Not only half duplex, but as you say in your post title (not the text), the resources are shared between all users within a physical area. Aggregate throughput drops quickly as the number of users on a WLAN increases.
If we replaced our copper connections with WLAN at my company, the network would become effectively useless. Too many users.
Another way to think of it is: For a typical user, even a 100Base-T wire to a switch will match even the latest and greatest MIMO high speed implementations (advertised 270-300, but in reality you'll be lucky to see 100 Mbps real world in a single direction).
Once you go above 2-3 users, the switch connections win hands-down.
Add gigabit into the mix (cheap nowadays) and wired wins by an even greater margin.
Where did I say ANYTHING about having the original file?
ECB mode does make a "guessed plaintext" cryptanalysis attack more feasible. For example, if you encrypt empty space (all zeroes) with ECB, an attacker will see lots of repeating ciphertext and guess that it's probably the ciphertext for a zero block.
Now for at least one value, you know the cleartext and ciphertext, which in theory makes it easier to discover the key. (With AES, it's still Really Damn Hard, and at best makes it easier to know you guessed right. With weaker block ciphers it can kill you - For example, some attacks against CSS used "guessed plaintext" attacks I believe, where they took advantage of a few things: 1) The CSS block cipher was weak and you gained a lot of information about the key by having a block of both plaintext and ciphertext 2) Black MPEG-2 frames encode very predictably 3) Nearly all movies started with a few black frames
You never had the original cleartext file, BUT you could use statistical analysis to guess the cleartext for some blocks with high certainty and get you closer to knowing the key.
Actually, if you use the wrong block cipher mode, it's easy to distinguish between an encrypted file and random noise. AES-256 encrypts 128 bits of data at a time (with a 256-bit key). If you use the same key and the same block of data (ECB mode), you get the same output and can determine that there's something there.
If you modify each block with some known quantity that is different from block to block, then the output becomes much less patterned. For example, Counter (CTR) mode XORs or adds an increasing count to each block of cleartext, so that if you have two identical blocks of cleartext, the output is very different. Cipher Block Chaining (CBC) takes the encrypted output of block N and XORs it with the cleartext of block N+1 before encrypting that block.
I agree. Back in college a lot of friends of mine and I (who were almost all Linux fans) all felt that the only division of MS that made good products was their input devices division. Even the Linux Users Group would often reccommend MS mice to people looking for a good mouse.
Their routers were OK too, not quite as high quality as the mice/keyboards/joysticks though.
I now carry a Logitech G5 in my laptop bag at all times (Yes, I bought a *second* G5 just for my laptop), like you I've basically given up on wireless mice (Bluetooth or proprietary). The only time I use a wireless mouse is with my HTPC in the living room.
I agree. It's next to impossible to find a full sized Bluetooth mouse now. They're all travel mice.
Strangely enough, despite being travel mice which SHOULD have improved performance on "odd" surfaces, my Microsoft Bluetoooth Notebook Mouse 6000 (don't remember the exact name, it's basically MS's only BT offering) is the worst mouse I've ever used on carpet surfaces or couches. Even my first generation MS Intellimouse Explorer (the first optical mouse ever released to my knowledge) is superior on most surfaces.
I've heard that Logitech's equivalent to the MS notebook mouse isn't any better, in fact on Newegg the MS unit got generally better reviews for "odd-surface" usability.
Another annoyance is that you can't change the pointer speed of Bluetooth mice separately from other pointing devices in any Linux distro I've tried (Ubuntu and Gentoo). In both cases if the pointer speed is adjusted for good trackpad usability, the cursor moves at Mach 1 for the Bluetooth mouse.
Comcast's approach (bogus RST injection) was even nastier in some ways, since it would outright kill a connection instead of slowing it down. If you were using a protocol that didn't resume partial uploads (like Lotus Notes) you were completely screwed.
At least with CV's approach, you could still upload stuff, it just took forever because it was so slow.
In both cases, the companies never acknowledged that they engaged in such practices, at least not until quite a while after the public outcry.
Time Warner has no influence over Cablevision, other than being "buddy buddy" with them.
Each has their own monopoly over their given geographic area. In fact, the big boys (and CV is DEFINATELY one of them, not a "little guy" by ANY means!) have their own effective cartel with CableLabs.
Yeah. Cablevision is notorious for some shady "stealth capping" policies, or at least was back in 2003-2004.
If you used too much of your upload bandwith (with "too much" being undocumented to the customer), you would have your upstream cap lowered to 150 kilobits/sec (from something like 1.5 Mbits) without any notice.
He's pissed because somehow, someone in his office interpreted "keep it on the down low" (including the "need to know" phrase) as "the mayor doesn't need to know".
Yes, he did nothing. How could he do anything when HE WASN'T EVEN NOTIFIED!
The mayor of the city that's getting flown over SHOULD be notified of such items, but instead of directly contacting the mayor, the FAA contacted some city hall flunkie who cowered in fear at the distribution statement so much that he didn't even tell those who had a "need to know".
It is possible to design a machine that is secure even from someone who has physical access, but doing so is expensive and involves compromises in usability that normal users would never accept. (Of the "you no longer own your own machine" kind.)
Sometimes there are cases where you're using a no-longer-maintained software tool that itself does not work on newer Windows version.
At work we have a Windows NT machine that performs one specific function, the software that performs this function fails on Win2k/XP/Vista - it's THAT old and it's unmaintained.
As a result that machine is firewalled off from the rest of the network.
Depending on the exact insulation and the environment (see a previous post regarding outdoor cabling), the insulation could degrade with age.
Also, the physical geometry of the cabling is important for high speed networks. If the cable gets moved around frequently, it could degrade to the point where it no longer works.
There is of course the whole upgradability thing - Cat5 is good enough for 100M, and 5e is good enough for gigabit, but what if a few years down the line you want to go 10GbE? It seems outlandishly fast now, but it's around a 5-8 year cycle between Ethernet generations. What's standard now (GbE) for new installations will be "old hat" in 5-8 years.
It's a lot easier to upgrade networking equipment at the endpoints than to upgrade cabling runs.
It sounds like the security of the "as designed" military system was implemented at the endpoint radios with no satellite involvement. For 1970s satellites this makes sense - keep the satellite (the REALLY expensive part) simple, unfortunately it does make the system susceptible to illicit use and jamming.
The linked article already talks about this. Data retention laws for antiterrorism purposes are already going through the legislative pipeline in Europe apparently.
You seem to forget that their current production model (which is backordered for at least a year or two I believe despite the current economic climate) is $110k. (Not sure if that's before or after the government tax credit). At that price point, a toy for the extremely rich, but still a backordered toy.
So the Model S, while expensive, is approximately HALF the price of the previous model. (Exactly what the ratio is depends on whether the Roadster's 110k is pre-tax-credit or post-credit.) At its price, it's definately restricted to the upper middle class, but it IS attainable for an upper middle class family.
If that trend continues, then Tesla's next model will likely hit the price point where the average consumer can afford it.
Slashdot Mirror
Sadly, stuff like that is quite common in job postings, for the reasons stated above - the hiring manager's requirements get filtered by HR.
Not only half duplex, but as you say in your post title (not the text), the resources are shared between all users within a physical area. Aggregate throughput drops quickly as the number of users on a WLAN increases.
If we replaced our copper connections with WLAN at my company, the network would become effectively useless. Too many users.
Another way to think of it is: For a typical user, even a 100Base-T wire to a switch will match even the latest and greatest MIMO high speed implementations (advertised 270-300, but in reality you'll be lucky to see 100 Mbps real world in a single direction).
Once you go above 2-3 users, the switch connections win hands-down.
Add gigabit into the mix (cheap nowadays) and wired wins by an even greater margin.
Where did I say ANYTHING about having the original file?
ECB mode does make a "guessed plaintext" cryptanalysis attack more feasible. For example, if you encrypt empty space (all zeroes) with ECB, an attacker will see lots of repeating ciphertext and guess that it's probably the ciphertext for a zero block.
Now for at least one value, you know the cleartext and ciphertext, which in theory makes it easier to discover the key. (With AES, it's still Really Damn Hard, and at best makes it easier to know you guessed right. With weaker block ciphers it can kill you - For example, some attacks against CSS used "guessed plaintext" attacks I believe, where they took advantage of a few things:
1) The CSS block cipher was weak and you gained a lot of information about the key by having a block of both plaintext and ciphertext
2) Black MPEG-2 frames encode very predictably
3) Nearly all movies started with a few black frames
You never had the original cleartext file, BUT you could use statistical analysis to guess the cleartext for some blocks with high certainty and get you closer to knowing the key.
I think it goes with the whole "Innocent until proven guilty beyond a reasonable doubt" mantra.
If a file is indistinguishable from random noise, then a court can't prove that it's encrypted data.
That said, in reality they can make your life a living hell.
Actually, if you use the wrong block cipher mode, it's easy to distinguish between an encrypted file and random noise. AES-256 encrypts 128 bits of data at a time (with a 256-bit key). If you use the same key and the same block of data (ECB mode), you get the same output and can determine that there's something there.
If you modify each block with some known quantity that is different from block to block, then the output becomes much less patterned. For example, Counter (CTR) mode XORs or adds an increasing count to each block of cleartext, so that if you have two identical blocks of cleartext, the output is very different. Cipher Block Chaining (CBC) takes the encrypted output of block N and XORs it with the cleartext of block N+1 before encrypting that block.
Apple's purchase of PASemi was big news. Scattered reports, my ass.
I agree. Back in college a lot of friends of mine and I (who were almost all Linux fans) all felt that the only division of MS that made good products was their input devices division. Even the Linux Users Group would often reccommend MS mice to people looking for a good mouse.
Their routers were OK too, not quite as high quality as the mice/keyboards/joysticks though.
I now carry a Logitech G5 in my laptop bag at all times (Yes, I bought a *second* G5 just for my laptop), like you I've basically given up on wireless mice (Bluetooth or proprietary). The only time I use a wireless mouse is with my HTPC in the living room.
I agree. It's next to impossible to find a full sized Bluetooth mouse now. They're all travel mice.
Strangely enough, despite being travel mice which SHOULD have improved performance on "odd" surfaces, my Microsoft Bluetoooth Notebook Mouse 6000 (don't remember the exact name, it's basically MS's only BT offering) is the worst mouse I've ever used on carpet surfaces or couches. Even my first generation MS Intellimouse Explorer (the first optical mouse ever released to my knowledge) is superior on most surfaces.
I've heard that Logitech's equivalent to the MS notebook mouse isn't any better, in fact on Newegg the MS unit got generally better reviews for "odd-surface" usability.
Another annoyance is that you can't change the pointer speed of Bluetooth mice separately from other pointing devices in any Linux distro I've tried (Ubuntu and Gentoo). In both cases if the pointer speed is adjusted for good trackpad usability, the cursor moves at Mach 1 for the Bluetooth mouse.
But it's just not as craptastic as Milorganite...
Comcast's approach (bogus RST injection) was even nastier in some ways, since it would outright kill a connection instead of slowing it down. If you were using a protocol that didn't resume partial uploads (like Lotus Notes) you were completely screwed.
At least with CV's approach, you could still upload stuff, it just took forever because it was so slow.
In both cases, the companies never acknowledged that they engaged in such practices, at least not until quite a while after the public outcry.
Time Warner has no influence over Cablevision, other than being "buddy buddy" with them.
Each has their own monopoly over their given geographic area. In fact, the big boys (and CV is DEFINATELY one of them, not a "little guy" by ANY means!) have their own effective cartel with CableLabs.
They've been hiding a "stealth capping" secret for 5-6 years now.
Or at least they were 5-6 years ago, my guess is that they still are. (See my previous post in this topic for details.)
Yeah. Cablevision is notorious for some shady "stealth capping" policies, or at least was back in 2003-2004.
If you used too much of your upload bandwith (with "too much" being undocumented to the customer), you would have your upstream cap lowered to 150 kilobits/sec (from something like 1.5 Mbits) without any notice.
Read TFA.
He's pissed because somehow, someone in his office interpreted "keep it on the down low" (including the "need to know" phrase) as "the mayor doesn't need to know".
Yes, he did nothing. How could he do anything when HE WASN'T EVEN NOTIFIED!
The mayor of the city that's getting flown over SHOULD be notified of such items, but instead of directly contacting the mayor, the FAA contacted some city hall flunkie who cowered in fear at the distribution statement so much that he didn't even tell those who had a "need to know".
ROT26 is too weak. You need to be using ROT104 or higher.
It is possible to design a machine that is secure even from someone who has physical access, but doing so is expensive and involves compromises in usability that normal users would never accept. (Of the "you no longer own your own machine" kind.)
But what license was the FICS code under? Was it really "open source"?
There are plenty of licenses that provide for distribution of source but are so restrictive that no one considers them to be "open source".
Sometimes there are cases where you're using a no-longer-maintained software tool that itself does not work on newer Windows version.
At work we have a Windows NT machine that performs one specific function, the software that performs this function fails on Win2k/XP/Vista - it's THAT old and it's unmaintained.
As a result that machine is firewalled off from the rest of the network.
Depending on the exact insulation and the environment (see a previous post regarding outdoor cabling), the insulation could degrade with age.
Also, the physical geometry of the cabling is important for high speed networks. If the cable gets moved around frequently, it could degrade to the point where it no longer works.
There is of course the whole upgradability thing - Cat5 is good enough for 100M, and 5e is good enough for gigabit, but what if a few years down the line you want to go 10GbE? It seems outlandishly fast now, but it's around a 5-8 year cycle between Ethernet generations. What's standard now (GbE) for new installations will be "old hat" in 5-8 years.
It's a lot easier to upgrade networking equipment at the endpoints than to upgrade cabling runs.
Based on the description in the article, I'm guessing the sats in question use linear transponders similar to many of the AMSAT Oscar (AO) satellites.
http://www.amsat.org/amsat-new/information/faqs/langdon.php has info on some of the AMSAT satellites.
It sounds like the security of the "as designed" military system was implemented at the endpoint radios with no satellite involvement. For 1970s satellites this makes sense - keep the satellite (the REALLY expensive part) simple, unfortunately it does make the system susceptible to illicit use and jamming.
The linked article already talks about this. Data retention laws for antiterrorism purposes are already going through the legislative pipeline in Europe apparently.
You don't need DHCP server logs for billing purposes.
You do need them for hunting file sharers.
Basically describes every Slashdot poster. :)
You seem to forget that their current production model (which is backordered for at least a year or two I believe despite the current economic climate) is $110k. (Not sure if that's before or after the government tax credit). At that price point, a toy for the extremely rich, but still a backordered toy.
So the Model S, while expensive, is approximately HALF the price of the previous model. (Exactly what the ratio is depends on whether the Roadster's 110k is pre-tax-credit or post-credit.) At its price, it's definately restricted to the upper middle class, but it IS attainable for an upper middle class family.
If that trend continues, then Tesla's next model will likely hit the price point where the average consumer can afford it.