No, it's not broccoli. Nor chocolate. It's creamed chipped beef on toast a/k/a sh*t on a shingle. Only on stale toast. You eat it because you are there to take orders. Not to think.
No, NAT will not die. NAT is a good idea, not a bad one. Virtually everyone uses firewalls nowadays, most of which do NAT, which adds a level of security (not enough by itself, but it helps).
It is a critical flaw in TCP/IP architecture that the application translates the name to the address and sees the IP address. And there's never a good reason for applications to have numeric IP addresses inside them. NAT only breaks broken applications. IPv6 is Just Plain Stupid. It's ugly and it wants to die. And it will. The people who are pushing it are the kind of people who seek out authority in order to obey it blindly.
Apple isn't using them because a) they're not out yet, b) they aren't mass-produced the way Apple needs them, and c) Apple has volume contracts for screens with its great friend Samsung.
(Yes, the irony is real -- they are suing Samsung while simultaneously buying tons from them.)
A lot of Toyota engines failed due to a cylinder cooling problem. They were replaced by Toyota, but it was a big costly deal for user and company alike. These were in Camrys and Siennas (minivans).
But my Corolla has been exceptionally reliable. The new ones look really cheap though. The LE of 2012 has a flimsier interior than the CE of 2004.
There still are collocators, just not as many. The FCC adopted policies in 2000-2004 that made it harder to be a CLEC, and many went out of business. There are still some, but they tend to concentrate on business customers, who can pay more. Collocation is also used by some CLECs that provide wholesale interconnection services to VoIP providers. In general, a VoIP provider needs a CLEC to get blocks of numbers and interconnection to the ILEC. Level 3 is probably the biggest wholesale player.
The huge savings in telephone company real estate happened over 20 years ago. Their big buildings were built for electromechanical switching systems, mostly installed between 1920 and 1970. The digital switches mostly installed in the 1980s were a fraction of the size, leaving lots of empty space in the big buildings. Some space has already been repurposed. And some is available, but the Bells don't want to give it up because it would make competition easier.
Most of the real estate still used by telco gear is for line drivers, the stuff needed to run analog phones. Whether these are fed by VoIP or TDM doesn't matter; 90 volt power ring and 48 volt battery take space. They also take power, but home-based analog terminal adapters (local battery) use even more, so centralized power (common battery) is a net savings.
Berninger is simply repeating Cisco memes, that somehow the magic pixie dust of IP makes everything wonderfuler. It's bullshit, but somebody has to call them on it.
The large (Bell and other) telephone companies are not regulated on rate of return any more. They are on "price caps". Only the smallest carriers, the mom'n'pops and subsidy-dependent rural ones, are on rate of return. That's why the Bells have laid off so many people and stopped investing - they are milking their old plant for all it's worth.
Most of these sites have the word "jersey" in them. It looks as if the NFL's licensing squad went a-hunting, and gave the list of unauthorized vendors to Uncle Sam. What's not obvious is whether all of these sites simply sell unauthorized jerseys, or whether other jersey vendors, or people from a certain island or state, also got nailed in the crosfire.
ITU is not the primary standards body for this, and their definition of 4G is irrelevant. Geeknet doesn't write COBOL standards either. 3G was defined by 3GPP (GSM -> WCDMA/UMTS/HSPDA) and 3GPP2 (CDMA2000). All 3G is based on CDMA. Now 3GPP has defined LTE, which is different and newer technology, OFDMA, so it's a genuine generational shift. That's 4G, no matter what the speed. It gets more bits per Hz though.
ATT and T-Mobile are flogging their HDPDA+ (WCDMA) networks as "4G", because under really good conditions, they can get more speed than older forms of 3G. But it's just a late-life kicker for 2000's technology. VZW, to their credit, already has LTE, which is the real 4G, while ATT will have it later. And Sprint/Clearwire's WiMAX is sort of 4G, though not equivalent to LTE.
But they really should get rid of the "C:\" convention for disks. Sure, you can do some remapping, but it's homage to the floppy-disk days of MS-DOS.
Cutler's previous OS, VMS, got it right, and better than Unix (dare I say it here!). A drive had a physical name that was based on its hardware: DJA1: But that was normally hidden and mapped to a Logical name, which could refer to any node in a directory tree, including a cluster of disks, or just a directory: SYS$SYSTEM: (which might point to DJA1:[SYSTEM] SLASHFILES: (which might point to DJA2:[CMDRTACO.SLASH]
Applications could then use the logical name, and if drives were added or subtracted, nobody worried about things breaking, so long as the logicals were correctly mapped.
These aren't just telco grants. Two of today's four grants go to rural telcos. One goes to a university consortium, the other to a state agency that will compete with the telcos. The Bells don't play in this pen.
The Telstra-NBN deal illustrates how the telecom industry should be restructured.
In the US, recent policy has moved in exactly the opposite direction, towards more vertical integration, so the telephone companies, who own wires built with monopoly money, don't have to let competing ISPs use them at all. They only have to let competitive phone companies (CLECs) use them under certain circumstances, which are shrinking; this basically is limited to old copper wire in urban areas and town centers.
A "LoopCo" would be a company that owns the outside wire and leases it equally to all comers, building fiber for all who want to rent it, even cable. One fiber plant is a lot easier to afford than two or three. The original NBN plan would have built a new fiber plant to compete with Telstra; as customers moved off of Telstra's old copper network, Telstra would have lost money. Telstra blinked: They're selling their existing plant to NBN, so that they will be the biggest wholesale customer, not a competitor. Telstra wins: They get to use the new network, and get paid A$11B for their old wire. The country wins: They get NBN's new fiber, and don't have to fight Telstra all the way, or pay twice.
The Bells in the US do not see it this way. Nor does the FCC, which is squarely in their pocket. Expect the US to fall farther and farther behind, as the farce called "National Broadband Plan" leads to more of the same, just with higher taxes to subsidize CenturyTel, TDS, and other rural subsidy whores who can use the subsidy money to put local wireless ISPs, who are not eligible for subsidies (only one subsidy recipient in a given place - it's literally a monopoly fund) out of business.
HR is really popular in Russia, where it is called Radiosport. It has a lot of geek-friendly sporting activities. For instance, DXing is trying to talk to as many countries (loosely defined!) as possible. Contests take place many weekends, to try to make as many contacts according to some set of rules (to a given place, to as many lat/long squares, to as many countries, on a certain band, etc.). Some people do fox hunts (hide a transmitter and try to find it with a portable radio and direction antenna).
And yes, there's the ability to just chat with fellow geeks, anywhere, without depending on somebody else's network.
It used to be true that the station license specified a location, and if you operated anywhere else, you were "/3" in code or "portable 3" or whatever district you were in. To get around it, you could pull "secondary" licenses, with separate call signs, at each address. I had a couple of those. But that went away by the early 1980s, or late 1970s. (I've been licensed a lot longer than that.) Now you get one call sign and can use it anywhere in the country, and take it with you when you move across district boundaries. So call districts really only apply to how they assign you a new call sign, based on mailing address.
Back in the 1960s, when the space capsules were being designed, 16k words was a lot. So programmers wrote tight code, optimized down to every instruction. It wasn't perfect but it was fairly easy to examine.
Nowadays memory is cheap. And we have programming techniques that take advantage of it. Object-oriented code, for instance. This speeds up some kinds of programming but it puts more effective distance between the source code and the executable. And the mere act of using an OS API, as noted in the original story, makes the application vulnerable to subtle bugs in the OS.
It's laughable that Toyota could have thoroughly debugged code that was written using modern, large-memory techniques.
Still, it's unfair to single out Toyota. There's too much drive-by-wire in a lot of cars.
If the engine accelerates suddenly and doesn't respond, you turn the ignition key one click to the left, to the off/unlocked position. This is how you listen to the radio when parked, for instance. The key stays in the lock and the steering wheel is unlocked.
WiMAX is optimized for single-frequency (time division duplex) use. It works on single channels in the 2.3, 3.65 or 2.6 GHz bands, for instance. Clearwire has lots of 2.6 licenses.
LTE is optimized for dual-frequency (frequency division duplex) use, as are cell phones. It will eventually replace TDMA (GSM) and CDMA. It will initially coexist with them; the carriers will roll out LTE on some frequencies while preserving their legacy digital networks. This is sort of how the analog-digital transition (and the 2G-3G, for GSM operators; CDMA 2G and 3G are compatible) worked.
Yes, there is dual-frequency WiMAX and there might even be a spec somewhere for single-frequency LTE. But the two specs are similar. Both (this refers to the later "mobile" WiMAX) use OFDMA transmission and multiple antennas (for range or speed). They license some of the same patents. So if a licensee (Clearwire) is sitting on unpaired spectrum, they'll use WiMAX, and if they're paired, they'll used LTE.
And it's thus likely that in practice, WiMAX will act more like the Internet, while LTE, owned mostly by VZW and ATT, will be constrained to "wireless web" crap, charging by the message or picture, restricting the "app" you're allowed to use, etc.
While many routers "support" IPv6, it is software support, not the hardware support for the "fast path" that IPv4 uses for standard packets. IPv6 packets are the slow exceptions. The total packet capacity is low. This isn't noticed yet much because v6 carries roughly 1/100 of 1% (i.e., 1/10,000) of the total traffic of v4, and a lot of that is just IETF dorkwads throwing around experimental packets to show that it can be done.
If GE has part of 4/8, then it's not pre-IANA. Net 4 was one of three Class A's belonging to BBN in the old days. 4 went to Genuity; two, used for government work, were returned. Genuity tanked and its assets went to Level 3, which kept 1/4 of the block.
Right. The transition must fail because at every stage, everyone needs v4 addresses, and thus there will be more v4-reachable destinations, and thus no need to move to v6.
But then v6 is so frabjulously flawed, so stupid in so many ways, that it will fail on its own. It is just a way for Cisco to force costly new hardware on people.
The whole "SUVs are safer" thing was a pile of crap. It managed to make the oil sheiks rich, and their buddies in Washington (recently departed), but it hurt the inhabitants and the planet.
SUVs are dangerous because they have a higher center of gravity than cars. Add the powerful engines and you have a recipe for rollover. So while an SUV is safer than a compact car in a head-on collision, head-on collisions just aren't all that common. SUVs added more deaths from rollover than they saved from their dead weight.
Plus the weight was a zero-sum game. SUV to SUV is like compact to compact. You win if you're the only SUV; you lose when everybody else has one. And you lose when the compact handles better, evades the collision, and the SUV rolls over while trying.
Childish whining like the OP about cable companies' not metering their own television broadcasts or telephone calls, but metering Internet, gets nowhere. You all want cake, and you want it free, and to eat it too. But the cake is a lie.
Cable runs telephone on reserved, engineered capacity (PacketCable) for which subscribers pay a fee. It doesn't touch the Internet; it goes to a media gateway into the phone network.
Cable runs video on many channels, some analog, most QAM nowadays. That's sent from the head end, mostly from satellite feeds, some from over-the-air receivers and ATSC-to-QAM remodulators.
Internet goes on a separate CMTS that goes over middle mile facilities to an ISP backbone. That all costs money. UPSTREAM capacity on cable is VERY limited; it only works upstream to 42 MHz, and broadband only above about 20 MHz. It is a terrible medium for providing content or running file servers, which is what Torrent is about.
So heavy uploaders in particular, and heavy users in general, tax the shared capacity of the Internet and worsen everyone else's usage (gaming response, data performance, etc.). So I'd rather be on a system that invites the heaviest users to go elsewhere, thank you.
Sure they could "buy" more capacity, but why should I pay more so that a handful of bozos can exchange movies? Tiered pricing allows my price, for using under 50 GB/month, to stay reasonable.
I would prefer a free market in ISPs, with DSL still open to any ISP so that there would be an open market. The FCC could fix that. But regulating ISPs per se is a truly, deeply dumb idea.
What a waste. I bought a cheapo no-name cable at our electronic discount parts store, but when I painted both ends of it with green Magic Marker, the bit error rate improved by an order of magnitude and my overall transfer speed went up by 17%!
Maybe I'll write a letter about it to Etherphile magazine.
Slashdot Mirror
No, it's not broccoli. Nor chocolate. It's creamed chipped beef on toast a/k/a sh*t on a shingle. Only on stale toast. You eat it because you are there to take orders. Not to think.
No, NAT will not die. NAT is a good idea, not a bad one. Virtually everyone uses firewalls nowadays, most of which do NAT, which adds a level of security (not enough by itself, but it helps).
It is a critical flaw in TCP/IP architecture that the application translates the name to the address and sees the IP address. And there's never a good reason for applications to have numeric IP addresses inside them. NAT only breaks broken applications. IPv6 is Just Plain Stupid. It's ugly and it wants to die. And it will. The people who are pushing it are the kind of people who seek out authority in order to obey it blindly.
Apple isn't using them because a) they're not out yet, b) they aren't mass-produced the way Apple needs them, and c) Apple has volume contracts for screens with its great friend Samsung.
(Yes, the irony is real -- they are suing Samsung while simultaneously buying tons from them.)
A lot of Toyota engines failed due to a cylinder cooling problem. They were replaced by Toyota, but it was a big costly deal for user and company alike. These were in Camrys and Siennas (minivans).
But my Corolla has been exceptionally reliable. The new ones look really cheap though. The LE of 2012 has a flimsier interior than the CE of 2004.
There still are collocators, just not as many. The FCC adopted policies in 2000-2004 that made it harder to be a CLEC, and many went out of business. There are still some, but they tend to concentrate on business customers, who can pay more. Collocation is also used by some CLECs that provide wholesale interconnection services to VoIP providers. In general, a VoIP provider needs a CLEC to get blocks of numbers and interconnection to the ILEC. Level 3 is probably the biggest wholesale player.
The huge savings in telephone company real estate happened over 20 years ago. Their big buildings were built for electromechanical switching systems, mostly installed between 1920 and 1970. The digital switches mostly installed in the 1980s were a fraction of the size, leaving lots of empty space in the big buildings. Some space has already been repurposed. And some is available, but the Bells don't want to give it up because it would make competition easier.
Most of the real estate still used by telco gear is for line drivers, the stuff needed to run analog phones. Whether these are fed by VoIP or TDM doesn't matter; 90 volt power ring and 48 volt battery take space. They also take power, but home-based analog terminal adapters (local battery) use even more, so centralized power (common battery) is a net savings.
Berninger is simply repeating Cisco memes, that somehow the magic pixie dust of IP makes everything wonderfuler. It's bullshit, but somebody has to call them on it.
The large (Bell and other) telephone companies are not regulated on rate of return any more. They are on "price caps". Only the smallest carriers, the mom'n'pops and subsidy-dependent rural ones, are on rate of return. That's why the Bells have laid off so many people and stopped investing - they are milking their old plant for all it's worth.
Most of these sites have the word "jersey" in them. It looks as if the NFL's licensing squad went a-hunting, and gave the list of unauthorized vendors to Uncle Sam. What's not obvious is whether all of these sites simply sell unauthorized jerseys, or whether other jersey vendors, or people from a certain island or state, also got nailed in the crosfire.
ITU is not the primary standards body for this, and their definition of 4G is irrelevant. Geeknet doesn't write COBOL standards either. 3G was defined by 3GPP (GSM -> WCDMA/UMTS/HSPDA) and 3GPP2 (CDMA2000). All 3G is based on CDMA. Now 3GPP has defined LTE, which is different and newer technology, OFDMA, so it's a genuine generational shift. That's 4G, no matter what the speed. It gets more bits per Hz though.
ATT and T-Mobile are flogging their HDPDA+ (WCDMA) networks as "4G", because under really good conditions, they can get more speed than older forms of 3G. But it's just a late-life kicker for 2000's technology. VZW, to their credit, already has LTE, which is the real 4G, while ATT will have it later. And Sprint/Clearwire's WiMAX is sort of 4G, though not equivalent to LTE.
But they really should get rid of the "C:\" convention for disks. Sure, you can do some remapping, but it's homage to the floppy-disk days of MS-DOS.
Cutler's previous OS, VMS, got it right, and better than Unix (dare I say it here!). A drive had a physical name that was based on its hardware:
DJA1:
But that was normally hidden and mapped to a Logical name, which could refer to any node in a directory tree, including a cluster of disks, or just a directory:
SYS$SYSTEM: (which might point to DJA1:[SYSTEM]
SLASHFILES: (which might point to DJA2:[CMDRTACO.SLASH]
Applications could then use the logical name, and if drives were added or subtracted, nobody worried about things breaking, so long as the logicals were correctly mapped.
Boeing bought Narus in 2010. Boeing is best known for its civilian aircraft, but it is also a huge military contractor.
These aren't just telco grants. Two of today's four grants go to rural telcos. One goes to a university consortium, the other to a state agency that will compete with the telcos. The Bells don't play in this pen.
The Telstra-NBN deal illustrates how the telecom industry should be restructured.
In the US, recent policy has moved in exactly the opposite direction, towards more vertical integration, so the telephone companies, who own wires built with monopoly money, don't have to let competing ISPs use them at all. They only have to let competitive phone companies (CLECs) use them under certain circumstances, which are shrinking; this basically is limited to old copper wire in urban areas and town centers.
A "LoopCo" would be a company that owns the outside wire and leases it equally to all comers, building fiber for all who want to rent it, even cable. One fiber plant is a lot easier to afford than two or three. The original NBN plan would have built a new fiber plant to compete with Telstra; as customers moved off of Telstra's old copper network, Telstra would have lost money. Telstra blinked: They're selling their existing plant to NBN, so that they will be the biggest wholesale customer, not a competitor. Telstra wins: They get to use the new network, and get paid A$11B for their old wire. The country wins: They get NBN's new fiber, and don't have to fight Telstra all the way, or pay twice.
The Bells in the US do not see it this way. Nor does the FCC, which is squarely in their pocket. Expect the US to fall farther and farther behind, as the farce called "National Broadband Plan" leads to more of the same, just with higher taxes to subsidize CenturyTel, TDS, and other rural subsidy whores who can use the subsidy money to put local wireless ISPs, who are not eligible for subsidies (only one subsidy recipient in a given place - it's literally a monopoly fund) out of business.
HR is really popular in Russia, where it is called Radiosport. It has a lot of geek-friendly sporting activities. For instance, DXing is trying to talk to as many countries (loosely defined!) as possible. Contests take place many weekends, to try to make as many contacts according to some set of rules (to a given place, to as many lat/long squares, to as many countries, on a certain band, etc.). Some people do fox hunts (hide a transmitter and try to find it with a portable radio and direction antenna).
And yes, there's the ability to just chat with fellow geeks, anywhere, without depending on somebody else's network.
You're really out of date.
It used to be true that the station license specified a location, and if you operated anywhere else, you were "/3" in code or "portable 3" or whatever district you were in. To get around it, you could pull "secondary" licenses, with separate call signs, at each address. I had a couple of those. But that went away by the early 1980s, or late 1970s. (I've been licensed a lot longer than that.) Now you get one call sign and can use it anywhere in the country, and take it with you when you move across district boundaries. So call districts really only apply to how they assign you a new call sign, based on mailing address.
What does the Toyota source code look like?
Back in the 1960s, when the space capsules were being designed, 16k words was a lot. So programmers wrote tight code, optimized down to every instruction. It wasn't perfect but it was fairly easy to examine.
Nowadays memory is cheap. And we have programming techniques that take advantage of it. Object-oriented code, for instance. This speeds up some kinds of programming but it puts more effective distance between the source code and the executable. And the mere act of using an OS API, as noted in the original story, makes the application vulnerable to subtle bugs in the OS.
It's laughable that Toyota could have thoroughly debugged code that was written using modern, large-memory techniques.
Still, it's unfair to single out Toyota. There's too much drive-by-wire in a lot of cars.
If the engine accelerates suddenly and doesn't respond, you turn the ignition key one click to the left, to the off/unlocked position. This is how you listen to the radio when parked, for instance. The key stays in the lock and the steering wheel is unlocked.
A Prius is a special case, with its electronics.
WiMAX is optimized for single-frequency (time division duplex) use. It works on single channels in the 2.3, 3.65 or 2.6 GHz bands, for instance. Clearwire has lots of 2.6 licenses.
LTE is optimized for dual-frequency (frequency division duplex) use, as are cell phones. It will eventually replace TDMA (GSM) and CDMA. It will initially coexist with them; the carriers will roll out LTE on some frequencies while preserving their legacy digital networks. This is sort of how the analog-digital transition (and the 2G-3G, for GSM operators; CDMA 2G and 3G are compatible) worked.
Yes, there is dual-frequency WiMAX and there might even be a spec somewhere for single-frequency LTE. But the two specs are similar. Both (this refers to the later "mobile" WiMAX) use OFDMA transmission and multiple antennas (for range or speed). They license some of the same patents. So if a licensee (Clearwire) is sitting on unpaired spectrum, they'll use WiMAX, and if they're paired, they'll used LTE.
And it's thus likely that in practice, WiMAX will act more like the Internet, while LTE, owned mostly by VZW and ATT, will be constrained to "wireless web" crap, charging by the message or picture, restricting the "app" you're allowed to use, etc.
While many routers "support" IPv6, it is software support, not the hardware support for the "fast path" that IPv4 uses for standard packets. IPv6 packets are the slow exceptions. The total packet capacity is low. This isn't noticed yet much because v6 carries roughly 1/100 of 1% (i.e., 1/10,000) of the total traffic of v4, and a lot of that is just IETF dorkwads throwing around experimental packets to show that it can be done.
If GE has part of 4/8, then it's not pre-IANA. Net 4 was one of three Class A's belonging to BBN in the old days. 4 went to Genuity; two, used for government work, were returned. Genuity tanked and its assets went to Level 3, which kept 1/4 of the block.
By 1986, every government computer was supposed to support OSI. So OSI backers thought that it would be commonplace.
Uncle's IPv6 mandate is just GOSIP II. But IPv6 makes OSI look like Shakespeare next to its own Ed Wood.
Right. The transition must fail because at every stage, everyone needs v4 addresses, and thus there will be more v4-reachable destinations, and thus no need to move to v6.
But then v6 is so frabjulously flawed, so stupid in so many ways, that it will fail on its own. It is just a way for Cisco to force costly new hardware on people.
The whole "SUVs are safer" thing was a pile of crap. It managed to make the oil sheiks rich, and their buddies in Washington (recently departed), but it hurt the inhabitants and the planet.
SUVs are dangerous because they have a higher center of gravity than cars. Add the powerful engines and you have a recipe for rollover. So while an SUV is safer than a compact car in a head-on collision, head-on collisions just aren't all that common. SUVs added more deaths from rollover than they saved from their dead weight.
Plus the weight was a zero-sum game. SUV to SUV is like compact to compact. You win if you're the only SUV; you lose when everybody else has one. And you lose when the compact handles better, evades the collision, and the SUV rolls over while trying.
Childish whining like the OP about cable companies' not metering their own television broadcasts or telephone calls, but metering Internet, gets nowhere. You all want cake, and you want it free, and to eat it too. But the cake is a lie.
Cable runs telephone on reserved, engineered capacity (PacketCable) for which subscribers pay a fee. It doesn't touch the Internet; it goes to a media gateway into the phone network.
Cable runs video on many channels, some analog, most QAM nowadays. That's sent from the head end, mostly from satellite feeds, some from over-the-air receivers and ATSC-to-QAM remodulators.
Internet goes on a separate CMTS that goes over middle mile facilities to an ISP backbone. That all costs money. UPSTREAM capacity on cable is VERY limited; it only works upstream to 42 MHz, and broadband only above about 20 MHz. It is a terrible medium for providing content or running file servers, which is what Torrent is about.
So heavy uploaders in particular, and heavy users in general, tax the shared capacity of the Internet and worsen everyone else's usage (gaming response, data performance, etc.). So I'd rather be on a system that invites the heaviest users to go elsewhere, thank you.
Sure they could "buy" more capacity, but why should I pay more so that a handful of bozos can exchange movies? Tiered pricing allows my price, for using under 50 GB/month, to stay reasonable.
I would prefer a free market in ISPs, with DSL still open to any ISP so that there would be an open market. The FCC could fix that. But regulating ISPs per se is a truly, deeply dumb idea.
What a waste. I bought a cheapo no-name cable at our electronic discount parts store, but when I painted both ends of it with green Magic Marker, the bit error rate improved by an order of magnitude and my overall transfer speed went up by 17%!
Maybe I'll write a letter about it to Etherphile magazine.