a) Should there be such a thing as illegal material, and then
b) If the answer to a) is "yes", then shouldn't it be censored on the internet.
You state you're not interested in discussing a), but want to focus on b), so let's do that:
There are a few ways that I can see that you can filter out illegal material:
i) Have a human prepare hashes of the banned material and refuse to transfer that material.
ii) Have a human look at all material that might be transferred as the transfer starts and ok it, or not.
iii) Have a computer look at all material that might be transferred as the transfer starts and use some heuristic matching scheme to ok it, or not.
i) will have no false positives or false negatives. Unfortunately, it is usually very easy to make small changes to a file so that it no longer matches the hash. This makes this scheme relatively ineffective, and hence a waste of resources to try.
ii) and iii) have a common problem: the mechanism for stopping a particular document is approximate and will get the answer wrong some percentage of the time. A human is more likely to get things right than a computer, but is much more expensive and there are privacy issues. A computer is going to have a horrid error rate. The article mentions 1%, but it isn't just a random 1%. It is a 1% around things related to the banned subjects. This means that people smuggling awareness sites are going to be falsely blocked at a much higher rate than banking web sites. This is a HUGE problem.
In summary: if it was possible to block all the illegal content and only the illegal content, then you can only argue a) above. Unfortunately it just isn't possible, and the collateral damage is worse than the original problem.
i) Use a revision control system. There are a bunch of good ones: git, monotone, darcs, bzr, subversion... This will give you ability to have people edit and share the documents.
It'll work better if you use document formats that are text based. e.g. unzipped ODF or latex for 'word processing'
These systems are very much collaborative, but are move away from 'instant' communications to 'parallel' editing with an assisted merge step.
ii) If you move away from RCS based thoughts to direct collaborative editing, then things get more complex. I only know of collaborative editors that edit one type of document. (and there is good reason for this - see http://en.wikipedia.org/wiki/Operational_transformation )
I know that inkscape (FOSS SVG editor) has a collaborative editing module based on Jabber comms. There are numerous collaborative text editors that people might point you towards.
iii) The third option is Google docs. For text this is ok. For images you can try to use their presentation software, but it is clunky for that purpose - use inkscape instead.
I'm not sure about this... but are they using GNU/Linux or just linux?
GNU/Linux refers to using the GNU toolchain on the linux kernel. I thought that Android used their own runtime environment (byteocde interpreter and libs) on a linux kernel.
If that is right, then they're using Linux and NOT GNU/Linux.
(I suspect there are GNU tools in the SDK, but I'd be surprised if there were any on the phone itself.)
There are a few interesting points in your post: It all hinges on the "reasonable expectation of privacy".
If I'm walking down a public road, and I look around and don't see anyone nearby, do I have a "reasonable expectation of privacy"?
Is there legal distinction between short term privacy and long term privacy? e.g. Is my expectation that people will not follow me around for any significant period of time "reasonable" under the US constitution?
If a police officer is patrolling in a marked police car, do they have a "reasonable expectation of privacy" or would it be ok to tag that police car with a GPS tracker and display the location real time in a Google Maps mashup? Is there some other law that would prevent this apart from the constitution?
If the above is ok, what about if the police office is parked behind some bushes/a billboard in a "Dukes of Hazard" style speed trap. Does that officer have a "reasonable expectation of privacy"?
What about if said officer is patrolling in an unmarked car (but one which was ID'd as a police car earlier), do they now have a "reasonable expectation of privacy"?
I'm guessing that most of these questions haven't been answered by US courts. I'd be particularly interested if there is a distinction between the expectation of privacy for police officers and the expectation of privacy for the general public.
I've long thought that they should get rid of.com,.net,.org,.gov etc, and only have country domains at the top level. That would make it clear which legal system governs that site.
You could mirror the current TLDs into.us to begin with - that would sort itself out pretty quick. You could also make any domain name without a county code resolve to the domain in the country you're in.
This resolves all the issues about trademark law, etc. They just use the law of that country.
Pity it'll never happen...
The Rules of the Swarm... on slashdot.
on
The Rules of the Swarm
·
· Score: 5, Insightful
When I read the summary my immediate response was that this was old news. It has been known for quite a while that cellular automata with simple rules can form universal Turing machines. That means that they can "form a collective brain" and "make decisions and move like a single organism". The interesting question then becomes; What "program" is your machine running, and how do you make it robust so that it works in the real world (with all the noise of nature)?
The article is a popular science article, but addresses this, more interesting, question much more than the summary. They discuss some of the rules involved in specific situations (ants), and even look at "human swarms" (although that bit is a little cheesy). There is no general theory posited about how to make these rule sets though, apart from trial and error (in simulation if you can). They say that the researchers are starting to see patterns, but don't talk about what those patterns are - pity really, as that would have been very interesting.
A feature you missed is the ability to install into a user account.
I don't have root access on every box I use. I want to install programs anyway. e.g. If I want to install a new version control system for my use, why should I need root access?
This is possible if I build from source, but the packaging system should make it easy.
- Sit on a chair with a rotating base
- Move your arm around an imaginary cone out from your shoulder pointing down to your side at 45 degrees. i.e. at the top of the cone your arm is pointing out to the side horozontal to the ground and moving forward, and at the bottom of the cone your arm is vertical by your side and moving backwards.
- As you move your arm forward at the top of the cone, your arm has a high moment of inertia. To maintain zero angular momentum your body will turn the other way (towards your arm).
- As you move your arm backwards at the bottom of the cone, it is near your side and hence has a lower moment of inertia. This means that your body doesn't have to rotate back as far to maintain zero angular momentum.
- You get a net turn towards your arm.
You can make the whole thing more effective by doing it with both arms at once. Just make sure your arms are moving the right way: if one is going forward at the top, the other must be going backwards at the top, else they'll cancel rather than add. (opposite directions because your arms are on opposite sides of your body).
The best bit is to make sure everyone in your office can see you waving your arms around. When they are you what you're doing, you have to tell them with a straight face, "conserving angular momentum".:)
Just throwing out an idea... There has been a lot of talk about whose laws should apply in cyberspace. One thought is that maybe it should be made explicit whose laws are applying by making the country explicit in the domain name. Ditch all.com,.org,.net, etc domains and just keep the country top level domains.
As an ease of use measure you could make.com redirect to.com.us in the US,.com.au in Australia, etc. Those names would only be useful as shorthand for people to type and would be deprecated as published URLs (because they would no longer mean the same thing everywhere).
When I access a.us site from Canada, the same laws apply as if I poked a stick over the border while standing in Canada.
Moreover, it removes the problem of VeriSign playing with the TLDs (at least for the rest of the world, I don't know who administers.us).
Pity it'll never happen.
Re:Tie IP protection to the tax code
on
Fair IP Laws?
·
· Score: 1
I also had this thought, but I decided I didn't like it. And then I came up with a modification that I did like. See below:
My issue with the suggestion above is that it ties the amount you can charge for a patent to the cost of production. That is not the way that most markets work. The value of something is the fair market value, not the cost of production.
Then I thought up a variation:
i) All IP must be available. You can't just say "Not for sale". ii) There is a cap on the price of IP licences: you can't charge more than your own net profit on the use of the IP. Again, this net profit is available through the IRS. (If you haven't made any money off the IP in ~3 years then the IP expires.)
Example: Company A has some Music. They make CDs for $12 and sell them for $20. This means Company A must license the music for no more than $20-$12=$8. Company B comes along and can make CDs more efficiently - it only costs them $8 to make the CDs. Company B has to pay $8 in licensing, but they can sell their CDs at $18 and still make $2 profit per CD. It also means that Company C can sell mp3s for $10 each and still make $2 per mp3.
An interesting thought: maybe companies should be able to choose which of these two systems (capital or net profit) they wish to use when they register the IP?
Another comment I have is WRT software patents. Software patents should be kept secret for two years. If they are independantly re-invented in that two years then they are invalid. If they're not original enough to survive two years without being independantly re-invented then they're not worth a patent.
Finally, I think the government needs to put some money into IP enforcement/education once the IP laws are fixed.
Slashdot Mirror
As you correctly note, there are two issues here:
a) Should there be such a thing as illegal material, and then
b) If the answer to a) is "yes", then shouldn't it be censored on the internet.
You state you're not interested in discussing a), but want to focus on b), so let's do that:
There are a few ways that I can see that you can filter out illegal material:
i) Have a human prepare hashes of the banned material and refuse to transfer that material.
ii) Have a human look at all material that might be transferred as the transfer starts and ok it, or not.
iii) Have a computer look at all material that might be transferred as the transfer starts and use some heuristic matching scheme to ok it, or not.
i) will have no false positives or false negatives. Unfortunately, it is usually very easy to make small changes to a file so that it no longer matches the hash. This makes this scheme relatively ineffective, and hence a waste of resources to try.
ii) and iii) have a common problem: the mechanism for stopping a particular document is approximate and will get the answer wrong some percentage of the time. A human is more likely to get things right than a computer, but is much more expensive and there are privacy issues. A computer is going to have a horrid error rate. The article mentions 1%, but it isn't just a random 1%. It is a 1% around things related to the banned subjects. This means that people smuggling awareness sites are going to be falsely blocked at a much higher rate than banking web sites. This is a HUGE problem.
In summary: if it was possible to block all the illegal content and only the illegal content, then you can only argue a) above. Unfortunately it just isn't possible, and the collateral damage is worse than the original problem.
Can it decrypt SSL/SSH in real time?
According to the article they use man-in-the-middle attacks. This is probably quite easy if the server is using self-signed certs.
There are three possibilities that I see here:
i) Use a revision control system. There are a bunch of good ones: git, monotone, darcs, bzr, subversion... This will give you ability to have people edit and share the documents.
It'll work better if you use document formats that are text based. e.g. unzipped ODF or latex for 'word processing'
These systems are very much collaborative, but are move away from 'instant' communications to 'parallel' editing with an assisted merge step.
ii) If you move away from RCS based thoughts to direct collaborative editing, then things get more complex. I only know of collaborative editors that edit one type of document. (and there is good reason for this - see http://en.wikipedia.org/wiki/Operational_transformation )
I know that inkscape (FOSS SVG editor) has a collaborative editing module based on Jabber comms. There are numerous collaborative text editors that people might point you towards.
iii) The third option is Google docs. For text this is ok. For images you can try to use their presentation software, but it is clunky for that purpose - use inkscape instead.
I'm not sure about this... but are they using GNU/Linux or just linux?
GNU/Linux refers to using the GNU toolchain on the linux kernel. I thought that Android used their own runtime environment (byteocde interpreter and libs) on a linux kernel.
If that is right, then they're using Linux and NOT GNU/Linux.
(I suspect there are GNU tools in the SDK, but I'd be surprised if there were any on the phone itself.)
There are a few interesting points in your post: It all hinges on the "reasonable expectation of privacy".
If I'm walking down a public road, and I look around and don't see anyone nearby, do I have a "reasonable expectation of privacy"?
Is there legal distinction between short term privacy and long term privacy? e.g. Is my expectation that people will not follow me around for any significant period of time "reasonable" under the US constitution?
If a police officer is patrolling in a marked police car, do they have a "reasonable expectation of privacy" or would it be ok to tag that police car with a GPS tracker and display the location real time in a Google Maps mashup? Is there some other law that would prevent this apart from the constitution?
If the above is ok, what about if the police office is parked behind some bushes/a billboard in a "Dukes of Hazard" style speed trap. Does that officer have a "reasonable expectation of privacy"?
What about if said officer is patrolling in an unmarked car (but one which was ID'd as a police car earlier), do they now have a "reasonable expectation of privacy"?
I'm guessing that most of these questions haven't been answered by US courts. I'd be particularly interested if there is a distinction between the expectation of privacy for police officers and the expectation of privacy for the general public.
I've long thought that they should get rid of .com, .net, .org, .gov etc, and only have country domains at the top level. That would make it clear which legal system governs that site.
You could mirror the current TLDs into .us to begin with - that would sort itself out pretty quick. You could also make any domain name without a county code resolve to the domain in the country you're in.
This resolves all the issues about trademark law, etc. They just use the law of that country.
Pity it'll never happen...
When I read the summary my immediate response was that this was old news. It has been known for quite a while that cellular automata with simple rules can form universal Turing machines. That means that they can "form a collective brain" and "make decisions and move like a single organism". The interesting question then becomes; What "program" is your machine running, and how do you make it robust so that it works in the real world (with all the noise of nature)?
The article is a popular science article, but addresses this, more interesting, question much more than the summary. They discuss some of the rules involved in specific situations (ants), and even look at "human swarms" (although that bit is a little cheesy). There is no general theory posited about how to make these rule sets though, apart from trial and error (in simulation if you can). They say that the researchers are starting to see patterns, but don't talk about what those patterns are - pity really, as that would have been very interesting.
A feature you missed is the ability to install into a user account.
I don't have root access on every box I use. I want to install programs anyway. e.g. If I want to install a new version control system for my use, why should I need root access?
This is possible if I build from source, but the packaging system should make it easy.
You know that you can do this youself at home:
:)
- Sit on a chair with a rotating base
- Move your arm around an imaginary cone out from your shoulder pointing down to your side at 45 degrees. i.e. at the top of the cone your arm is pointing out to the side horozontal to the ground and moving forward, and at the bottom of the cone your arm is vertical by your side and moving backwards.
- As you move your arm forward at the top of the cone, your arm has a high moment of inertia. To maintain zero angular momentum your body will turn the other way (towards your arm).
- As you move your arm backwards at the bottom of the cone, it is near your side and hence has a lower moment of inertia. This means that your body doesn't have to rotate back as far to maintain zero angular momentum.
- You get a net turn towards your arm.
You can make the whole thing more effective by doing it with both arms at once. Just make sure your arms are moving the right way: if one is going forward at the top, the other must be going backwards at the top, else they'll cancel rather than add. (opposite directions because your arms are on opposite sides of your body).
The best bit is to make sure everyone in your office can see you waving your arms around. When they are you what you're doing, you have to tell them with a straight face, "conserving angular momentum".
Just throwing out an idea... There has been a lot of talk about whose laws should apply in cyberspace. One thought is that maybe it should be made explicit whose laws are applying by making the country explicit in the domain name. Ditch all .com, .org, .net, etc domains and just keep the country top level domains.
.com redirect to .com.us in the US, .com.au in Australia, etc. Those names would only be useful as shorthand for people to type and would be deprecated as published URLs (because they would no longer mean the same thing everywhere).
.us site from Canada, the same laws apply as if I poked a stick over the border while standing in Canada.
.us).
As an ease of use measure you could make
When I access a
Moreover, it removes the problem of VeriSign playing with the TLDs (at least for the rest of the world, I don't know who administers
Pity it'll never happen.
I also had this thought, but I decided I didn't like it. And then I came up with a modification that I did like. See below:
My issue with the suggestion above is that it ties the amount you can charge for a patent to the cost of production. That is not the way that most markets work. The value of something is the fair market value, not the cost of production.
Then I thought up a variation:
i) All IP must be available. You can't just say "Not for sale".
ii) There is a cap on the price of IP licences: you can't charge more than your own net profit on the use of the IP. Again, this net profit is available through the IRS. (If you haven't made any money off the IP in ~3 years then the IP expires.)
Example: Company A has some Music. They make CDs for $12 and sell them for $20. This means Company A must license the music for no more than $20-$12=$8. Company B comes along and can make CDs more efficiently - it only costs them $8 to make the CDs. Company B has to pay $8 in licensing, but they can sell their CDs at $18 and still make $2 profit per CD. It also means that Company C can sell mp3s for $10 each and still make $2 per mp3.
An interesting thought: maybe companies should be able to choose which of these two systems (capital or net profit) they wish to use when they register the IP?
Another comment I have is WRT software patents. Software patents should be kept secret for two years. If they are independantly re-invented in that two years then they are invalid. If they're not original enough to survive two years without being independantly re-invented then they're not worth a patent.
Finally, I think the government needs to put some money into IP enforcement/education once the IP laws are fixed.