to more accurately address. You would not accuse a weatherman of spreading fear for profit by warning of a 25 percent chance of showers and saying 'tune in later for more information.'"
(Log in to post comments)
Patent problems plague Linux (ZDNet) Posted Aug 5, 2004 22:30 UTC (Thu) by JoeBuck (subscriber, #2330) [Link] But in this case, the weatherman is saying "give me a lot of money and I'll chase the storm away for you".
Patent problems plague Linux (ZDNet) Posted Aug 6, 2004 0:44 UTC (Fri) by piman (subscriber, #8957) [Link] It's more like an insurance company having an weatherman on staff, and hawking more insurance when he predicts disasters. It still isn't entirely on the up-and-up in my opinion, but it's not as bad as you make it out to be.
Patent problems plague Linux (ZDNet) Posted Aug 6, 2004 8:56 UTC (Fri) by hppnq (subscriber, #14462) [Link] It's more like having someone around who will start blowing in the opposite direction until s/he is out of breath. You could do this yourself, but it would not be so expensive.
Patent problems plague Linux (ZDNet) Posted Aug 6, 2004 15:17 UTC (Fri) by mmarsh (subscriber, #17029) [Link] I think a better analogy would be a weatherman who says, "There's a 25% chance of rain. Would you like to buy an umbrella?" Or perhaps "Would you like to buy flood insurance?"
Patent problems plague Linux (ZDNet) Posted Aug 6, 2004 17:00 UTC (Fri) by piman (subscriber, #8957) [Link] Your analogy is much better. :)
Patent problems plague Linux (ZDNet) Posted Aug 20, 2004 6:08 UTC (Fri) by xoddam (subscriber, #2322) [Link] Nah, it's Noah selling tickets for seats on the Ark.
How many affect Microsoft Windows? Posted Aug 5, 2004 22:41 UTC (Thu) by dwheeler (guest, #1216) [Link] I would expect at least the same number of patents, if not many more, to affect Microsoft Windows. The patent office allows patents of prior art and obvious concepts, even though it's required to not do so.
How many affect Microsoft Windows? Posted Aug 6, 2004 0:39 UTC (Fri) by DaveK (subscriber, #2531) [Link] The real point is liability, Since both systems claim to offer a similar set of end usertools (the reality may be different) it is reasonable to assume that many patents that affect one may affect the other. One is an offering from a large corporation based in Redmond, most of its users would expect said corporation to indemnify them aginst any patent violations - it is a commercial product after all - or, more to the point, to have acquired such licenses as would be required to use such patented technologies in a commercial product. Ultimatly they have one of the biggest bank accounts there is, and some of the most feared lawyers. (The EULA may say something much different - buyer beware!). Small users of GNU/Linux systems - without large bank accounts, or high priced lawyers - it seems may be liable for violations in code that some one else wrote, and yet another distributed. This is where the real fear, uncertainty and doubt lies - especially if said patent is of questionable validity.
Who is liable? Posted Aug 6, 2004 11:20 UTC (Fri) by phip (subscriber, #1715) [Link] Who is liable for patent infringment?
For example, if company X patents a new kind of toaster, and company Y makes and sells a competing toaster that infringes the patent, who is breaking the law?
Company Y?
If publishing a description of a patented invention is not illegal, why should it be illegal to publish a description of a patented algorithm or software method, whether in English, pseudocode, or source code form?
Cheers,
Who is liable? Posted Aug 6, 2004 12:10 UTC (Fri) by utoddl (subscriber, #1232) [Link] If publishing a description of a patented invention is not illegal, why should it be illegal to publish a description of a patented algorithm or software method, whether in English, pseudocode, or source code form?Which is precisely why software should not be patentable. Fundamentally, software is math. Granted, it math that can do things, but it's still math.
Who is liable? Posted Aug 6, 2004 13:27 UTC (Fri) by DaveK (subscriber, #2531) [Link] Seems like we're getting back to the 'Is source code a language and thus likespeech?' question once again. Can I invent a language, patent it, then charge everyone else to use it? (obviously you'd all carry on using your own languages, but...). If so, and given the record of many patent offices around thw world, then it I can forsee a situation in which I might patent 'a method of verbal and written communication for the peoples of England and other linguistically compatible countries' despite the obvious existance of several hundred years of 'prior art'. But I digress... PS. By the way that'll be GBP 10 please for anyone who wishes to respond in ENGLISH!
How many affect Microsoft Windows? Posted Aug 6, 2004 15:22 UTC (Fri) by vmole (guest, #111) [Link] One is an offering from a large corporation based in Redmond, most of its users would expect said corporation to indemnify them aginst any patent violations - it is a commercial product after all One might expect that, but it turns out not to be the case. A few years ago, it was discovered that MS had knowingly included technology patented by another company (Timeline) in SQL Server, had licensed said technology for the purposes of distributing SQL Server, but had explicitly NOT licensed the rights for SQL Server developers to use the technology in custom applications. All this would be okay, except MS also failed to notify developers that the patent existed. Oops. Given the state of software patents in the US, I'm pretty sure that any significant software system violates some patents. I have some confidence that the Linux kernel does not knowingly violate significant patents (i.e. those aren't just landgrab claims of 20 year old common knowledge.) I've no such confidence in commercial products.
Patent problems plague Linux (ZDNet) Posted Aug 6, 2004 9:15 UTC (Fri) by hppnq (subscriber, #14462) [Link] There are a couple of things that I don't understand.1) How was OSRM's research conducted? Surely it's no trivial task to map the entire Linux kernel against all potentially applicable patents? Yet, Ravicher claims I could do this myself -- I sure hope that the research involved more than a quick glance at Grokline and the patent database. 2) If I subscribe to one of OSRM's offerings, isn't that just as bad as looking up potentially infringed patents myself anyway? In court, I could hardly claim unawareness of any patent infringement if I pay someone $BIG_NUMBER for legal assistance in a patent case, right?
Patent problems plague Linux (ZDNet) Posted Aug 6, 2004 10:41 UTC (Fri) by DaveK (subscriber, #2531) [Link] IANAL, but my understanding is that taking up OSRM's offer does not mean that youknow you are violating any patents, but that since the kernel is too big to check yourself, you insure yourself against the risk of the possibility of violating one or more patent or patents unknown, in the same way that buying house insurance doesn't mean that you know you're going to be burgled, or even that you expect to be, you simply insure yourself against the possibility (risk) that you might. On the other hand, doing the research yourself would reveal the truth of whether there are any violations or not, and there may be none, but if you find even a single violation, and continue anyway you would be doing so knowing specifically which patent(s) you were violating. It is suggested that that would be a wilful violation. The difference is between the grey area of 'maybe' and the black and white of 'knowing for sure'. Who will be willing to take the chance of lining themselves up against a wilful infringement lawsuit versus the possibility of proving the kernel clean. Would such a suit follow 'automatically', Would the violation then automatically become wilful, could it be licensed, how much would that cost? This is the kind of uncertainty on which the FUDmeisters thrive!
Patent problems plague Linux (ZDNet) Posted Aug 6, 2004 14:00 UTC (Fri) by hppnq (subscriber, #14462) [Link] First off: ITANAL. ;-)The burglary analogy is not correct, IMHO, for several reasons. The insurance company doesn't pay if I set my own house on fire. Also, I cannot get an insurance that covers the case where I accidentally burgle someone, can I? ('Cause that would be a better analogy, no?) No wonder OSRM warns their indemnification program is not an insurance, even though other public statements seem to suggest otherwise. (If I am wrong, please let me know, I am not a troll. ;-)
Patent problems plague Linux (ZDNet) Posted Aug 6, 2004 14:02 UTC (Fri) by hppnq (subscriber, #14462) [Link] Sorry: ... if I set my own house on fire ...What am I thinking?! This should be, of course: ... if I steal my own stuff ... Pfff...
Patent problems plague Linux (ZDNet) Posted Aug 6, 2004 14:53 UTC (Fri) by DaveK (subscriber, #2531) [Link] I see what you mean, notwithstanding that stealing your own stuff is like violatingyour own patent, you can't, since you already own it and any rights it entails, and if you try to claim compensation, I believe its called fraud. Perhaps a better analogy is motor insurance (excluding the fact that in this country at least, it is a legal requirement) whereby you indemnify yourself at a minimum against claims for damages and losses caused to others by your actions.
Patent problems plague Linux (ZDNet) Posted Aug 6, 2004 15:46 UTC (Fri) by hppnq (subscriber, #14462) [Link] Well, I can think of a company that would accuse me of stealing my own code, but basically you're right. ;-) It all boils down to the really awful patent law (at least in the States) and the fact that you're absolutely powerless if you haven't got deep financial pockets.I think that offering "insurance" against patent infringement sends the wrong message to the world, and I blame OSRM for not thinking about that carefully enough. (Yeah, that means I am assuming they really mean to be a good citizen.)
Patent problems plague Linux (ZDNet) Posted Aug 7, 2004 20:33 UTC (Sat) by mmarq (guest, #2332) [Link] " you simply insure yourself against the possibility (risk) that you might.On the other hand, doing the research yourself would reveal the truth of whether there are any violations or not "
ok... but i cant help thinking that exactly that is the stupidiest thing that a single 'independent' developer can do... If someone cames after me saying that (hypothetical) "my" code infringes on some of their patents then they have to reveal (point to) themselfs the *Correct Patent Application*... the only risk i take is that they only show that application in court... because "i" can get a unfair tryal!!...
THAT IS WHERE THE FUD IS...remenber SCO
I might be that some 'independent' developer gets waxed to pay damages,... but that could only happen, by the few judicial system knowledge i have, in context of 'unfairness',... judges helping the powers that are,... as it still can happen with the SCO vs IBM, althought very unlikely because SCO really is caught in its lie, ... because the most than obvious outcame is that "i" replace the algoritm or drop the code entirely, nothing else...
Also in this Patent romance there is clearly powers trying to *INTIMIDATE* and scatter free software developers, as was obviously in the SCO case, and or taking advantages by *control power*!,.. and in those cases OSRM *insurance* only makes sense to get proper judicial support,... but only if its free of charge for registered developers, because its supported by the commercial players(the gainers) of the Linux industry... otherwise if "i" have to pay to develop then "my" code can't really be free of charge, can it ?... (can see greed executives licking their fingers)
But "i" can *STILL* get, althought unlikely, an unfair tryal either "i" have a OSRM indemnification or not... so to "me" OSRM, in the present incarnation, is in most respects void of substance.
A really, very good Patent Busting Organisation, is what is most required... it will *INTIMIDATE BACK* those that might want to prosecute "me"... of course we wont see IBM or HP backing one of those... urgh!?...
Patent problems plague Linux (ZDNet) Posted Aug 7, 2004 22:07 UTC (Sat) by DaveK (subscriber, #2531) [Link] For the average Kernel developer, the risk would appear to be less than that to businesses. If they are taken to court, AFAIK (but IANAL) they simply need to convince the court that they were unaware of the patent, and agree to cease and desist, and rewrite the contested code in a non-infringing manner. I would have thought that corporate patent holders are less likely to hound individulas as the cash rewards simply aren't there - how many could settle for millions - and it also makes bad PR However, proving - or even settling such a suit would be great FUD fodder for those who oppose Free Software - even if the outcome was cease and desist, and a small rewrite. On the other hand, if you are a business especially a non IT related one which doesn't write software, but you simply use it to power your servers, you may not have the wherewithall to deal with rewriting or cutting out contested sections of code. Chasing your business will probably have better financial rewards, and better PR than chasing the individual.
Even a cease and desist order served on such a business would hurt -cease and desist using the Linux Kernel - OK, so what then,
Whichever you settle for, its more great FUD fodder. These are the people that OSRM wish to attract. At this point in time I am uncertain how much of a 'for profit' organisation OSRM wishes to be, but if it wishes to be 'for profit' then it is in its commercial interest to drag the patent uncertainty out, and fuel the FUD whilst offering indemnities against violations at a price. Simply using the cash to patent bust any dodgy patents which could be used against GNU/Linux would quickly start to stifle their own cash flow as they reduce the percieved 'risk'.
Patent problems plague Linux (ZDNet) Posted Aug 9, 2004 19:54 UTC (Mon) by mmarq (guest, #2332) [Link] "For the average Kernel developer, the risk would appear to be less than that to businesses..."
That is my opinion too.
"I would have thought that corporate patent holders are less likely to hound individulas as the cash rewards simply aren't there..."
Well SCO 'was' a complete bluff, but they had threaten to prosecute single developers as well as users,... , my point is that if those corporate patent holders think that is usefull to *INTIMIDATE* a whole developer community, than they would not hesitate a second,... remember Sklyarov ?... he was the victim to send a *intimidating* message about DCMA power ( http://www.eff.net/IP/DMCA/US_v_Elcomsoft/ )... and precisely because of this, i belive that an arrest and ridiculous prosecution of some single developer, about patents, is something that could not be put entirely aside, and in that case OSRM plans couldn't do nothing, be it 'free of charge' or with 'profit' intention.
So in my opinion EFF deserves all atention in his patent busting effort, ( http://www.eff.org/patent/ ), because that is a mechanism that could intimidate back potencial 'greed for FUD' ridiculous patent prosecuters,... even to business that are more likely victims of potencial 'extortion' prosecutions, and for those i dont see how OSRM could be anything more than a band-aid.
Patent problems plague the USofA and maybe the EU Posted Aug 6, 2004 10:01 UTC (Fri) by hummassa (subscriber, #307) [Link] The problem with software patents are, in summary:1. Software patents are absurd. 2. Even in the rare case where they could be considered less than absurd, they are granted by the USPTO at least without proper research. 3. There isn't ONE software patent granted that I know of that fulfills the requirements of novelty and/or non-obviousness. Thanks $DEITY software patents are strictly _forbidden_ by BR law. And I'll go to the Congress to lobby myself if any Congressperson tries to change it.
Patent problems plague the USofA and maybe the EU Posted Aug 6, 2004 13:11 UTC (Fri) by utoddl (subscriber, #1232) [Link] Thanks $DEITY software patents are strictly _forbidden_ by BR law.That's good news indeed, but what do you mean by "BR law?" Basic Reasoning? Business Regulations? Benefits/Rewards? Brains Running? Help me out here...
BR = Brazil? Posted Aug 6, 2004 13:55 UTC (Fri) by dwheeler (guest, #1216) [Link] I suspect "BR" means "Brazil" since the top-level domain .br is Brazil.
Patent problems plague the USofA and maybe the EU Posted Aug 6, 2004 15:30 UTC (Fri) by vmole (guest, #111) [Link] 3. There isn't ONE software patent granted that I know of that fulfills the requirements of novelty and/or non-obviousness. How about the RSA public key algorihm? Or the LZW compression algorithm? It think it can be reasonably argued that those meet the intention of the US patent system.
Patent problems plague the USofA and maybe the EU Posted Aug 6, 2004 17:07 UTC (Fri) by piman (subscriber, #8957) [Link] Some people (myself included) are of the opinion any mathematics fails the "novel and non-obvious" test. One train of logic is that since such properties are (formally, computationally, i.e. "automatically") derivable from axiomatic theories, nothing new was discovered, only an *explanation* of an *effect* was (i.e. a proof from some axiom set).
The "short" version is -- you can't invent math, you can only discover it, and patents only cover inventions.
Patent problems plague the USofA and maybe the EU Posted Aug 6, 2004 17:36 UTC (Fri) by AJWM (guest, #15888) [Link] "patents only cover inventions"Alas, that hasn't been the case for a long time. I'm mindful in particular of numerous biological patents and patents for various uses of substances (enzymes, proteins, various other biochemicals) found in nature. Technically I think they weasel-word around it by patenting a particular use of said substance, and I suppose arguably Edison & co. didn't so much invent the carbon-filament light bulb as discover (through exhaustive research) that carbon made a good light bulb filament. (The general idea that an electrically heated wire might make a good light source, if it would last long enough, having been around already.) I guess that's where "obviousness" comes into it. Most software patents have been pretty obvious cases of "apply new technology X to well-understood process Y", where they haven't been blatent grabs of prior art.
|
|
||||||||||||