You should take a look at http://www.ietf.org/ID-nits.html: you're still going to need some sort of authentication to prevent spoofing, and to provide nonrepudiation so you can blacklist spammers later.
In practice, the social aspects of trust would be a pain to implement. I personally would need to get my mail server trusted by everyone else in the net? Sounds like an inverse RBL, with the same sorts of problems (maybe worse because of scalability issues)
Make people send you digital cash with each email. You return it if the email isn't spam (if you don't return it for nonspam, then you're a bastard) Unfortunately, it's impossible to make this work in a back-compatible way, so...
Example protocol:
220 foo.bar.com CASHMAIL System
HELO
250 foo.bar.com Hello
MAIL FROM: spammer@mail.com
250 spammer@mail.com... Sender ok
RCPT TO: foo@bar.com
666 foo@bar.com requires payment of 20 cents
CASH: 82kd0xma893mcos0
667 foo.bar.com accepts payment of 20 cents
DATA ...
I think some work in the IETF has been done on spam prevention, but no one has even tried to standardize it.
The problem is very closely related to fair play in wireless ad hoc network routing. In wireless ad hoc networks, nodes forward packets for each other; a selfish node could save battery power and still get their packets routed. At least two papers in the published literature make attempts at this problem:
Enforcing Service Availability in Mobile Ad-Hoc WANs uses secure hardware to achieve this result. Obviously, this makes it open to law-enforcement attack, since the issuer of the hardware is a single point of failure. Also, it's a lot easier to get someone to download something than to buy a piece of secure hardware.
Mitigating Routing Misbehavior in Mobile Ad Hoc Networks doesn't try to stop misbehaving nodes; rather, they try to stop using misbehaving nodes for forwarding. (If you think this scheme is not directly applicable, think of the case of requesting a download of a file you just uploaded.)
Since this is an ongoing area of research, it'll be interesting to see what happens; any workable solution for ad hoc network routing fairness will also ensure p2p fairness. It doesn't work the other way around, since the routing mechanism itself is under attack.
It seems that in a p2p system, including digital signatures in shares, in combination with some kind of reputation system, might be a good way to both achieve fairness and eliminate spam. Maybe allowing leechers in times of excess bandwidth would jumpstart the system (a problem for warezers), and using "moderation point" like things to mod people up and down.
[Disclaimer: I only work in a somewhat related area; I haven't actually considered how one might solve either problem]
How about military bases? Trying to achieve fairness in national spending is somewhat problematic; maybe federal taxes should be paid to the states, and then congress tells the state how to spend the tax dollars the collected. =)
But realistically, money is the federal govt's only way to have any control: that's why the drinking age is 21, and why Bush can set education standards for schools.
USB.org's FAQ says USB is 12Mbps (v2 is 480Mbps), and FC-AL is 3.2Gbps. Even commodity ATA/133 gives 1064Mbps. USB can't replace disk interfaces. FireWire can. (As can FC-AL, but FC-AL is as commodity as Space Stations).
For two-channel I/O, USB provides plenty of bandwidth. There's a sledgehammer killing a fly argument to be made against making this device 1394. You could argue that everything should be replaced with firewire, including ATA and SCSI. I actually rather dislike the disappearance of RS-232, since it'll make hardware harder to hack. Putting together something that talks RS232 is so much easier for the average geek than something that talks USB/1394.
Anyhow, if you want a 1394 interface, check out MOTU. They have some killer audio interfaces for 1394.
I'm also a musician geek... I wish they'd bring the Digi001 interface into a PCMCIA card, for the same sorts of reasons. Of course, if you have firewire, you've been able to have MOTU's stuff for quite some time now (2408 was the first, but now the 896 gives 24bit/96kHz, 8 mic inputs (with individually switchable phantom), 8 outs (-10/+4 switchable) + stereo mains, and ADAT I/O.
Paper money/Digital cash rocks [Re:the coin route]
on
The Euro
·
· Score: 2
Men have no practical way to carry change... As soon as I get home, I also dump all my change in a bin. Unfortunately, I never have time to sort it out. What that means is a lower velocity of money and a higher savings rate, but that savings never makes it out as a loan, so it doesn't cause more monetary growth.
When I traveled in Europe and Canada before, the high denomination coins annoyed me (the Canadian twoonie and the 10ff coin are notable offenders), since I felt obligated to carry them and use them, or I'd never have a use for them when I got home. Italians must feel equally miffed, since their largest coin is 500Lira (about 25 cents) and you probably could go a week in Italy without ever using a coin.
The right solution if you're worried about reissue cost is smart cards and cryptographically anonymous cash. One reason I use credit cards everywhere I can is to reduce the amount of cash I need to keep around. I'd definatly carry a smart card around if it were legal tender...
Re:Using PCs to *route* gigabit ethernet?
on
Apartments for Techies?
·
· Score: 2, Interesting
Switch when you can, route when you must.
He's unlikely to get more than 10Mbps in, and a Pentium 90 running FreeBSD can route and NAT to saturate a 10Mbps link. All he needs is to switch each apartment with 100Mbps switchs with 1Gbps uplink, switch the 100Mbps switches with a 1Gbps switch, and plug the router into that switch complex to route traffic out over the ADSL line. Another advantage of using a FreeBSD/Linux box at that level is that you can firewall... in fact, you could firewall for the clueless users and punch holes in the firewall for those who want it...
They offer Fast Ethernet (for quite a high price), and I think it's good for up to 100Mbps to all Walden locations. The pipe to the ISP is 45MBps; probably T3 (I stand corrected).
I get some pretty killer data rates, in the hundreds of kb per second. There are probably more than 200 users, but statistical multiplexing works. Sure, 200 users all downloading MP3s at the same time gives a theoretical 230kbps, but if everyone offsets their download by just a little bit, average latency on a 5MB file is not long enough to even worry about waiting for.
No firewall. I think it rocks-- it allows you to run arbitrary servers. People often get their boxen 0wned; I see a lot of attacks against my server from.waldenweb.com addresses. It's the price you pay for completely open access, and you can always run your own firewall.
WaldenWeb has a few apartment complexes in the Houston area; they run an OC-3 from an ISP to their NOC, and run OC-3 from their NOC to each of their apartments. My apartment has 3 RJ-45 drops (only one of which I can make active at a time, but that's what a hub is for). Rent is reasonable, Internet access runs about $50/month.
Delta compression is the main reason
why you can get 100/1 ratios for video feeds.
Obviously, but the point is that the information content is still fixed. Using DSP technology, you can use better refresh to get better resolution, or vice versa.
If everyone had a local cache, it would be practical. And that's the point: you can't only rely on the wireless network.
No, compression would be worse, because the idea with DVD compression is to compress-out the things that wouldn't be observable. For example, consider how well you can compress a 1024x768 picture and have it still make sense, vs what you can do to a 640x480 picture.
It has to work for more than one person, unless you're going to set up the system for just yourself...
I understand the tradeoff with phones; I don't even own a land line. But the point is that lotsa storage can be had really small, so bringing your own data rather than sucking it over the network seems to be worse than having it locally...
wireless connections we have in the here and now are enough to send as much information as the human eyes take in real-time
DVD speed is 5.2Mbps (approximately). State of the art wireless is 54Mbps. Imagine a room of 100 people, each watching a different movie, and interacting with it. Even if IEEE 802.11 DCF was 100% efficient (it isn't), and even if we never want higher data rates than DVD (we will), and even if we won't want stimulus besides sight and soudn (we will), deploying enough base stations to have enough bandwidth everywhere will be prohibitively expensive.
The beauty of local storage and processing is that every person brings processing and storage. To the extent that we can use local resources, the available wireless bandwidth can provide better quality of service to those who actually need it/are willing to pay for it
Incidentally, there's a battery power tradeoff to grabbing things off a wireless network too... unfortunately, battery capacity doesn't double every 18 months.
Teledesic has been saying they're going to do this for years. However, this is really bleeding-edge technology, and if you know how often iridium phones drop calls on hand-offs, you won't be looking forward to an iridium-style internet =)
Realistically, though, I don't think there's a real market for it. Deploying 844 satellites (or whatever it is) is prohibitively expensive for covering that 1% of the population of people that can afford this stuff.
Not only would PK come crashing down; the use of ciphers with IVs would suck (as in CBC, CFB, probably OFB). Any time you have a known plaintext/ciphertext pair, it's over.
So, you'd rather have high-speed chases and errattic driving than gun battles?
I don't know about where you live, but here (Houston) we have street-races all the time. Those are pretty much high-speed chases, just neither party is a cop. People participating in said races generally do drive erratically. Kill switches would solve neither high-speed, reckless driving, nor would it solve erratic driving.
Or get the bogus law changed.
Slash has been so efficient at getting the DMCA changed. It's hard to get laws changed.
Or, better yet, *leave it in your car.*
Doesn't change the fact that you're breaking the law. It just makes it harder to enforce. And stowing it in the trunk doesn't help the stalker situation.
So, you'd rather have high-speed chases and errattic driving than gun battles?
If someone has a criminal history, I suppose I wouldn't mind their driver's license having a "kill-switch only" restriction (not that they're likely to follow it). In general, our law enforcement system relies on busting people after a crime is committed, not by curtailing rights.
Maybe because the kill-switch broadcasts its ability to work, and turning it off will draw the attention of every Cop around?
Please. How could it tell if it could work? By periodically self-testing? Easily defeatable, unless you make the engine and the kill-switch a single unit, sealed in tamper-proof hardware. But then I'll just mount two engines in my car, one with a kill switch and one without.
Besides, if you're being followed, *eventually* someone can and will find you, and kill you.
So let's just make it more efficient for the crooks.
1: If there is an officer of the law nearby, you rely on her to keep you safe.
Rare.
2: If the officer is not present, or is not helping, you rely on the duly licensed and registered assault rifle you have "stored" within easy reach.
I personally could live without gun battles in the streets. If some "evildoer" (to borrow a Bush term) tries to make me stop my vehicle, I would hope that I had the opportunity to first use skillful driving to escape, in addition to the potential threat of lethal force. Secondly, if your workplace is a university campus, it's a federal offense to bring a weapon.
3: If you're lacking anything to deter / defeat the criminal, you go along quietly and hope you survive it. If use of the system flags a GPS query, your chance of being rescued go up.
GPS jamming? And how do you relay the coordinates back to somewhere safe? And why do I believe that my death being "avenged" will be sufficient?
There's a reason they tell you to drive to a well-lit area if you're being followed. Giving cops a kill-switch is asking for trouble: currently, your "being stalked" technique is
0: drive to a well-lit area
1: if you can't make it, and there's a cop nearby...
Installing car-kill switches eliminates option 0 and disarms your most powerful weapon, merely for the convienence of a few officers participating in chases. This is a clear lose in my book: why would a crook buy/drive a car without disabling the car-kill switch?
Let's see... I do research in security, I know people who've dedicated their lives to researching computer security. Few people use locking screensavers, few people would tolerate having to type their password/phrase over and over again to decrypt things. There are techinques (see the ZIA talk at http://www.research.ibm.com/compsci/mobile/seminar.html) for doing this more efficiently (from a user perspective), but they're a few years out.
BTW locking screensavers like xlock are dangerous unless you can challenge the screensaver to authenticate itself before you give it a password.
and an auto-kill switch to aid in law enforcement would complety eliminate any high-speed chases. (if the cop can tell you to stop ANYWAY, what's wrong with him flipping a switch that kills your enigne for ten minutes and thus *forcing* you to do what you're legally required to do anyway?)
And what happens when a criminal can flip the same switch to carjack you? Or a stalker can more easily catch his/her target? Or Clayton Lee Wagner can pull over his next target and kill them in a more rural area than he might otherwise?
Any tools you give the cops will eventually be used by crooks (guns, wiretaps, pulling people over to kill them,...)
People don't use secure crypto because secure crypto is a pain in the rear from a user's perspective. Look at the way w2k requires reauthentication on wakeup. Nobody I know leaves that on once they know how to turn it off.
If you take a secure program (say, PGP) and give a little on the UI (say, retain keys for ten minutes), you lose lots of your security.
Crypto built into office apps? Why, if your OS is going to write your decrypted document all over your swap space?
Besides, would you really want non-repudiation for all your emails? =)
7 pounds was US$10 last time I went to London (August 2001), so the prices seem to be $10/1MB with plan, $30/MB without? Voicestream's best rates are $4/MB w/ plan.
Slashdot Mirror
You should take a look at http://www.ietf.org/ID-nits.html: you're still going to need some sort of authentication to prevent spoofing, and to provide nonrepudiation so you can blacklist spammers later.
In practice, the social aspects of trust would be a pain to implement. I personally would need to get my mail server trusted by everyone else in the net? Sounds like an inverse RBL, with the same sorts of problems (maybe worse because of scalability issues)
Example protocol:
220 foo.bar.com CASHMAIL System
...
HELO
250 foo.bar.com Hello
MAIL FROM: spammer@mail.com
250 spammer@mail.com... Sender ok
RCPT TO: foo@bar.com
666 foo@bar.com requires payment of 20 cents
CASH: 82kd0xma893mcos0
667 foo.bar.com accepts payment of 20 cents
DATA
I think some work in the IETF has been done on spam prevention, but no one has even tried to standardize it.
The problem is very closely related to fair play in wireless ad hoc network routing. In wireless ad hoc networks, nodes forward packets for each other; a selfish node could save battery power and still get their packets routed. At least two papers in the published literature make attempts at this problem:
Enforcing Service Availability in Mobile Ad-Hoc WANs uses secure hardware to achieve this result. Obviously, this makes it open to law-enforcement attack, since the issuer of the hardware is a single point of failure. Also, it's a lot easier to get someone to download something than to buy a piece of secure hardware.
Mitigating Routing Misbehavior in Mobile Ad Hoc Networks doesn't try to stop misbehaving nodes; rather, they try to stop using misbehaving nodes for forwarding. (If you think this scheme is not directly applicable, think of the case of requesting a download of a file you just uploaded.)
Since this is an ongoing area of research, it'll be interesting to see what happens; any workable solution for ad hoc network routing fairness will also ensure p2p fairness. It doesn't work the other way around, since the routing mechanism itself is under attack.
It seems that in a p2p system, including digital signatures in shares, in combination with some kind of reputation system, might be a good way to both achieve fairness and eliminate spam. Maybe allowing leechers in times of excess bandwidth would jumpstart the system (a problem for warezers), and using "moderation point" like things to mod people up and down.
[Disclaimer: I only work in a somewhat related area; I haven't actually considered how one might solve either problem]
Can't ProTools Free be used with MOTU's 896? If I had any free time to do multitrack recording, I'd seriously consider a TiBook + 896.
Maybe, but in ATA, two disks are talking simultaneously, and if they're replying from cache, they can have some pretty impressive bandwidth numbers.
How about military bases? Trying to achieve fairness in national spending is somewhat problematic; maybe federal taxes should be paid to the states, and then congress tells the state how to spend the tax dollars the collected. =)
But realistically, money is the federal govt's only way to have any control: that's why the drinking age is 21, and why Bush can set education standards for schools.
USB.org's FAQ says USB is 12Mbps (v2 is 480Mbps), and FC-AL is 3.2Gbps. Even commodity ATA/133 gives 1064Mbps. USB can't replace disk interfaces. FireWire can. (As can FC-AL, but FC-AL is as commodity as Space Stations).
Firewire's easy to use too. Plug and pray...
For two-channel I/O, USB provides plenty of bandwidth. There's a sledgehammer killing a fly argument to be made against making this device 1394. You could argue that everything should be replaced with firewire, including ATA and SCSI. I actually rather dislike the disappearance of RS-232, since it'll make hardware harder to hack. Putting together something that talks RS232 is so much easier for the average geek than something that talks USB/1394.
Anyhow, if you want a 1394 interface, check out MOTU. They have some killer audio interfaces for 1394.
I'm also a musician geek... I wish they'd bring the Digi001 interface into a PCMCIA card, for the same sorts of reasons. Of course, if you have firewire, you've been able to have MOTU's stuff for quite some time now (2408 was the first, but now the 896 gives 24bit/96kHz, 8 mic inputs (with individually switchable phantom), 8 outs (-10/+4 switchable) + stereo mains, and ADAT I/O.
Men have no practical way to carry change... As soon as I get home, I also dump all my change in a bin. Unfortunately, I never have time to sort it out. What that means is a lower velocity of money and a higher savings rate, but that savings never makes it out as a loan, so it doesn't cause more monetary growth.
When I traveled in Europe and Canada before, the high denomination coins annoyed me (the Canadian twoonie and the 10ff coin are notable offenders), since I felt obligated to carry them and use them, or I'd never have a use for them when I got home. Italians must feel equally miffed, since their largest coin is 500Lira (about 25 cents) and you probably could go a week in Italy without ever using a coin.
The right solution if you're worried about reissue cost is smart cards and cryptographically anonymous cash. One reason I use credit cards everywhere I can is to reduce the amount of cash I need to keep around. I'd definatly carry a smart card around if it were legal tender...
Switch when you can, route when you must.
He's unlikely to get more than 10Mbps in, and a Pentium 90 running FreeBSD can route and NAT to saturate a 10Mbps link. All he needs is to switch each apartment with 100Mbps switchs with 1Gbps uplink, switch the 100Mbps switches with a 1Gbps switch, and plug the router into that switch complex to route traffic out over the ADSL line. Another advantage of using a FreeBSD/Linux box at that level is that you can firewall... in fact, you could firewall for the clueless users and punch holes in the firewall for those who want it...
I get some pretty killer data rates, in the hundreds of kb per second. There are probably more than 200 users, but statistical multiplexing works. Sure, 200 users all downloading MP3s at the same time gives a theoretical 230kbps, but if everyone offsets their download by just a little bit, average latency on a 5MB file is not long enough to even worry about waiting for.
No firewall. I think it rocks-- it allows you to run arbitrary servers. People often get their boxen 0wned; I see a lot of attacks against my server from .waldenweb.com addresses. It's the price you pay for completely open access, and you can always run your own firewall.
WaldenWeb has a few apartment complexes in the Houston area; they run an OC-3 from an ISP to their NOC, and run OC-3 from their NOC to each of their apartments. My apartment has 3 RJ-45 drops (only one of which I can make active at a time, but that's what a hub is for). Rent is reasonable, Internet access runs about $50/month.
why you can get 100/1 ratios for video feeds.
Obviously, but the point is that the information content is still fixed. Using DSP technology, you can use better refresh to get better resolution, or vice versa.
If everyone had a local cache, it would be practical. And that's the point: you can't only rely on the wireless network.
It has to work for more than one person, unless you're going to set up the system for just yourself...
I understand the tradeoff with phones; I don't even own a land line. But the point is that lotsa storage can be had really small, so bringing your own data rather than sucking it over the network seems to be worse than having it locally...
DVD speed is 5.2Mbps (approximately). State of the art wireless is 54Mbps. Imagine a room of 100 people, each watching a different movie, and interacting with it. Even if IEEE 802.11 DCF was 100% efficient (it isn't), and even if we never want higher data rates than DVD (we will), and even if we won't want stimulus besides sight and soudn (we will), deploying enough base stations to have enough bandwidth everywhere will be prohibitively expensive.
The beauty of local storage and processing is that every person brings processing and storage. To the extent that we can use local resources, the available wireless bandwidth can provide better quality of service to those who actually need it/are willing to pay for it
Incidentally, there's a battery power tradeoff to grabbing things off a wireless network too... unfortunately, battery capacity doesn't double every 18 months.
Wouldn't we all love Xerox to pursue the Ebola pattern with Microsoft and their WinCE product?
=)
Realistically, though, I don't think there's a real market for it. Deploying 844 satellites (or whatever it is) is prohibitively expensive for covering that 1% of the population of people that can afford this stuff.
Not only would PK come crashing down; the use of ciphers with IVs would suck (as in CBC, CFB, probably OFB). Any time you have a known plaintext/ciphertext pair, it's over.
So, you'd rather have high-speed chases and errattic driving than gun battles?
I don't know about where you live, but here (Houston) we have street-races all the time. Those are pretty much high-speed chases, just neither party is a cop. People participating in said races generally do drive erratically. Kill switches would solve neither high-speed, reckless driving, nor would it solve erratic driving.
Or get the bogus law changed.
Slash has been so efficient at getting the DMCA changed. It's hard to get laws changed.
Or, better yet, *leave it in your car.*
Doesn't change the fact that you're breaking the law. It just makes it harder to enforce. And stowing it in the trunk doesn't help the stalker situation.
So, you'd rather have high-speed chases and errattic driving than gun battles?
If someone has a criminal history, I suppose I wouldn't mind their driver's license having a "kill-switch only" restriction (not that they're likely to follow it). In general, our law enforcement system relies on busting people after a crime is committed, not by curtailing rights.
Maybe because the kill-switch broadcasts its ability to work, and turning it off will draw the attention of every Cop around?
Please. How could it tell if it could work? By periodically self-testing? Easily defeatable, unless you make the engine and the kill-switch a single unit, sealed in tamper-proof hardware. But then I'll just mount two engines in my car, one with a kill switch and one without.
Besides, if you're being followed, *eventually* someone can and will find you, and kill you.
So let's just make it more efficient for the crooks.
Rare.
2: If the officer is not present, or is not helping, you rely on the duly licensed and registered assault rifle you have "stored" within easy reach.
I personally could live without gun battles in the streets. If some "evildoer" (to borrow a Bush term) tries to make me stop my vehicle, I would hope that I had the opportunity to first use skillful driving to escape, in addition to the potential threat of lethal force. Secondly, if your workplace is a university campus, it's a federal offense to bring a weapon.
3: If you're lacking anything to deter / defeat the criminal, you go along quietly and hope you survive it. If use of the system flags a GPS query, your chance of being rescued go up.
GPS jamming? And how do you relay the coordinates back to somewhere safe? And why do I believe that my death being "avenged" will be sufficient?
There's a reason they tell you to drive to a well-lit area if you're being followed. Giving cops a kill-switch is asking for trouble: currently, your "being stalked" technique is
0: drive to a well-lit area
1: if you can't make it, and there's a cop nearby...
Installing car-kill switches eliminates option 0 and disarms your most powerful weapon, merely for the convienence of a few officers participating in chases. This is a clear lose in my book: why would a crook buy/drive a car without disabling the car-kill switch?
BTW locking screensavers like xlock are dangerous unless you can challenge the screensaver to authenticate itself before you give it a password.
And what happens when a criminal can flip the same switch to carjack you? Or a stalker can more easily catch his/her target? Or Clayton Lee Wagner can pull over his next target and kill them in a more rural area than he might otherwise?
Any tools you give the cops will eventually be used by crooks (guns, wiretaps, pulling people over to kill them,
People don't use secure crypto because secure crypto is a pain in the rear from a user's perspective. Look at the way w2k requires reauthentication on wakeup. Nobody I know leaves that on once they know how to turn it off.
If you take a secure program (say, PGP) and give a little on the UI (say, retain keys for ten minutes), you lose lots of your security.
Crypto built into office apps? Why, if your OS is going to write your decrypted document all over your swap space?
Besides, would you really want non-repudiation for all your emails? =)
7 pounds was US$10 last time I went to London (August 2001), so the prices seem to be $10/1MB with plan, $30/MB without? Voicestream's best rates are $4/MB w/ plan.
OTOH, the UK is generally more expensive.