Many grocery stores now label where produce comes from. (Many produce companies do, too.) Usually state of origin within the US and country of origin otherwise. As long as you're okay with produce from California and Mexico, it's pretty easy to avoid produce from distant lands.
If the phone is going to send the credit card number to a nearby device (which is what it's doing with NFC), then it needs to have access to the plaintext credit card number. So the phone can store the CC# encrypted, but it has to be encrypted with a key based on user input (which is generally short) or stored on the phone (which doesn't actually provide substantial security).
I'm not even a motorcyclist and I agree with that. Almost every time, when someone is driving dangerously badly, they're busy talking on their phone. The other times, it's 2 AM on a Friday night.
Well, tunneling over HTTPS will shout to any analyzer that you're doing an SSL handshake, which they can block by policy.
It's true that you can write down the start sequences to almost every meaningful L7 protocol and then look for them. That doesn't let you detect encrypted data in the fully general case, but it does let you block lots of protocols you don't like.
Since BitTorrent is a custom wire protocol and not HTTP, what it would need to do is detect non-HTTP traffic over port 80 as a problem. Encrypted data cannot be meaningfully differentiated from compressed data -- though it's possible to look for common protocols (like an SSL handshake).
You don't need to talk about P2P in general. This is specifically talking about BitTorrent. The BitTorrent protocol is well-known. Data transfer between peers is done via a custom wire protocol over TCP.
Further, it's already known that the IP harvesting that produced the data discussed in TFA is performed by BitTorrent tracker scraping, which often reports the IP addresses supplied via the (layer 7) BitTorrent tracker protocol (which is HTTP-based).
Incidentally, one of the reasons TCP is preferred by BitTorrent is that it works better than UDP over NATing routers, particularly ones that don't do UPnP. Unfortunately, TCP connections require NATing routers to do connection-tracking. BitTorrent can create so many TCP connections that low-grade consumer routers crash from the high connection-tracking demands.
Of course, BitTorrent isn't really decentralized. That is, the original, tracker-based BitTorrent system isn't decentralized, and that's where the data in TFA comes from. Actual decentralized networks, particularly those based on distributed hash tables (including the DHT later integrated into BitTorrent) does use UDP.
Port 80 is popular because it's far and away the most likely port to be open on a firewall, more so than 443.
This is already a library -- libxul. They can break it into smaller pieces, but that's not the problem. It actually compiles and links fine. The problem is that it's an aggressive form of link-time optimization, which requires that (a) all of the objects being linked together be in memory at once and (b) the linker stores a ton of extra data for performing the optimization. If you give up on link-time optimization (which is very resource-intensive), code modularity works just fine.
In fact, it would probably be better if web browsers were more similar to operating systems -- two of the things that browsers seem to be bad at is memory and cache management and task scheduling (particularly, multiprocessing).
No, I'm not thinking of actual IP spoofing at all.
I'm referring to the ip parameter of the GET request that's sent to the tracker. Some trackers honor this, some don't. You'd think it would be redundant information, since your IP address is already transmitted as part of the IP packet header, but it isn't always the case that the IP address other peers should contact is the same as the IP address making the tracker GET request (e.g., if you have particular kinds of proxies). This is documented in the BitTorrent specification (Tracker HTTP Procotol: Request Parameters).
Also, connections to peers aren't over UDP; they're also TCP connections. (Same document, "Peer wire protocol".)
The BitTorrent tracker protocol does allow you to tell the tracker what your "real" IP address is, in the event it's different from the IP address the HTTP request is coming from. Some trackers honor this field, so it can be used to obtain a list of peers from a tracker while posting a completely-false IP address to the tracker list (since this field is unverified).
Reliably? Yes. Sure, it's easy to delete the copy in the production database. It's harder to prove that if the disks backing the production database were stolen and analyzed, it would be impossible to recover the data. It's harder still to locate and redact every backup of the database that contains the data. (It's even harder still to prove that a copy of the data doesn't persist on another user's hard drive as a result of having viewed the data in a web browser.)
This is the Cloud Era; you can't reliably delete data any more.
Most of the rules don't significantly hamper your ability to wage war. The ones that do -- restrictions on the use of chemical, biological, and nuclear weapons -- are very likely to provoke international retaliation.
So, in general, it's the group that obeys the rules that win, because the guys with the better militaries also tend to be the ones that politely follow the rules.
IIRC, medics that are part of a combat unit aren't protected and don't mark themselves with a cross. Separate medical units are protected and generally aren't targets.
Slashdot Mirror
Many grocery stores now label where produce comes from. (Many produce companies do, too.) Usually state of origin within the US and country of origin otherwise. As long as you're okay with produce from California and Mexico, it's pretty easy to avoid produce from distant lands.
Something like chicken nuggets would be a challenge, but there's nothing stopping processors from making single-animal ground meat and sausage.
Drinking doesn't actually damage your brain cells, it just inhibits their action temporarily.
If the phone is going to send the credit card number to a nearby device (which is what it's doing with NFC), then it needs to have access to the plaintext credit card number. So the phone can store the CC# encrypted, but it has to be encrypted with a key based on user input (which is generally short) or stored on the phone (which doesn't actually provide substantial security).
I'm not even a motorcyclist and I agree with that. Almost every time, when someone is driving dangerously badly, they're busy talking on their phone. The other times, it's 2 AM on a Friday night.
Well, tunneling over HTTPS will shout to any analyzer that you're doing an SSL handshake, which they can block by policy.
It's true that you can write down the start sequences to almost every meaningful L7 protocol and then look for them. That doesn't let you detect encrypted data in the fully general case, but it does let you block lots of protocols you don't like.
Since BitTorrent is a custom wire protocol and not HTTP, what it would need to do is detect non-HTTP traffic over port 80 as a problem. Encrypted data cannot be meaningfully differentiated from compressed data -- though it's possible to look for common protocols (like an SSL handshake).
You don't need to talk about P2P in general. This is specifically talking about BitTorrent. The BitTorrent protocol is well-known. Data transfer between peers is done via a custom wire protocol over TCP.
Further, it's already known that the IP harvesting that produced the data discussed in TFA is performed by BitTorrent tracker scraping, which often reports the IP addresses supplied via the (layer 7) BitTorrent tracker protocol (which is HTTP-based).
Incidentally, one of the reasons TCP is preferred by BitTorrent is that it works better than UDP over NATing routers, particularly ones that don't do UPnP. Unfortunately, TCP connections require NATing routers to do connection-tracking. BitTorrent can create so many TCP connections that low-grade consumer routers crash from the high connection-tracking demands.
Of course, BitTorrent isn't really decentralized. That is, the original, tracker-based BitTorrent system isn't decentralized, and that's where the data in TFA comes from. Actual decentralized networks, particularly those based on distributed hash tables (including the DHT later integrated into BitTorrent) does use UDP.
Port 80 is popular because it's far and away the most likely port to be open on a firewall, more so than 443.
That's definitely the solution, but not one that's available to Mozilla developers.
This is already a library -- libxul. They can break it into smaller pieces, but that's not the problem. It actually compiles and links fine. The problem is that it's an aggressive form of link-time optimization, which requires that (a) all of the objects being linked together be in memory at once and (b) the linker stores a ton of extra data for performing the optimization. If you give up on link-time optimization (which is very resource-intensive), code modularity works just fine.
In fact, it would probably be better if web browsers were more similar to operating systems -- two of the things that browsers seem to be bad at is memory and cache management and task scheduling (particularly, multiprocessing).
You mean natural gas, which is what is actually being recovered by this technique?
What keeps it from causing contamination through those fissures is that the fissures don't exist until the hydraulic fracturing is performed.
64-bit Windows lets 32-bit applications access 4 GB of virtual address space. A 64-bit application can access much more, but the linker isn't 64-bit.
No, I'm not thinking of actual IP spoofing at all.
I'm referring to the ip parameter of the GET request that's sent to the tracker. Some trackers honor this, some don't. You'd think it would be redundant information, since your IP address is already transmitted as part of the IP packet header, but it isn't always the case that the IP address other peers should contact is the same as the IP address making the tracker GET request (e.g., if you have particular kinds of proxies). This is documented in the BitTorrent specification (Tracker HTTP Procotol: Request Parameters).
Also, connections to peers aren't over UDP; they're also TCP connections. (Same document, "Peer wire protocol".)
The BitTorrent tracker protocol does allow you to tell the tracker what your "real" IP address is, in the event it's different from the IP address the HTTP request is coming from. Some trackers honor this field, so it can be used to obtain a list of peers from a tracker while posting a completely-false IP address to the tracker list (since this field is unverified).
Who has lost a civil case based only on an IP address?
Reliably? Yes. Sure, it's easy to delete the copy in the production database. It's harder to prove that if the disks backing the production database were stolen and analyzed, it would be impossible to recover the data. It's harder still to locate and redact every backup of the database that contains the data. (It's even harder still to prove that a copy of the data doesn't persist on another user's hard drive as a result of having viewed the data in a web browser.)
This is the Cloud Era; you can't reliably delete data any more.
That's not good science at all. You don't even have a control group.
So, you would fall into the category of "those those don't understand ternary".
Yes, that's what I meant. Though if it's never accessed, it'll probably only take up lines in the page tables and not in the TLB.
A resource-guaranteeing OS may allocate unused pages, but you're probably not going to see Java running one one of those anyway.
Thanks to virtual memory systems, almost all of that 2 GB isn't actually consumed, because it's all zero pages that have never been touched.
The Internet is a unicast network or, rarely, a multicast network. Not a broadcast network.
The Librarian of Congress is in charge of making exemptions to the DMCA.
Most of the rules don't significantly hamper your ability to wage war. The ones that do -- restrictions on the use of chemical, biological, and nuclear weapons -- are very likely to provoke international retaliation.
So, in general, it's the group that obeys the rules that win, because the guys with the better militaries also tend to be the ones that politely follow the rules.
IIRC, medics that are part of a combat unit aren't protected and don't mark themselves with a cross. Separate medical units are protected and generally aren't targets.