Rather than try to use "men with guns" to protect inventors, how about we level the playing field and require people to actually create items and find a way to market them before they're knocked off?
Of course "men with guns" are not only involved with just patents, but with other types of IP laws as well. So if "men with guns involved" means somethings needs to be done away with that means:
No Patents. as per your speculation.
No Copyrights. Feel free to copy those bits. $20 Mil to make a movie. With just the help a friend at the development lab, anyone can undercut your price without the megabucks investment
No Trademarks. Worked hard on your brand? Tought dookie. Someone else can call thier product the same thing and milk your reputation. Its not that they are counterfiting, because there is no such thing. They just like the name. Or they can claim its a now a generic term. Who is to stop them? Men with guns?
I don't see any public interest achievements in NASA
Thats because you probably have not looked.
For example, you could see how NASA research can benefit you if you are handicapped or as you grow older by reading Robert Heinlein's non-fiction essay "Spinoff", based on his testimony before comittee in Congress. Its found in the collection _Expanded Universe_.
My understanding of the GPL was that basically you couldn't plan to distribute just an executable binary. That the person who the software is distributed too can also get the source.
Now if the government is producing code based on GPL products, then typically they will be the only customer. The only one the code would be distributed to would be the NZ goverment itself. So the government would be the only customer that could ask for the source code.
Its going to worry about asking itself?
Just don't ask. Take the position that the product is an in-house development, and is never distributed outside of 'in-house'. No outside distribution, no GPL problem.
About the only thing I can think of that might propose a problem would be if the government produced standard programs for third parties. Like standardized tax preperation programs in lieu of distributing paper forms.
But as that would make sense, I don't think we have to worry much about a government doing it.:)
Joking aside, if the government doesn't go into the business of distributing software outside itself, this issue is a no-starter.
There seems to be no backup way of opening this lock.
What if the firmware crashes in it. If the space it locks does not have an alternate means of entry, they couldn't even pull the hinges off to get at the lock. They'd need a fire axe or a jaws of life to get in again.
All that would have to be done to wedge the door is jam the sonic 'knock' channel. It seems that it would be fairly easy to make a device that sent pseudo random knocks into the door on a constant basis. That would overlap if any knock code sent into the door by a real code box.
The door would not open.
Hide it on the inside of the door, with a timer to activate it, and then you have a veto on that door ever unlocking again from the outside.
Wow, its almost as if the ISPs were trying to say that people would have to pay more if they wanted thier packets routed with a high standard for delivery time.
Where have I heard that
recently?
The better choice is futures options. For a few thousand dollars you can buy an option whose value will be 1000 times (oil price - $100). If oil got to $200 your few thousand dollars would turn into $100,000...
But what if after inflation (due to rapid increase in oil prices and economic chaos) $100,000 is basically 6 months of electricity?
Now that there is a way to make sound with network activity, the next step will be someone will write an app that plays actual music by generating varying rates of activity.
Sort of like they used to do with lineprinters. Print patterns that make the mechanism make certain tones.
Picture a sysop sitting in the NOC late at night , when all of a sudden, the sonic network montior starts emitting Darth Vader's theme.
Well, not for network operations, but for control systems in a space ship.
Walter Jon Williams had a story in which the starship control panel emitted sound patterns related to the functioning of the ship. If the characteristics of the soundscape changed that alerted the pilot that something had altered and be checked.
The end result is an incoming bittorrent tagged stream and an outgoing line noise stream with some token placeholder tags. That prevents it from being birectional noise, but how does it achieve the stated goal of evading the ISP's bandwidth throttle?
Firstly, there should not be any obvious bittorrent tagged stream, because encryption is being used. And there could be more than one data stream.
But to answer your question, because when the phrase "traffic shaping" is used that is not a general throttle on all bandwidth, but a throttle on specific sort of connection. If you make it difficult to detect that type of connection, you make it difficult to trigger the throttling.
The only thing that could be detected would be basic bandwidth use. And that would present a danger of many more false positives than traffic shaping specificly targeting P2P. There is an increased risk that such false positives would updet the entire customer base, not just the P2P users.
Fact of the matter is that P2P traffic looks very different from normal web browsing on the grandest of scales. A P2P user transfers many gigabytes, both upstream AND downstream to many other low-bandwidth users -- most of which have IPs that are trivially attributed to DSL/cable providers, making it clear that they are users, not businesses (no major websites are hosted by SBC DSL users, comcast cable customers, etc. etc...) You can probably profile a P2P user by raw upload:download ratio alone, with 90% accuracy. The average web user downloads many times more than they send upstream.
That's brahm's point.
But the supposition is that the ISP will 'shape' the traffic. If they do it by identifying customer who are apparently using P2P, but not being able to identify which TCP/IP sessions are the one carrying the traffic, they'll be forced to reduce bandwidth on every connection. That's so ham-handed its really difficult to call it traffic shaping at all. It basically giving a customer not even the suggestion of what they advertised to him as service.
Maybe the ISPs would not care about driving away P2P high bandwidth users. Maybe they wouldn't be concerned about false advertising claims. But I'd think they's worry about the reputation that will come from the screams of 'false positive' customers who will witness that the ISPs service is worse than dialup even though they did nothing out of the oridinary.
The more agile the target, the more computation it costs to track it. I can't see it as being practical to have firmware in the cable modem/dsl slam or in a cable headend that can track statistics on every unique session passing through and squash P2P sessions. Espcially if individual TCP connections complete in a matter of seconds.
The idea is to make shaping cost more than its worth, so that the ISPs can't selectively welch on the services they advertised that they provide.
How would that prevent it from being lots of bidirectional line noise?
If they are sniffing based on IP/port tuples, using one pipe one way and other for the other obfuscates the bidirectional part. For noise, you'd have to salt it with something that doesn't look random. Maybe just a "HTML" string at the beginning.
Modifiable, of course, to keep it a moving target.:)
Actually on further thought, I think it might be more difficult to track a transaction that kept rolling onto a new TCP/IP session every few seconds. Let the peers do the handshaking of the new connection in the background, and when its open for business, move the data off the old stream and onto the new on.
"...a wire protocol which transfers a lot of data bidirectionally and consistently looks like line noise with no header is only marginally more difficult to identify then one which uses fixed ports."
Sounds like a call to camoflage the traffic as several pipes between peers. Not just one tcp/ip connection, but several, with a jitter function to pick which pipe is used at the moment so it does not look consistant
You speak about perl/tk as a rejected option, because you need it small and easy to use.
Since you are probably using ActivePerl, have you considered using the perlcc command to make a single EXE file of your application? Its easy to install as any other stand alone EXE file.
'Small' might be another story. Its certainly smaller than a full perl install on your target system. The EXE will be bigger than a meg certainly. Large compared to true compiled application, larger, and a bit slower to start up, as it unpacks its required modules from its internal package.
But a few meg isn't really that big of a deal these days.
I can make my IP change every hour, by changing the MAC address.
Why was this guy mod'd troll?
He expressed a technical point that its not difficult to get a different IP address. That leads to the obvious conclusion that banning an IP address is not the same as banning a person from editing.
Slashdot Mirror
Of course "men with guns" are not only involved with just patents, but with other types of IP laws as well. So if "men with guns involved" means somethings needs to be done away with that means:
Is the Lynx browser now officially against the law in the UK?
Thats because you probably have not looked.
For example, you could see how NASA research can benefit you if you are handicapped or as you grow older by reading Robert Heinlein's non-fiction essay "Spinoff", based on his testimony before comittee in Congress. Its found in the collection _Expanded Universe_.
You can read some of it via Amazon.com here.
It starts about page 501.
For a moment there I was worried I'd have to have an EKG reader before I could use Ghostscript
I had it approximately right! :)
Maybe that's what NASA really stands for:
Needs Another Succesive Approximation.
My understanding of the GPL was that basically you couldn't plan to distribute just an executable binary. That the person who the software is distributed too can also get the source.
:)
Now if the government is producing code based on GPL products, then typically they will be the only customer. The only one the code would be distributed to would be the NZ goverment itself. So the government would be the only customer that could ask for the source code.
Its going to worry about asking itself?
Just don't ask. Take the position that the product is an in-house development, and is never distributed outside of 'in-house'. No outside distribution, no GPL problem.
About the only thing I can think of that might propose a problem would be if the government produced standard programs for third parties. Like standardized tax preperation programs in lieu of distributing paper forms.
But as that would make sense, I don't think we have to worry much about a government doing it.
Joking aside, if the government doesn't go into the business of distributing software outside itself, this issue is a no-starter.
There seems to be no backup way of opening this lock.
What if the firmware crashes in it. If the space it locks does not have an alternate means of entry, they couldn't even pull the hinges off to get at the lock. They'd need a fire axe or a jaws of life to get in again.
All that would have to be done to wedge the door is jam the sonic 'knock' channel. It seems that it would be fairly easy to make a device that sent pseudo random knocks into the door on a constant basis. That would overlap if any knock code sent into the door by a real code box.
The door would not open.
Hide it on the inside of the door, with a timer to activate it, and then you have a veto on that door ever unlocking again from the outside.
The lock would be effectively glued.
1) it drops a vowel.
2) it rhymes with gay (bengay was a fluke).
Those Stingray bikes were considered pretty cool back when I was a kid.
Tor can be pretty scarey, but I'm not sure he would make logs worthless.
No, they announced prospective plans about Florida before this annoucement.
http://www.floridatoday.com/apps/pbcs.dll/article
If Singapore canes people for spitting in the subway, what are they going to do when you vomit on a airplane, even if it is in zero-gee.
Wow, its almost as if the ISPs were trying to say that people would have to pay more if they wanted thier packets routed with a high standard for delivery time. Where have I heard that recently?
If he just leaned that way, wouldn't his platform make the turn for him?
Screw rationality "Use the Force Luke! Let Go!"
Seems like its not that the subconcious mind makes better decisions, but that the subconcious mind can make your life miserable if it disagrees.
But what if after inflation (due to rapid increase in oil prices and economic chaos) $100,000 is basically 6 months of electricity?
Now that there is a way to make sound with network activity, the next step will be someone will write an app that plays actual music by generating varying rates of activity.
Sort of like they used to do with lineprinters. Print patterns that make the mechanism make certain tones.
Picture a sysop sitting in the NOC late at night , when all of a sudden, the sonic network montior starts emitting Darth Vader's theme.
Well, not for network operations, but for control systems in a space ship.
Walter Jon Williams had a story in which the starship control panel emitted sound patterns related to the functioning of the ship. If the characteristics of the soundscape changed that alerted the pilot that something had altered and be checked.
I think it was in the novel _Angel Station_
Firstly, there should not be any obvious bittorrent tagged stream, because encryption is being used. And there could be more than one data stream.
But to answer your question, because when the phrase "traffic shaping" is used that is not a general throttle on all bandwidth, but a throttle on specific sort of connection. If you make it difficult to detect that type of connection, you make it difficult to trigger the throttling.
The only thing that could be detected would be basic bandwidth use. And that would present a danger of many more false positives than traffic shaping specificly targeting P2P. There is an increased risk that such false positives would updet the entire customer base, not just the P2P users.
That's brahm's point.
But the supposition is that the ISP will 'shape' the traffic. If they do it by identifying customer who are apparently using P2P, but not being able to identify which TCP/IP sessions are the one carrying the traffic, they'll be forced to reduce bandwidth on every connection. That's so ham-handed its really difficult to call it traffic shaping at all. It basically giving a customer not even the suggestion of what they advertised to him as service.
Maybe the ISPs would not care about driving away P2P high bandwidth users. Maybe they wouldn't be concerned about false advertising claims. But I'd think they's worry about the reputation that will come from the screams of 'false positive' customers who will witness that the ISPs service is worse than dialup even though they did nothing out of the oridinary.
The more agile the target, the more computation it costs to track it. I can't see it as being practical to have firmware in the cable modem/dsl slam or in a cable headend that can track statistics on every unique session passing through and squash P2P sessions. Espcially if individual TCP connections complete in a matter of seconds.
The idea is to make shaping cost more than its worth, so that the ISPs can't selectively welch on the services they advertised that they provide.
Opens router config; set port port forwarding on for 1 through 65535.
NAT sure can make life difficult sometimes. (sigh)
Don't they have firewalls that let you use SNMP to modify rules yet?
I suppose if they did then malware would be messing with it too.
How would that prevent it from being lots of bidirectional line noise?
:)
If they are sniffing based on IP/port tuples, using one pipe one way and other for the other obfuscates the bidirectional part. For noise, you'd have to salt it with something that doesn't look random. Maybe just a "HTML" string at the beginning.
Modifiable, of course, to keep it a moving target.
Actually on further thought, I think it might be more difficult to track a transaction that kept rolling onto a new TCP/IP session every few seconds. Let the peers do the handshaking of the new connection in the background, and when its open for business, move the data off the old stream and onto the new on.
FTA:
"...a wire protocol which transfers a lot of data bidirectionally and consistently looks like line noise with no header is only marginally more difficult to identify then one which uses fixed ports."
Sounds like a call to camoflage the traffic as several pipes between peers. Not just one tcp/ip connection, but several, with a jitter function to pick which pipe is used at the moment so it does not look consistant
Since you are probably using ActivePerl, have you considered using the perlcc command to make a single EXE file of your application? Its easy to install as any other stand alone EXE file.
'Small' might be another story. Its certainly smaller than a full perl install on your target system. The EXE will be bigger than a meg certainly. Large compared to true compiled application, larger, and a bit slower to start up, as it unpacks its required modules from its internal package.
But a few meg isn't really that big of a deal these days.
YMMV
Why was this guy mod'd troll?
He expressed a technical point that its not difficult to get a different IP address. That leads to the obvious conclusion that banning an IP address is not the same as banning a person from editing.
Come on moderators!