Note that the Ninth Circuit says that Sony doesn't have any patents that it says Connectix infringed.
What if Sony had used a patented subsystem to deny access unless certain patented code was copied? The patent could be so narrow that it was a "picture patent" but anyone who wanted access would have to make and use the specific patented material to be able to play the game.
There is no fair use defense in patent law. Would the case now come out the opposite way?
Something like this happened in the Sega-Accolade case. Accolade won the copyright round against Sega's access-denial system that used copyright law. But Accolade stumbled over Sega's second line of defense, a patented access-denial system. (Note that Connectix here has the same lawyer Accolade had in its case against Sega.)
Do you suppose Sega has a patent up its sleeve in this case too? (If so, it should have used it the first time around. Didn't it learn anything from the Sega-Accolade case?) If Sega has some patented scheme, will this victory turn out to be a charade?
More generally, is reverse engineering still vulnerable to being squashed? Just that you need a patent to squash it rather than a copyright?
One of the factors weighed in deciding whether to grant a preliminary injunction is whether the plaintiff seeking the p.i. is likely to succeed on the merits of its copyright infringement claim. In this case the district court held that Sony was likely to prevail because Connectix's downloading of code (reproduction of the work or intermediate copying) in order to reverse engineer was not privileged as fair use. The court of appeals reversed, saying that the intermediate copying for purposes of reverse engineering WAS fair use.
Given that ruling, it will be a steep uphill battle for Sony now to win the case - just about impossible.
The point of the proposed database law is that it is not a copyright law. Therefore it is not subject to rules like that saying "nobody can copyright a fact."
The idea of the law is to penalize unauthorized appropriation of the contents of the database, for example, by unauthorized large scale downloading. (How large? Not clear.)
Maybe if you use twice what you paid to download once, you're an infringer.
"...but what is to keep the holding company that keeps the patents in "trust" from getting gobbled up by one of the big fish? I mean, big fish could abuse legal process to bankrupt holding company, then buy the holding company off the auction block. This kind of thing has happened before...."
If you make the 'holding company' a non-profit corporation committed to educational purposes, assuming that the holding company can so qualify, state laws generally prohibit the assets of the chairity (i.e., holding company) from being gobbled up and exploited by a for-profit corporation (your Big Fish Corp.).
In the alternative, the holding company could assign any patent it desired to enforce. It would assign it to 'New Corp.,' which would be set up solely for the purpose of owning and exploiting the particular patent. (New Corp. would be a for-profit corporation or LLC [= GmbH].) If New Corp. tanked, your Big Fish Corp. would be able to gobble up only the assets of New Corp., not those of the holding company. The assets of the holding company would be insulated, assuming that all of the proper corporation law formalities were observed.
So, I don't think that the problem concerning you is likely to happen, if proper steps are taken to avoid it.
It would be unrealistic for a company like Rio to think that it would cost it less than $1.5 million to defend the suit, and $10 million is not unlikely.
You suggest responding to the suit in a perfunctory manner, but the risk to Rio would be to lose a bet-your-company threat. The damage to Rio could be hundreds of $million if it screwed up in defending itself. That would lead to a stockholder suit against the management for running the company in an idiotic manner. No sane managemnent will risk that.
The cost of defending a patent is substantial because the risk of failing to defend in a prudent manner is even more substantial.
Many good and bad ideas have emerged from the 'Open Defensive Patents' discussion. This comment tries to build on that discussion to develop a practical program for dealing with the IP 'Enclosure Movement' now threatening software, Inet, and e-commerce technology.
The comment is directed to a proposal to create a 'bore from within' organization using the GNU and Open Software - copyleft models, adapted for patents. I should say, as a preliminary matter, that this wheel will need to be reinvented without any help from Richard Stallman and the copyleft movement. They are too committed to opposing software patents in any and all forms, and in any and all circumstances, to have any interest in a patent utilizing proposal, even one to bore from within. Nonetheless, their idea for dealing with software copyright problems is basically sound and just needs some tweaking to adapt it to a patent environment. The major increment to their approach proposed here is to use revenue gained from software patents to finance opposition to software patents, so that just as GNU's not UNIX, software patenting's no friend of software patenting.
Much of the most profitable patenting in the last 20 years has followed what might be called the Jerry Lemelson model (Jerry being the late genius who was a master of this technique). You anticipate which way technology is going, describe it in general terms in a patent, claim it broadly, and wait to see whether you guessed right. If you did, then you hold out your hand and demand payment from those who followed the moving wave of technology and actually invented and developed an actual product that works in accordance with what you predicted from your armchair.
If feasible, you modify your patent application while it's pending to describe more specifically the inventions of others that occur after you filed your original patent application. That makes it even easier for you to hold up their corporate employers for royalty ransom. Or, to look at it from another vantage point, it makes it even more attractive for their corporate employers to do a deal with you (the patent owner) in order to clobber their competitors more effectively by using your patent against them under an exclusive license.
Who are in a better position to be software technology prophets or soothsayers than the slashdot nerd community? Collectively, a group of them could tie up the future development of software technology tighter than the army of corporate patenters now trying to enclose the 'Net and e-commerce using the 'Net. The idea would be not simply to mulct corporate America but rather to use the proceeds to finance efforts to cut back on software patents, perhaps by lobbying, bringing test cases (perhaps asserting your own patents in court in so outrageous a manner that the courts in reaction set legal precedents limiting software patents in general), financing a prior art depository, establishing a copyleft-like licensing program, or spreading information to the public. Certainly, you use the proceeds of patenting software to fund the underlying program. (Everything here is recursive.)
Now, the first thing to dissipate is the $10,000 cost myth. (The second thing to forget about is the self-addressed letter. It won't work. The third thing to forget about is publication. It does not usually work, either, and at best works only after a year has passed. A rival can 'swear back' of your publication and thus avoid it as a prior art reference, even if it identically discloses everything that the rival invents, which is unlikely in itself.) But let's get back to cost.
Yes, it costs more than $10K in attorney fees to get a regular software patent ($20K-30K is more realistic - of that, government fees amount to $2K or less while fees for attorney services amount to the rest). And if you multiply $10K by the number of applications a technology soothsayer needs to file in order to cover the waterfront sufficiently, the resulting dollar amount makes the project infeasible (unless you teach yourself how to write your own patent applications). But this arithmetic ignores the recently created 'Provisional Patent Application' (PPA) of the US Patent Office (PTO). It is now possible to file a do-it-yourself, skeleton patent application simply by:
preparing a good technical writeup of an invention (you are not obliged to write claims or otherwise go through the effort of preparing a regular patent application),
filling out a one-page cover sheet (PTO Form 16, downloadable at http://www.uspto.gov/web/forms/sb0016.pdf),
stapling the form and writeup together,
writing the PTO a check for $75,
filling out another one page form (PTO Form 9, downloadable at http://www.uspto.gov/web/forms/sb0009.pdf) that says you are a sole inventor or a small business organization or a non-profit corporation (these latter two use similar PTO forms) and that thus makes you eligible for the reduced $75 fee, and
sending the whole package to the PTO.
If you file a PPA, you have one year to decide whether to file a full, regular patent application. During that year, the PPA gives you first dibs over anybody else who tries to patent the same invention. You can decide whether the invention is worth pursuing by patent. If not, do nothing and after the year the PTO just tubes the PPA and nothing further occurs. If yes, you file a regular patent application and get the benefit of the earlier filing date. There is no significant downside, particularly if the alternative was doing nothing.
The point is, at $75 a pop you can be much more casual about how many applications you file. Moreover, the one year dibs period gives you the opportunity to make a better judgment whether it is worth spending $10K or more to prepare a real patent application. (Keep in mind that you cannot sue anyone for infringing a PPA. A PPA gives you nothing, legally, but first dibs against later patent applications for the same thing, IF you get a regular patent based on the PPA. The real benefit of the PPA, in this context, is that it enables you to cast a wider net, evaluate the market, and regroup within a year - so that you avoid spending large sums on too many regular patent applications.)
But let's look at the $10K issue a little more closely. First of all, if the project actually works, the cost of future patent applications should be financed by the revenue from licensing earlier ones. It should be possible to find some patent lawyers willing to take a chance on such a scheme. (By that I mean preparing 'on spec' proper patent applications for the proposed organization for those PPAs that seem worth pursuing, when the one year dibs period is almost over. Their incentive would be to be rewarded from future revenues.) There might even be idealists or cranks among the ranks of patent lawyers who were willing to donate some time pro bono publico. I can think of some like that.
Enforcement of patents is a separate issue from their procurement. When an organization has a really valuable patent, and it is being infringed in a way and to an extent that substantial damages can be collected, the organization can find patent lawyers who will take the case on a contingent fee basis. I have at times found such lawyers. Furthermore, suppose that the proposed organization is willing to grant exclusive commercial licenses (these exclude the copyleft-equivalent or GPL-equivalent field from the license). Then the licensee will fund the enforcement if doing so is to its competitive advantage. Certainly, that obligation can be bargained for as a term of the exclusive license.
Finally, the objection has been raised in various forms that you cannot preemptively patent all future software advances. Of course you cannot. But it does not follow that you cannot make a big dent in the system. That would be a worthy goal.
The idea of a GPL or copyleft for software patents is challenging. But to accomplish anything, the project must be on a technically sound basis. That means technically adept from a software technology standpoint and also technically correct from a patent law standpoint.
This may be an idea whose time has come. Are any of you willing to join in forming a SIGSWPAT or other forum to discuss this further with a view toward establishing an organization to implement these ideas? Is there an existing organization into which this could properly be incorporated? Would anybody want to start a Web site or include a site for this as part of an existing, related one?
Slashdot Mirror
What if Sony had used a patented subsystem to deny access unless certain patented code was copied? The patent could be so narrow that it was a "picture patent" but anyone who wanted access would have to make and use the specific patented material to be able to play the game.
There is no fair use defense in patent law. Would the case now come out the opposite way?
Something like this happened in the Sega-Accolade case. Accolade won the copyright round against Sega's access-denial system that used copyright law. But Accolade stumbled over Sega's second line of defense, a patented access-denial system. (Note that Connectix here has the same lawyer Accolade had in its case against Sega.)
Do you suppose Sega has a patent up its sleeve in this case too? (If so, it should have used it the first time around. Didn't it learn anything from the Sega-Accolade case?) If Sega has some patented scheme, will this victory turn out to be a charade?
More generally, is reverse engineering still vulnerable to being squashed? Just that you need a patent to squash it rather than a copyright?
Given that ruling, it will be a steep uphill battle for Sony now to win the case - just about impossible.
The idea of the law is to penalize unauthorized appropriation of the contents of the database, for example, by unauthorized large scale downloading. (How large? Not clear.)
Maybe if you use twice what you paid to download once, you're an infringer.
If you make the 'holding company' a non-profit corporation committed to educational purposes, assuming that the holding company can so qualify, state laws generally prohibit the assets of the chairity (i.e., holding company) from being gobbled up and exploited by a for-profit corporation (your Big Fish Corp.).
In the alternative, the holding company could assign any patent it desired to enforce. It would assign it to 'New Corp.,' which would be set up solely for the purpose of owning and exploiting the particular patent. (New Corp. would be a for-profit corporation or LLC [= GmbH].) If New Corp. tanked, your Big Fish Corp. would be able to gobble up only the assets of New Corp., not those of the holding company. The assets of the holding company would be insulated, assuming that all of the proper corporation law formalities were observed.
So, I don't think that the problem concerning you is likely to happen, if proper steps are taken to avoid it.
You suggest responding to the suit in a perfunctory manner, but the risk to Rio would be to lose a bet-your-company threat. The damage to Rio could be hundreds of $million if it screwed up in defending itself. That would lead to a stockholder suit against the management for running the company in an idiotic manner. No sane managemnent will risk that.
The cost of defending a patent is substantial because the risk of failing to defend in a prudent manner is even more substantial.
Many good and bad ideas have emerged from the 'Open Defensive Patents' discussion. This comment tries to build on that discussion to develop a practical program for dealing with the IP 'Enclosure Movement' now threatening software, Inet, and e-commerce technology.
The comment is directed to a proposal to create a 'bore from within' organization using the GNU and Open Software - copyleft models, adapted for patents. I should say, as a preliminary matter, that this wheel will need to be reinvented without any help from Richard Stallman and the copyleft movement. They are too committed to opposing software patents in any and all forms, and in any and all circumstances, to have any interest in a patent utilizing proposal, even one to bore from within. Nonetheless, their idea for dealing with software copyright problems is basically sound and just needs some tweaking to adapt it to a patent environment. The major increment to their approach proposed here is to use revenue gained from software patents to finance opposition to software patents, so that just as GNU's not UNIX, software patenting's no friend of software patenting.
Much of the most profitable patenting in the last 20 years has followed what might be called the Jerry Lemelson model (Jerry being the late genius who was a master of this technique). You anticipate which way technology is going, describe it in general terms in a patent, claim it broadly, and wait to see whether you guessed right. If you did, then you hold out your hand and demand payment from those who followed the moving wave of technology and actually invented and developed an actual product that works in accordance with what you predicted from your armchair.
If feasible, you modify your patent application while it's pending to describe more specifically the inventions of others that occur after you filed your original patent application. That makes it even easier for you to hold up their corporate employers for royalty ransom. Or, to look at it from another vantage point, it makes it even more attractive for their corporate employers to do a deal with you (the patent owner) in order to clobber their competitors more effectively by using your patent against them under an exclusive license.
Who are in a better position to be software technology prophets or soothsayers than the slashdot nerd community? Collectively, a group of them could tie up the future development of software technology tighter than the army of corporate patenters now trying to enclose the 'Net and e-commerce using the 'Net. The idea would be not simply to mulct corporate America but rather to use the proceeds to finance efforts to cut back on software patents, perhaps by lobbying, bringing test cases (perhaps asserting your own patents in court in so outrageous a manner that the courts in reaction set legal precedents limiting software patents in general), financing a prior art depository, establishing a copyleft-like licensing program, or spreading information to the public. Certainly, you use the proceeds of patenting software to fund the underlying program. (Everything here is recursive.)
Now, the first thing to dissipate is the $10,000 cost myth. (The second thing to forget about is the self-addressed letter. It won't work. The third thing to forget about is publication. It does not usually work, either, and at best works only after a year has passed. A rival can 'swear back' of your publication and thus avoid it as a prior art reference, even if it identically discloses everything that the rival invents, which is unlikely in itself.) But let's get back to cost.
Yes, it costs more than $10K in attorney fees to get a regular software patent ($20K-30K is more realistic - of that, government fees amount to $2K or less while fees for attorney services amount to the rest). And if you multiply $10K by the number of applications a technology soothsayer needs to file in order to cover the waterfront sufficiently, the resulting dollar amount makes the project infeasible (unless you teach yourself how to write your own patent applications). But this arithmetic ignores the recently created 'Provisional Patent Application' (PPA) of the US Patent Office (PTO). It is now possible to file a do-it-yourself, skeleton patent application simply by:
If you file a PPA, you have one year to decide whether to file a full, regular patent application. During that year, the PPA gives you first dibs over anybody else who tries to patent the same invention. You can decide whether the invention is worth pursuing by patent. If not, do nothing and after the year the PTO just tubes the PPA and nothing further occurs. If yes, you file a regular patent application and get the benefit of the earlier filing date. There is no significant downside, particularly if the alternative was doing nothing.
The point is, at $75 a pop you can be much more casual about how many applications you file. Moreover, the one year dibs period gives you the opportunity to make a better judgment whether it is worth spending $10K or more to prepare a real patent application. (Keep in mind that you cannot sue anyone for infringing a PPA. A PPA gives you nothing, legally, but first dibs against later patent applications for the same thing, IF you get a regular patent based on the PPA. The real benefit of the PPA, in this context, is that it enables you to cast a wider net, evaluate the market, and regroup within a year - so that you avoid spending large sums on too many regular patent applications.)
But let's look at the $10K issue a little more closely. First of all, if the project actually works, the cost of future patent applications should be financed by the revenue from licensing earlier ones. It should be possible to find some patent lawyers willing to take a chance on such a scheme. (By that I mean preparing 'on spec' proper patent applications for the proposed organization for those PPAs that seem worth pursuing, when the one year dibs period is almost over. Their incentive would be to be rewarded from future revenues.) There might even be idealists or cranks among the ranks of patent lawyers who were willing to donate some time pro bono publico. I can think of some like that.
Enforcement of patents is a separate issue from their procurement. When an organization has a really valuable patent, and it is being infringed in a way and to an extent that substantial damages can be collected, the organization can find patent lawyers who will take the case on a contingent fee basis. I have at times found such lawyers. Furthermore, suppose that the proposed organization is willing to grant exclusive commercial licenses (these exclude the copyleft-equivalent or GPL-equivalent field from the license). Then the licensee will fund the enforcement if doing so is to its competitive advantage. Certainly, that obligation can be bargained for as a term of the exclusive license.
Finally, the objection has been raised in various forms that you cannot preemptively patent all future software advances. Of course you cannot. But it does not follow that you cannot make a big dent in the system. That would be a worthy goal.
The idea of a GPL or copyleft for software patents is challenging. But to accomplish anything, the project must be on a technically sound basis. That means technically adept from a software technology standpoint and also technically correct from a patent law standpoint.
This may be an idea whose time has come. Are any of you willing to join in forming a SIGSWPAT or other forum to discuss this further with a view toward establishing an organization to implement these ideas? Is there an existing organization into which this could properly be incorporated? Would anybody want to start a Web site or include a site for this as part of an existing, related one?
rstern@computer.org
http://www.law.gwu.edu/facweb/claw/rhs1.htm