Wow, can I get your filing fees? Then, I'll go ahead and file your patent, as a small entity, for a whooping $500. That's selling *10* devices, not 100. And if you can't sell 10 of them, then trust me, it's not worth patenting.
Reality is, the Patent Office does try to make this affordable for small inventors. Almost everyone can swing $500 for filing fees.
Now if you want to hire a patent attorney to write your application, well then you're looking at closer to $10,000. But, if you don't have the money, but have the time, I highly recommend Patent It Yourself by Nolo Press. It'll help you write your patent for a grand total of $40. So, with filing fees, Assignments, and postage, the exercise might cost you $600. That's 12 of your devices at $50 profit. Not too shabby.
The Court actually simply decided that there was no case in controversy. Edelman sued, before he did anything for a declaratory judgement giving him permission to reverse engineer. The court said they wouldn't decide the case, because it's not sure that N2H2 would sue, or that Edelman would do anything. The only time a court will make a decision if nothing has happened yet (like here, about future potential lawsuits) is if a fundamental right is at question. While copyright Fair Use is important, it is clearly not a fundamental right. So, chill. And wait for someone to actually be sued under the DMCA for reverse engineering for research purposes. If the Court then holds that there is no right to reverse engineer anymore (in spite of Sega v. Accolade.
This type of billion dollar settlement is a typical artifact of high-technology management styles for conflicts. I doubt that anyone over at Sony that caused this problem will take any fallout from it.
More likely, this conflict slowly progressed where no product manager wanted to take the sales hit for moving away from the protected PlayStation brand. After all, the loss of brand recognition would be felt immediately in the sales figures that measure the success of a product manager and the current settlement is only felt long after those product managers have been promoted.
A good dictator could have fixed this problem before PlayStation was such an entrenched brand.
Wrt to the copyright/defamation/etc. suit, California has a wonderful form for this called SLAPP (strategic lawsuits against public participation.) SLAPP lawsuits are basically lawsuits that try to shut down criticism, and California has a specific statute dealing with this (as do a number of other states). If you win a SLAPP claim, you can automatically get attorneys fees and costs.
It's probably way to late to bring this up, but, yes, Java can and probably will beat C++ performance. There are two basic reasons:
First, Java byte code operation is specified completely. This means that is a calling tree never calls any sort of native method implementation, the optimizer can play 'fun with graphs'. This gives the optimizer the oportunity to use tricks like non-blocking I/O, and aggressive inlining. Most importantly, it allows register optimization across method calls. In C/C++, optimizing across method calls runs afoul of odd and random pointers.
Second, Java can reoptimize based on current runtime usage pattern. Imagine a banking transaction program that processes a million checks and deposits. Cashing a check involves looking up the account, withdrawing the money, applying overdraft, etc. In reality, 99.99% are correct without overdrafts. The Java optimizer rewrites the code as 'look up account or panic, then subtract check amount from account leaving a positive balance or panic'. The program is shorter and can make better use of registers in the usual case, and the 'panic' means to unwind the optimized code back to the Java byte code. This beats C++ when you do a million checks followed by a million deposits.
There really is a much larger theoritical cieling for Java optimization. It comes with the cost of stepping back from the hardware.
Always this argument...
on
The Wireless City
·
· Score: 5, Insightful
I am tired of the usual diatribe from security people that bandwidth is this great outlet for danger. Any system could be used for DOS, DDOS, Spam, spoofing, hacking onto other machines in the park, secret Chinese spy deals, and more. Get over it.
Some would liken IP connectivity to a printing press, and argue the company providing the press must watch each item printed against copyrighted, subversive, or pornographic works. Others would argue it is like electricity, a utility that is provided fairly cheaply after the initial wiring is installed, and need not be charged for at all for small amounts. The few who see it as a wilderness, full of abuse and crime and desparados checking for weakness tend to sell computer security services.
The OS wants to store something in RAM, trys to allocate RAM and fails.
The OS decides to swap out some memory based on an analysis of page usage statistics, so it grabs the handle to the swap file.
Then, the OS streams the offending pages through file I/O subsystem, worrying about waits and updating page counters and such. This includes dragging all the data over a busy system bus.
The receiving device stores it in RAM.
Cool! What operating system could aspire to such levels of efficiency?
Here is a decent writeup from The Register. The accusation is that "results could only be accessed via a 40 character ID code." Now whether this is an extended address, or a password is unclear. It also notes that there are a couple of other firms that have also accused Reuters of hacking into their systems to get early access to reports.
Actually, this does raise an interesting question. If a page is put on the web that cannot be spidered, and cannot be reached from any publicly available page, can we assume that anyone who accesses that page has some sort of unauthorized information? I have never heard of hackers systematically trying IP addresses for content. And it is in fact likely that Reuters got the info from an employee... in violation of the employment agreement.
This should be a fascinating case, and not nearly as easy as the writeup makes it seem.
I used to work in the group at Sun that promoted Solaris on Intel. There is a core group of morons that is very good at dodging layoffs, signing large contracts that don't deliver revenue, and bitching to Scott McNealy that Solaris on Intel really isn't dead. This leads to all sorts of pathological decisions.
Solaris is an operating system, and a pretty good one. Solaris generally has oddly optimized drivers for large boxes that make it very useful for large sites. Also, Solaris is the vehicle for pushing Sun's special talent; networking more processors more effectively. Solaris on SPARC works well.
Solaris on Intel is the bastard child of an unresolved angst over controlling the client desktop. Sun has never figured out that it has a special weakness against making a decent client. Sun has never turned around to the niche market and embraced Apple clients, or PC clients, or anyone else. The wierd waffling on Solaris on Intel is a sickness from a lack of decision.
The problem will not go away until the group is fired. Deal with it.
Why are you surprised? The same ratings scheme applies on television. If it's just people getting shot, well then the kiddies can watch no problem. If it's nudity, that's a problem. If it's MALE nudity, then it belongs on the triple-X channel only. Americans are a puritan bunch... and the puritans did approve of killing people (remember, they burned witches for entertainment value)... just not of sex.
This is a typical case of frivolous lawsuits. There are, unsurprisingly, rules against this. If anyone actually fights it out in court, I expect that the Court would in fact grant attorneys fees to Google or the other defendants/counter-plaintiffs, as they have done in other cases. There is precedent for this which is binding in California.
In fact, some courts would consider this a SLAPP case (Strategic Litigation Against Public Participation), trying to restrict the free speech rights of the critics of the web site. Anti-SLAPP suits are great because attorneys fees are automatically granted if you win.
The problem is the up-front cost of litigation. Unfortunately, there is nothing that can be done until at least one case goes through the courts. BUT, in the past, some people have been prevented by judgement from suing anyone else, because they were "court abusers." Hopefully this will happen to Mr. Novak.
Oh, and if Slashdot is sued, I'd be happy to help with the defense. I'm no litigator, but writing the reply brief to this would be entertainment, not work.
Nope, I went to school in the US, and we took technical drafting. And then we had a class on CAD/CAM drafting tools. But neither of these two is engineering. They're simply tools we can use to solve problems.
An NT admin is a SOFTWARE engineer. Thi sis quite different from a HARDWARE engineer. Software folks don't need to know how a circuit works, they just need to know what the inputs and outputs are. Both groups are needed... and they shouldn't be confused. Hardcore hardware folks still know how to draw a circuit diagram... and "real electrical engineers" know analog too.
The reason we'll never be replaced by Engineer-In-A-Box is the same reason why computers will not take over design. They're not creative. They cannot find a new solution to an old problem. They can optimize the known solutions, but that's it. Creativity is uniquely human (well, AI, hypothetically could do this... but then AI is only hypothetical for now.) And it's this type of creativity that makes everything happen.
Actually, given the current economy, very few patent lawyers are willing to do the work for a percentage. And even fewer inventors, most of whom believe that their invention is worth something, are willing to offer it. The deal we (patent attorneys) are usually offered is stock ownership. But after the dot-com bust very few of us are willing to do that.
Most patent attorneys charge by the hour, so you can significantly reduce the time you need to buy by preparing most of the work yourself. If you know an agent who does good work, that's another good way of reducing cost (average agent charges $100-150/hour compared to the attorney $200-400/hour) But remember that quality is paramount. So you're better of getting a better lawyer for fewer hours.
Go directly to Nolo.com and give them your money. They're absolutely wonderful. They also do a pretty good line in landlord-tenant law, and a book about getting out of a speeding ticket. Oh yes, and they have a fabulous law encyclopedia to boot.
Bullshit. I AM a lawyer, and I can tell you that the only way you will lose a patent is if you try to assert it (enforce it against someone) and you lose your case. Otherwise, if you keep paying your maintenance fees, you'll have a handy piece of paper you can wave at potential investors, nothing more. IF you chose to enforce your patent, you may lose it. But patent enforcement is justly called the game of kings... average cost is about $2 Million.
You do not have to enforce your patent in order for it to remain valid. Ditto for copyright. The only IP right you may have to enforce is trademark. Trademark is used to identify the source of your goods (it's an Apple computer -- Apple is the trademark). If others use your trademark long enough so it is no longer associated with you (like Kleenex or Xerox, which are both fighting a losing battle), you may lose your trademark. But you will NEVER lose your patent this way.
As to finding help, Patent it Yourself is fabulous (buy it directly from nolo.com). If you can afford an attorney, look for one who has written patents in your area (if you're doing prior art searches, look at the "attorney or agent" line on the patent. If they're doing a good job, give them a call.
Thalia
Miss the point; Free Wireless because of economics
on
War Car Offers Wi-Fi
·
· Score: 4, Insightful
Why rant about should there be free wireless; rant about there will be free wireless. It just makes sense.
The secondary profits of providing free wireless access are more than the primary profits of commerical wireless access. A free site tends to have about $500 in fixed costs, $100 per month in variable costs for bandwith at the site, and, well, nothing for billing and back end systems. A commercial site has these costs, plus about $50 per site to keep the back end and user accounts running, plus $10 per user per month in technical support costs. The technical support costs are the killer.
I chatted with a marketing droid for our local coffee shop's provider. It took him a bit to understand that $20 per month really was the maximum I would pay, that I wouldn't pay more for lighting fast access or access to any of the hundred of so sites nationwide or for special software so that I didn't need to keep a browser window minimized. I have access elsewhere; it's only worth $20 per month. Too bad they have technical support from people trying to get their 802.11(b) card to work under Windows 95/98/98SE/2K/ME/NT/XP/XPpro with NetScape/IE4/IT5/....
That provider tanked, someone else took over a month later. The coffee house noticed a significant drop during the slow hours when geeks drank coffee and typed on computers. The other coffee house is not as good but had free wireless and I spend $120 in the month I was there.
Hold it you say, $120 per month for wireless? Well, about $80 per month in profit. I drink a lot of coffee. It's worth more to the coffee shop to provide the wireless point, provide no support, and sell coffee than it will ever be worth to the wireless service provider to deal with billing and customer support for $20 per month.
Wireless wants to be free, just because it makes more money that way.
We used to argue this in the computer science lab at college. Can the human mind gain visualization skills in four dimensional geometry? We came up with the following interesting answers:
1. It's hard. We never see four diminsions. The brain would keep wanting to make one dimension some known continuim such as time, a color sequence, tone, or intensity. Only after this intermediate step would you get a true four dimensional geometry in your head.
2. You would need to have a true 3D display. Current rendering of three dimensional pictures flattened onto simple two dimensional screens would never work. Imagine using a laser pointer as a point source, and imagine that you had never seen a three dimensional object; now draw a three dimensional picture of a pick-up truck using the laser pointer. At the time, we were trying to get a simple three dimensional output, like <a href="http://www.stereographics.com/frames/frame-p rod.html">Crystal Eyes</a>. Now there are liquid crystal on silicon solutions that are much cleaner, if not cheaper.
Capitalism works on competing businesses raising capital and dividing markets. The resulting competition gets companies, and individuals, moving forward.
Consumer broadband languished for years until both cable and twisted pair solutions were available. This means you can buy residential broadband from either your cable company or from your phone company. Or you could try Covad, Speakeasy, Wifi Metro, or other services.
As any good market, the top broadband player will make most of the money, the second player will make a reasonable return, the third player loses a little money and hopes someone pulls an Enron, and everyone else has a dream. The market isn't yet mature, and there are business uncertainties about marking the boundry of the market and dividing costs. This is how people get into arguments on owning the loop to the phone company's Central Office, or the home owner owning the right to move the drop cable to a cable overbuild, or the right to dump the ISP side of an internet connection and pay only for physical routing.
It's also how people argue about costs. One way of looking at wireless internet nodes and household private networks is that they all usurp service and place undue burden on the provider; an open network is theft. The other way is to view the service as providing a utility, like power, to a residence; an open node is like running an extension cord out to the front yard. The market will sort this out.
SBC nee Pacific Bell doesn't mind if you run a local network or open node, and has a long history of not worrying about extra phone extensions. Cable companies have a long history of worrying about cable descramblers, people using cable for two televisions in the house, and people using cable for public display. The terms and enforcements follow the corporate histories.
Who is correct? Let the dollars decide.
Chasm/Thaila
P.s.: Looking for a Silicon Valley SE to sell products to developers? Email me at jobhunt@truegift.com
It sounds like a reasonable idea. Basically, it's just a fancy way of saying "cross licensing." In other words, if you license me your patents under these terms, I will license you my patents under the same terms. This is done on a daily basis by large corporations.
The only possible problem is that you will likely never see revenue from the patent, from other sources. Most licensees will insist on a "favored nation clause" that basically says that they won't be charged more than any other licensee. Clearly in this case, that would not be possible. So, I doubt you'd find many takers for the license. Remember that getting a patent is relatively expensive. If you hire a patent attorney, it'll run between $10-15K, just for the US. If you do it on your own, expect to spend 100 hours if you're inexperienced (about 40 with experience), and approximately $2000 for a patent. If you want international protection, multiply by 10 or 20. So, if you'll never see revenue from it, you're unlikely to get funding based on it, and you can't trade it for protection against other "evil" guys, what's the point of spending the money? You'd be better of using either Statutory Invention Registrations (cheaper, no examination, simply publishes the idea & documents your date), or publishing your ideas.
No, actually the courts are the arbiter of whether this case has merit.
The PTO gets its chunk of change (~$2000) for examining the application, and seeing if the patents they have make it obvious. It is the applicants' duty to disclose all references they're aware of.
FOIA (Freedom of Information Act) suits force an agency to divulge how they came to a particular decision. In fact, the PTO's decision process is public. You, or anyone else, can order a copy fo the File Wrapper of the patent, including every interchange between the PTO and the Applicant. It's about 25 cents per page.
You could make that assumption, but the income differentials are sufficiently large that it makes no sense. (Something on the order of 300X more money comes in from applications than from reexam requests.) A better theory is that the PTO gets a lot more applications than they can handle. Unfortunately a large percentage of their income in taken into the "general" fund, and they simply cannot hire enough competent examienrs. (They're hiring, if you feel masochistic.)
Oh yes, and they pay folks with CS degrees ~$35K to start, to live in DC. Odds that they'll get competent examiners at that salary are between slim and none.
If they have sued others under this patent, then you may also want to contact those companies (American Airlines et al.) They probably did a prior art search, and a thorough evaluation of the patent scope. You may be able to obtain copies of their work.
Slashdot Mirror
Wow, can I get your filing fees? Then, I'll go ahead and file your patent, as a small entity, for a whooping $500. That's selling *10* devices, not 100. And if you can't sell 10 of them, then trust me, it's not worth patenting.
Reality is, the Patent Office does try to make this affordable for small inventors. Almost everyone can swing $500 for filing fees.
Now if you want to hire a patent attorney to write your application, well then you're looking at closer to $10,000. But, if you don't have the money, but have the time, I highly recommend Patent It Yourself by Nolo Press. It'll help you write your patent for a grand total of $40. So, with filing fees, Assignments, and postage, the exercise might cost you $600. That's 12 of your devices at $50 profit. Not too shabby.
Thalia, why yes, I am a patent attorney
The Court actually simply decided that there was no case in controversy. Edelman sued, before he did anything for a declaratory judgement giving him permission to reverse engineer. The court said they wouldn't decide the case, because it's not sure that N2H2 would sue, or that Edelman would do anything. The only time a court will make a decision if nothing has happened yet (like here, about future potential lawsuits) is if a fundamental right is at question. While copyright Fair Use is important, it is clearly not a fundamental right. So, chill. And wait for someone to actually be sued under the DMCA for reverse engineering for research purposes. If the Court then holds that there is no right to reverse engineer anymore (in spite of Sega v. Accolade.
Thalia
This type of billion dollar settlement is a typical artifact of high-technology management
styles for conflicts. I doubt that anyone over
at Sony that caused this problem will take any
fallout from it.
More likely, this conflict slowly progressed where no product manager wanted to take the sales hit for moving away from the protected PlayStation brand. After all, the loss of brand recognition would be felt immediately in the sales figures that measure the success of a product manager and the current settlement is only felt long after those product managers have been promoted.
A good dictator could have fixed this problem before PlayStation was such an entrenched brand.
Wrt to the copyright/defamation/etc. suit, California has a wonderful form for this called SLAPP (strategic lawsuits against public participation.) SLAPP lawsuits are basically lawsuits that try to shut down criticism, and California has a specific statute dealing with this (as do a number of other states). If you win a SLAPP claim, you can automatically get attorneys fees and costs.
Thalia
It's probably way to late to bring this up, but, yes, Java can and probably will beat C++ performance. There are two basic reasons:
First, Java byte code operation is specified completely. This means that is a calling tree never calls any sort of native method implementation, the optimizer can play 'fun with graphs'. This gives the optimizer the oportunity to use tricks like non-blocking I/O, and aggressive inlining. Most importantly, it allows register optimization across method calls. In C/C++, optimizing across method calls runs afoul of odd and random pointers.
Second, Java can reoptimize based on current runtime usage pattern. Imagine a banking transaction program that processes a million checks and deposits. Cashing a check involves looking up the account, withdrawing the money, applying overdraft, etc. In reality, 99.99% are correct without overdrafts. The Java optimizer rewrites the code as 'look up account or panic, then subtract check amount from account leaving a positive balance or panic'. The program is shorter and can make better use of registers in the usual case, and the 'panic' means to unwind the optimized code back to the Java byte code. This beats C++ when you do a million checks followed by a million deposits.
There really is a much larger theoritical cieling for Java optimization. It comes with the cost of stepping back from the hardware.
I am tired of the usual diatribe from security people that bandwidth is this great outlet for danger. Any system could be used for DOS, DDOS, Spam, spoofing, hacking onto other machines in the park, secret Chinese spy deals, and more. Get over it.
Some would liken IP connectivity to a printing press, and argue the company providing the press must watch each item printed against copyrighted, subversive, or pornographic works. Others would argue it is like electricity, a utility that is provided fairly cheaply after the initial wiring is installed, and need not be charged for at all for small amounts. The few who see it as a wilderness, full of abuse and crime and desparados checking for weakness tend to sell computer security services.
Deja vu all over again. But it's been four and a half months... so we probably forgot the whole discussion already.
Thalia
- The OS wants to store something in RAM, trys to allocate RAM and fails.
- The OS decides to swap out some memory based on an analysis of page usage statistics, so it grabs the handle to the swap file.
- Then, the OS streams the offending pages through file I/O subsystem, worrying about waits and updating page counters and such. This includes dragging all the data over a busy system bus.
- The receiving device stores it in RAM.
Cool! What operating system could aspire to such levels of efficiency?Actually, this does raise an interesting question. If a page is put on the web that cannot be spidered, and cannot be reached from any publicly available page, can we assume that anyone who accesses that page has some sort of unauthorized information? I have never heard of hackers systematically trying IP addresses for content. And it is in fact likely that Reuters got the info from an employee... in violation of the employment agreement.
This should be a fascinating case, and not nearly as easy as the writeup makes it seem.
Thalia
I used to work in the group at Sun that promoted Solaris on Intel. There is a core group of morons that is very good at dodging layoffs, signing large contracts that don't deliver revenue, and bitching to Scott McNealy that Solaris on Intel really isn't dead. This leads to all sorts of pathological decisions.
Solaris is an operating system, and a pretty good one. Solaris generally has oddly optimized drivers for large boxes that make it very useful for large sites. Also, Solaris is the vehicle for pushing Sun's special talent; networking more processors more effectively. Solaris on SPARC works well.
Solaris on Intel is the bastard child of an unresolved angst over controlling the client desktop. Sun has never figured out that it has a special weakness against making a decent client. Sun has never turned around to the niche market and embraced Apple clients, or PC clients, or anyone else. The wierd waffling on Solaris on Intel is a sickness from a lack of decision.
The problem will not go away until the group is fired. Deal with it.
Why are you surprised? The same ratings scheme applies on television. If it's just people getting shot, well then the kiddies can watch no problem. If it's nudity, that's a problem. If it's MALE nudity, then it belongs on the triple-X channel only. Americans are a puritan bunch... and the puritans did approve of killing people (remember, they burned witches for entertainment value)... just not of sex.
Thalia
As always google cache rocks.
Thalia
This is a typical case of frivolous lawsuits. There are, unsurprisingly, rules against this. If anyone actually fights it out in court, I expect that the Court would in fact grant attorneys fees to Google or the other defendants/counter-plaintiffs, as they have done in other cases. There is precedent for this which is binding in California.
In fact, some courts would consider this a SLAPP case (Strategic Litigation Against Public Participation), trying to restrict the free speech rights of the critics of the web site. Anti-SLAPP suits are great because attorneys fees are automatically granted if you win.
The problem is the up-front cost of litigation. Unfortunately, there is nothing that can be done until at least one case goes through the courts. BUT, in the past, some people have been prevented by judgement from suing anyone else, because they were "court abusers." Hopefully this will happen to Mr. Novak.
Oh, and if Slashdot is sued, I'd be happy to help with the defense. I'm no litigator, but writing the reply brief to this would be entertainment, not work.
Thalia
Nope, I went to school in the US, and we took technical drafting. And then we had a class on CAD/CAM drafting tools. But neither of these two is engineering. They're simply tools we can use to solve problems.
An NT admin is a SOFTWARE engineer. Thi sis quite different from a HARDWARE engineer. Software folks don't need to know how a circuit works, they just need to know what the inputs and outputs are. Both groups are needed... and they shouldn't be confused. Hardcore hardware folks still know how to draw a circuit diagram... and "real electrical engineers" know analog too.
The reason we'll never be replaced by Engineer-In-A-Box is the same reason why computers will not take over design. They're not creative. They cannot find a new solution to an old problem. They can optimize the known solutions, but that's it. Creativity is uniquely human (well, AI, hypothetically could do this... but then AI is only hypothetical for now.) And it's this type of creativity that makes everything happen.
Thalia
Actually, given the current economy, very few patent lawyers are willing to do the work for a percentage. And even fewer inventors, most of whom believe that their invention is worth something, are willing to offer it. The deal we (patent attorneys) are usually offered is stock ownership. But after the dot-com bust very few of us are willing to do that.
Most patent attorneys charge by the hour, so you can significantly reduce the time you need to buy by preparing most of the work yourself. If you know an agent who does good work, that's another good way of reducing cost (average agent charges $100-150/hour compared to the attorney $200-400/hour) But remember that quality is paramount. So you're better of getting a better lawyer for fewer hours.
Thalia
If you actually want a paying job, instead of going out on your own, I strongly suggest an CS or EE degree. Not much else is getting hired these days.
Thalia
Go directly to Nolo.com and give them your money. They're absolutely wonderful. They also do a pretty good line in landlord-tenant law, and a book about getting out of a speeding ticket. Oh yes, and they have a fabulous law encyclopedia to boot.
Thalia
Bullshit. I AM a lawyer, and I can tell you that the only way you will lose a patent is if you try to assert it (enforce it against someone) and you lose your case. Otherwise, if you keep paying your maintenance fees, you'll have a handy piece of paper you can wave at potential investors, nothing more. IF you chose to enforce your patent, you may lose it. But patent enforcement is justly called the game of kings... average cost is about $2 Million.
You do not have to enforce your patent in order for it to remain valid. Ditto for copyright. The only IP right you may have to enforce is trademark. Trademark is used to identify the source of your goods (it's an Apple computer -- Apple is the trademark). If others use your trademark long enough so it is no longer associated with you (like Kleenex or Xerox, which are both fighting a losing battle), you may lose your trademark. But you will NEVER lose your patent this way.
As to finding help, Patent it Yourself is fabulous (buy it directly from nolo.com). If you can afford an attorney, look for one who has written patents in your area (if you're doing prior art searches, look at the "attorney or agent" line on the patent. If they're doing a good job, give them a call.
Thalia
The secondary profits of providing free wireless access are more than the primary profits of commerical wireless access. A free site tends to have about $500 in fixed costs, $100 per month in variable costs for bandwith at the site, and, well, nothing for billing and back end systems. A commercial site has these costs, plus about $50 per site to keep the back end and user accounts running, plus $10 per user per month in technical support costs. The technical support costs are the killer.
I chatted with a marketing droid for our local coffee shop's provider. It took him a bit to understand that $20 per month really was the maximum I would pay, that I wouldn't pay more for lighting fast access or access to any of the hundred of so sites nationwide or for special software so that I didn't need to keep a browser window minimized. I have access elsewhere; it's only worth $20 per month. Too bad they have technical support from people trying to get their 802.11(b) card to work under Windows 95/98/98SE/2K/ME/NT/XP/XPpro with NetScape/IE4/IT5/....
That provider tanked, someone else took over a month later. The coffee house noticed a significant drop during the slow hours when geeks drank coffee and typed on computers. The other coffee house is not as good but had free wireless and I spend $120 in the month I was there.
Hold it you say, $120 per month for wireless? Well, about $80 per month in profit. I drink a lot of coffee. It's worth more to the coffee shop to provide the wireless point, provide no support, and sell coffee than it will ever be worth to the wireless service provider to deal with billing and customer support for $20 per month.
Wireless wants to be free, just because it makes more money that way.
We used to argue this in the computer science lab at college. Can the human mind gain visualization skills in four dimensional geometry? We came up with the following interesting answers:
p rod.html">Crystal Eyes</a>. Now there are liquid crystal on silicon solutions that are much cleaner, if not cheaper.
1. It's hard. We never see four diminsions. The brain would keep wanting to make one dimension some known continuim such as time, a color sequence, tone, or intensity. Only after this intermediate step would you get a true four dimensional geometry in your head.
2. You would need to have a true 3D display. Current rendering of three dimensional pictures flattened onto simple two dimensional screens would never work. Imagine using a laser pointer as a point source, and imagine that you had never seen a three dimensional object; now draw a three dimensional picture of a pick-up truck using the laser pointer. At the time, we were trying to get a simple three dimensional output, like <a href="http://www.stereographics.com/frames/frame-
We were students once, and poor.
Capitalism works on competing businesses raising capital and dividing markets. The resulting competition gets companies, and individuals, moving forward.
Consumer broadband languished for years until both cable and twisted pair solutions were available. This means you can buy residential broadband from either your cable company or from your phone company. Or you could try Covad, Speakeasy, Wifi Metro, or other services.
As any good market, the top broadband player will make most of the money, the second player will make a reasonable return, the third player loses a little money and hopes someone pulls an Enron, and everyone else has a dream. The market isn't yet mature, and there are business uncertainties about marking the boundry of the market and dividing costs. This is how people get into arguments on owning the loop to the phone company's Central Office, or the home owner owning the right to move the drop cable to a cable overbuild, or the right to dump the ISP side of an internet connection and pay only for physical routing.
It's also how people argue about costs. One way of looking at wireless internet nodes and household private networks is that they all usurp service and place undue burden on the provider; an open network is theft. The other way is to view the service as providing a utility, like power, to a residence; an open node is like running an extension cord out to the front yard. The market will sort this out.
SBC nee Pacific Bell doesn't mind if you run a local network or open node, and has a long history of not worrying about extra phone extensions. Cable companies have a long history of worrying about cable descramblers, people using cable for two televisions in the house, and people using cable for public display. The terms and enforcements follow the corporate histories.
Who is correct? Let the dollars decide.
Chasm/Thaila
P.s.: Looking for a Silicon Valley SE to sell products to developers? Email me at jobhunt@truegift.com
It sounds like a reasonable idea. Basically, it's just a fancy way of saying "cross licensing." In other words, if you license me your patents under these terms, I will license you my patents under the same terms. This is done on a daily basis by large corporations.
The only possible problem is that you will likely never see revenue from the patent, from other sources. Most licensees will insist on a "favored nation clause" that basically says that they won't be charged more than any other licensee. Clearly in this case, that would not be possible. So, I doubt you'd find many takers for the license. Remember that getting a patent is relatively expensive. If you hire a patent attorney, it'll run between $10-15K, just for the US. If you do it on your own, expect to spend 100 hours if you're inexperienced (about 40 with experience), and approximately $2000 for a patent. If you want international protection, multiply by 10 or 20. So, if you'll never see revenue from it, you're unlikely to get funding based on it, and you can't trade it for protection against other "evil" guys, what's the point of spending the money? You'd be better of using either Statutory Invention Registrations (cheaper, no examination, simply publishes the idea & documents your date), or publishing your ideas.
Thalia
No, actually the courts are the arbiter of whether this case has merit.
The PTO gets its chunk of change (~$2000) for examining the application, and seeing if the patents they have make it obvious. It is the applicants' duty to disclose all references they're aware of.
FOIA (Freedom of Information Act) suits force an agency to divulge how they came to a particular decision. In fact, the PTO's decision process is public. You, or anyone else, can order a copy fo the File Wrapper of the patent, including every interchange between the PTO and the Applicant. It's about 25 cents per page.
Thalia
You could make that assumption, but the income differentials are sufficiently large that it makes no sense. (Something on the order of 300X more money comes in from applications than from reexam requests.) A better theory is that the PTO gets a lot more applications than they can handle. Unfortunately a large percentage of their income in taken into the "general" fund, and they simply cannot hire enough competent examienrs. (They're hiring, if you feel masochistic.)
Oh yes, and they pay folks with CS degrees ~$35K to start, to live in DC. Odds that they'll get competent examiners at that salary are between slim and none.
Thalia
If they have sued others under this patent, then you may also want to contact those companies (American Airlines et al.) They probably did a prior art search, and a thorough evaluation of the patent scope. You may be able to obtain copies of their work.
Thalia