Such mission critical systems should NEVER have untrusted media inserted, and they should NEVER be on the public internet. Further, inserting a media such as a USB stick should be safe because nothing should be automatically run.
How about removing the commodity black-box software, chock full of known vulnerabilities, that is wide open to infection by such paths, replacing it with software where you CAN disable or control such access.
Lots of IT pros have been screaming for a DECADE that only complete fucking morons put a SCADA system on anything that is connected to an external network. Let me repeat that. ONLY A COMPLETE MORON will hook up a scada system to a pc that bridges the internet and the secured network, OR puts the whole damn thing on a unsecured network.
It's not just the network. Malware predates general Internet accessibility by a number of years. The earliest ones were spread by removable disks carried via sneakernet.
"Only a complete moron" would build into a scada system a machine loaded with software that has THOUSANDS of wide-open known ways to infect it, if malware comes in on ANY vector: Network, removable disks, storage sticks, infrared flickering, WiFi signals,...
Such a machine is an agar plate waiting for the first bacterium to land. And a well designed chunk of malware (and this one looks like a masterpiece) can spread from network to machine to storage device to whatever and try them ALL, so that if there is even ONE POSSIBLE PATH it will be found.
Which apparently is what happened to Iran's uranium enrichment system, since reports are that it WASN'T connected to the net.
And that is how we ended up with the FTP over VOIP protocol.
And if the ISP is using QoS with traffic shaping you'll get only some very limited fraction of your data rate with parameters suitable for streaming. The rest will either be dropped or bumped down to "best-effort". (Probably dropped: Even if the ISP DOES have a box smart enough to demote it, you marked it as being useless-if-delayed. So it gets dropped if it's held up, to avoid chewing up backplane datarate with packets that won't be used when received.)
Net result of using FTP over VoIP is to make your REAL VoIP packets lose their priority and/or be dropped, while not appreciably helping (and possibly drastically throttling) your FTP traffic.
= = = =
This was done before - a LONG time ago (in network time scales). QoS / Diffserv was defined in the network protocols back at the start of things. But an early Microsoft product shipped with a network stack that "improved" its performance by marking everything for higher QoS than it needed - long before VoIP applications were common (or even generally available). Net result was that for several generations of buildout the backbone hasn't trusted the user's class-of-service markings - and won't in the foreseeable future, either (until there's some benefit for the ISPs to honoring SOME of them, along with a billing structure that pays for packets that get premium service at the expense of others).
It was this institutionalized "cheating" that led to the development of packet inspection, so ISPs could identify customers' traffic that really did need special service to give the user a good "internet experience" and adjust its handling appropriately (thus giving the ISP an edge on the competition). Of course, like fire, tools to identify traffic type can also be misused to the customers' disadvantage (such as protocols that compete with a product of the ISP's owner). And the requirement for the ISP to pick winning and losing protocols stifles innovation.
The problem, with bullet wound is... they are not always clean...
Another problem is emergency room doctors who believe the myth of "hydrostatic shock" damage and chop out a core of tissue around the bullet's path (like a linear cancer) rather than treating it properly by cleaning and closing it (like a puncture-with-displacement wound).
Yo, Docs! Even if the bullet WERE somehow traveling faster than the speed of sound in tissue (about mach 4.4), shock waves in tissue aren't any big deal! Think "Lithotripter".
Another problem with bullet wounds is emergency room doctors who believe the myth of "hydrostatic shock" damage and chop out a core of tissue around the bullet's path (as if it were a linear cancer), rather than treating it properly by cleaning and closing the wound (as if it were any other puncture-and-displacement trauma).
Yo, Docs! Even if the bullet somehow WAS traveling faster than the speed of sound in flesh (like about mach 4.4) shock waves aren't any big deal for soft tissue. Think Lithotripter.
The spores germinate only in very alkaline environments...
So what's the limit? The well at my NV place has a pH of eight and the water at my townhouse a pH of nine. Will the city's water system and/or my residential well be plugged with bacterial pseudo-cement, strong as the real stuff? (Note that the well casing has a cement wall - just ideal for them to treat the boundary between it and the dirt as a crack and follow it down.)
Lots of alkaline soil out there (like around my townhouse). Before adding soil amendments it was mostly clay - hard but still workable with a tiller. Will these bugs turn it to concrete requiring a jackhammer?
Notice that the growth only occurs if the muscle was injured. This implies that there's a valid biological mechanism inspiring the body builders' slogan "No pain, no gain!"
I seem to recall that the amygdala mediates a LOT more than just fear. Damage to (both of) it apparently produces:... overreaction to all objects, hypoemotionality, loss of fear, hypersexuality, and hyperorality, a condition in which inappropriate objects are placed in the mouth.... an inability to recognize familiar objects, approaching animate and inanimate objects indiscriminately,... loss of fear towards [possibly threatening organisms]... reduction in maternal behaviors towards infants, often physically abusing or neglecting them.... memory loss (amygdala apparently modulates whether memory is referenced), docility,...
See the Wikipedia articles on the amygdala and Klüver-Bucy syndrome.
A person without fear would rapidly become dead, as there would be nothing preventing them from taking crazy risks.
Actually, a person with extremely attenuated fear tends to become psychopathic.
Apparently an occasional dose of fear/love/hate/excitement is rewarding. (Perhaps not having occasional scary/angry/loving/excited events indicates you're not taking enough risks with your life and as a result are probably missing opportunities.) Thus scary movies and TV shows, risky sports like racing, skydiving, mountain climbing, etc.
One observation about psychopaths is that they do not get excited without extreme stimulus, leading to the speculation that they are experiencing a phenomenon akin to sensory deprivation and part of their behavior (risk taking, harming others, seeking extraordinary rewards with no regard to side-effects, etc.) is an attempt to produce a strong enough stimulus to actually feel some activation of this path.
When the dominant model is to buy the phone with the plan, why should the networks pay extra for the millions of phones they ship with plans when the only benefit is to make it easier for the customer to switch to a competitor? Better for them to ship a cheap phone that can't use all the competitors' services.
(the unshielded cable in the images being for suspension only, and the other two conductors for power).
Sorry, that's wrong.
In the US the typical drop has three wires with split-phase 240V. The two insulated wires are the "hot" ones - 120V each to ground, 240V between them. The bare wire is both for support and for the "neutral" wire, which approximately at ground potential (if it hasn't broken, come lose at one end, had a connection corrode,...).
When you aren't pulling the same amount of current from the two hot wires the difference current appears primarily in the neutral wire - which may make it a bit "hotter" than ground due to resistive voltage drop. The neutral will typically be tied to the building grounding system at one point near the service entrance of each house and to the pole grounding system at the base of each pole containing a transformer that feeds it (and sometimes other poles as well). But the resistance of the dirt is high enough that the bulk of the neutral current flows through the drop wire.
Besides, if you split the internet into two pipes, one neutral and one non-neutral, you kill net neutrality because you can prioritize the non-neutral bit over the neutral bit. In other words, you can't be a little bit pregnant.
Which is exactly what they need to do anyhow.
The reason the telecom-ISPs need to be able to prioritize certain packets is because they're trying to "converge" the old connection-based infrastructure into an IP-based backbone. This gives great economies by having only one set of boxes, one set of fibers, and so on.
But the connection-based services include quality-of-service guarantees. Once you have established (and are paying for) a connection you have guaranteed packet delivery, data rate, latency, and jitter. Drop too many packets, delay them, or vary the delay too much, and you've failed to perform. This is very bad.
Now basic IP doesn't make any guarantees on delivery and such. That's acceptable if you limit the bandwidth that can be sent into the backbone to less than it can carry. But you'd rather not enforce those limits. Better to do what the net does currently: Throw 'em into the pool and see if they make it to the other side. If there's a traffic jam in the great water-polo game and some of 'em drown, that's fine - more of 'em still get through.
But that's not acceptable when the packets that DIDN'T get through, or got delayed too much, are the ones for which the carriers wrote contracts giving delivery guarantees.
The common internet protocols, including especially TCP itself, work by increasing the number of packets they send until some don't get through, then backing off and ramping up again. If done properly this does a good job of fairly splitting the available data rate among users and using nearly all of it. But it drops OTHER protocols' packets, too. So it guarantees that things like file transfers break the connection-based service guarantees. (It also intermittently fouls up streaming, like VoIP and video.) The two approaches don't play well together in the "fair" environment.
Now the routing equipment COULD split the data rate of the connections between two (or more) services - say a "telecom" service and an "open internet" service - and give each a fixed fraction. Then the packets with guarantees could run in their "private" fraction and meet the requirements, while the "open net" did what the current net does on its fraction. The "open net" would get the bulk of the bandwidth, because the pricier guaranteed stuff is a small part of the total load. And nobody would be arguing with the carriers doing whatever they want with their special packets because they'd be in the special part of the net.
But doing it that way is more expensive than a unified solution. And the bulk of the "telecom" chunk would be idle most of the time. Wouldn't it be better to use that for more internet data?
Problem with doing it that way is that, when the packets with guarantees come along, you have to treat them better than the generic packets. Oops! Now you're not "neutral" by some definitions: You're treating some packets better than others. "Best effort" delivery is no longer your best effort - some packets get treated even better.
One solution is to go ahead and do it that way, marking some packets for special treatment and only allow packets for "connections" that have negotiated their reservations get the magic mark. Then the "best effort" / "fair and neutral internet" gets to use that extra bandwidth when it's not in use for the telephone-like connection service. Good! And it's easy to do it this way. The carrier can also contract with OTHER carriers to let THEM make connections using that reserved bandwidth, and thus emulate the connection-based networks correctly. They can also let internet customers make connections and admit packets up to the agreed data rate, so the service isn't limited to just themselves and other telephone companies. This is called QoS (quality-of-service) marking or DI
Interesting that EDS also processes the insurance claims for Medicare.
By the way: Perot sold EDS a while back. To GM.
I used to think that he shafted GM three times with EDS:
- Once when he convinced them to outsource all their programming and IT to EDS and fire all their non-EDS software consultants. (This in the month before cutting over payroll to a new system at a major plant. LOTS of extra cost - including buying a mainframe for the consultants who had done the work on the payroll, and who refused to walk back into an EDS shop to finish the job for 'em. B-) Lots of other costs with other consultants, too.)
- Once where he sold it to 'em in return for a bunch of stock and a seat on the board.
- Once when he made him pay him more to go away. B-)
But the extra costs to 'em for various consultants who now have to work through middlemen thanks to this mined-harbor business counts as a fourth one. And it wouldn't surprise me if these taken together and/or something else he did contributed to their bankruptcy, which would make it five.
The restrictions which have existed (until Obama overturned them) regard limiting federally funded research to certain pre-existent lines.
If "federally funded" includes "done in a facility which has built part of its infrastructure or is supporting part of its common infrastructure with federal funding", and essentially all organizations capable of doing the work and making the results available publicly (such as university medical centers) meet that definition, how is that different from a federal ban?
Sometimes walk buttons do something. I do know some traffic lights around Austin which will have reds all four ways if the buttons are pressed.
And sometimes they turn on the walk signal and hold the cross-traffic stopped for a long enough time for a pedestrian to cross. Sometimes they also are necessary to even get the light to give the main drag a red with no car traffic on the cross street.
A few elevator "close door" buttons remove or shorten the delay before the door closes. (I've been assuming the rest only did anything useful in the elevator test or fire department modes, or were part of a standard panel design that was used even with controllers that {currently} didn't pay any attention to them.)
It's not just healthcare that keeps people from starting up their own businesses.
There's also the 1980-ish inversion of the "Safe Harbor" provision of the tax code.
In many occupations - from programming to maid service - it is essentially impossible to work directly for a client. You must work through a middleman you don't own.
No matter how much documentation, contractual paperwork, and external stuff you've got in place (office, health insurance, DBA or incorporation, etc.), if you don't keep up your tax payments the IRS will go to your client, say you were really an employee, and demand the withholdings, half the social-security payments, plus penalties and interest. Even if you work through a middleman corporation, if you and/or your family members own half or more of it the IRS will treat it as a scam and non-existent.
So clients mostly won't deal with you unless you're a serf of a company you don't control.
This got flipped about 30 years ago. Before that the "safe harbor" provision of the tax code said essentially that if your client contracted with you in good faith only you were on the hook for your taxes. The flip was due to lobbying by Ross Perot, owner of EDS, a big computer consulting firm.
Those given the current from right to left across the parietal lobe did significantly better when given, compared to those who were given no electrical stimulation. The direction of the current was important -- those given stimulation running in the opposite direction, left to right, did markedly worse at these puzzles than those given no current, with their ability matching that of an average six-year-old.
Right-to-left good, left-to-right bad, eh?
But was this electrons going right-to-left? Or was it classical current (direction of hypothetical positive charge carriers, thanks to Franklin's wrong hypothetical coin flip). It would clearly make a LOT of difference.
(Which also brings up the question of what the charge carriers in the brain are. Cue the "holes in the head" jokes in 3, 2, 1...)
Actually, they were written to go after anybody that violated them, and whack them with draconian penalties to make up for the low probability of whacking any particular mole.
That's how they are written now.
It's also how they were written and interpreted as of 1974, when CONTU was formed to study the existing US copyright law and recommend whether/how it should be applied to/modified to cover software source, object code, databases, other compositions and art "fixed in computer media", photocopying, and other "New Technological Uses". They described the choirmaster precedent in their draft (and, I presume, final) report.
As I've pointed out else where in this thread, it was only in 1997 with the passage of the No-Electronic Theft Act that non-commercial infringement (i.e. trading copies for other copies) was made a criminal offense.
No argument there. But I'm talking about the civil law. My point was that, at least according to CONTU's interpretation of the precedent, the civil statutory damages were intended to be punitive and/or compensatory in the aggregate, so the low probability of whacking any particular mole didn't lead to a situation where, on the average, the moles profit despite some of them occasionally being whacked. (Think of the drug war model.)
(Please note I'm not endorsing CONTU's conclusions, either. Just pointing out that draconian statutory penalties for even relatively innocent, low-volume, non-commercial copyright violation is NOT new.)
I guess that ITU - the organization that defines what constitutes as 4G and what doesn't - does know what 4G means.
The ITU has no more standing to define what 4G means for advertising purposes within the US than you or I do. They can define what it means within their standards regime, but so what? The English language (unlike French) does not have an official standards board.
"nG" is a set of buzzwords used by executives roadmapping their network equipment offerings and network designs. It is not subject to standardization by the ITU - or ANSI, or ISO, or Bellcore, or IEEE, or IETF, or CEN, or ETSI, or (I could go on for pages).
If the ITU standard ever becomes sufficiently adopted in US telecom that ITU-4G is the generally recognized common form of what is sold in the US as "4G service", a competitor that conforms might sue one that does not, or a customer might claim he was defrauded when he was sold something that didn't conform. But with a service that has been advertised as "4G" being the only one sold that way, and being sold that way before the ITU standard was promulgated, it will be hard to claim fraud or false advertising against either that vendor or any other with a comparable product.
The ITU is an INTERnational standards organization promulgating standards for interconnection of telephone equipment. Telephone equipment manufacturers pay attention to their regulations because the carriers want equipment that works together and often puts ITU standards in their requirements when they ask vendors to bid on supplying them with equipment.
That does not give it any standing at all in the US domestic advertising market, or in trademark law. (If ANSI had done it they MIGHT have had more clout...)
Because one US company telecom has used the term in its advertising (to ITS customers) BEFORE ITU promulgated a definition (for the purposes of labeling equipment in situations where the TELECOMS are the customers) it will be easy for their competitors to argue that the term means what the first user meant by it and that any comparable offering is properly labeled with the same term so that consumers can compare them.
IMHO this won't come up anywhere unless/until some competitor of BOTH Clearwire/Sprint and Verizon rolls out a network that DOES meet the ITU definition. Then they'll put out an ad campaign claiming they have REAL 4G and their competitors only have fake, accept no substitutes...
There was never supposed to be a major nasty punishment in law for noncommercial activity. Copyright laws weren't written to assault someone who photocopies a chapter of a book to study from, or even the whole book. They were written to go after the guy who set up a printing press in a warehouse, or an LP-pressing machine, or a CD-burning machine, started burning bootleg copies and then selling them on the street corner or somewhere else for profit while undercutting the copyright owner.
Actually, they were written to go after anybody that violated them, and whack them with draconian penalties to make up for the low probability of whacking any particular mole.
The classic precedent is a church choir director who bought sheet music for a song that was scored for voices far squeakier than his choir, transposed it down to make it singable by ordinary people, ran off a few copies for his choir, and sent the modification back to the publisher, gratis, to be used in future editions if the publisher chose to do so. He got whacked for the statutory penalty big bucks (which were worth a lot more in those days) for each of the copies of the derived work he made.
As to "the guy who sets up... in a warehouse... a CD-burning machine... started burning bootleg copies and then selling them... for profit while undercutting the copyright owner", it also applies to the guy who does the same except he gives them away. THAT's what she's charged with. The file-sharing software made them available for some potentially large number of downloads by others.
I don't like it either. But let's not misconstrue the law or the situation. It confuses the issue.
I've seen induction heating used to temper truck axles, among other things.
I saw it used in the manufacture of commutators for starter motors:
- Copper bar bent into circle.
- Induction heat to orange in about 5 seconds, to weld the joint and take the stresses out of it. Result: Stress-free donut.
- Smash the donut into shape (segmented hollow top-hat) with dies.
- Mold plastic into it - to support it and make an insulated press-fit for the shaft).
- Saw the segments apart.
I've been trying to figure out how to make slip rings for a windmill. Seems like bending, welding, and annealing a copper bar would do the trick. And an induction hot plate ought to be just the ticket for the welding/annealing step.
Slashdot Mirror
I do not mind the government telling industry that they must secure their systems. Who else is going to do that? Customers?
Stockholders. B-)
Such mission critical systems should NEVER have untrusted media inserted, and they should NEVER be on the public internet. Further, inserting a media such as a USB stick should be safe because nothing should be automatically run.
How about removing the commodity black-box software, chock full of known vulnerabilities, that is wide open to infection by such paths, replacing it with software where you CAN disable or control such access.
Belt and suspenders.
Lots of IT pros have been screaming for a DECADE that only complete fucking morons put a SCADA system on anything that is connected to an external network. Let me repeat that. ONLY A COMPLETE MORON will hook up a scada system to a pc that bridges the internet and the secured network, OR puts the whole damn thing on a unsecured network.
It's not just the network. Malware predates general Internet accessibility by a number of years. The earliest ones were spread by removable disks carried via sneakernet.
"Only a complete moron" would build into a scada system a machine loaded with software that has THOUSANDS of wide-open known ways to infect it, if malware comes in on ANY vector: Network, removable disks, storage sticks, infrared flickering, WiFi signals, ...
Such a machine is an agar plate waiting for the first bacterium to land. And a well designed chunk of malware (and this one looks like a masterpiece) can spread from network to machine to storage device to whatever and try them ALL, so that if there is even ONE POSSIBLE PATH it will be found.
Which apparently is what happened to Iran's uranium enrichment system, since reports are that it WASN'T connected to the net.
And that is how we ended up with the FTP over VOIP protocol.
And if the ISP is using QoS with traffic shaping you'll get only some very limited fraction of your data rate with parameters suitable for streaming. The rest will either be dropped or bumped down to "best-effort". (Probably dropped: Even if the ISP DOES have a box smart enough to demote it, you marked it as being useless-if-delayed. So it gets dropped if it's held up, to avoid chewing up backplane datarate with packets that won't be used when received.)
Net result of using FTP over VoIP is to make your REAL VoIP packets lose their priority and/or be dropped, while not appreciably helping (and possibly drastically throttling) your FTP traffic.
= = = =
This was done before - a LONG time ago (in network time scales). QoS / Diffserv was defined in the network protocols back at the start of things. But an early Microsoft product shipped with a network stack that "improved" its performance by marking everything for higher QoS than it needed - long before VoIP applications were common (or even generally available). Net result was that for several generations of buildout the backbone hasn't trusted the user's class-of-service markings - and won't in the foreseeable future, either (until there's some benefit for the ISPs to honoring SOME of them, along with a billing structure that pays for packets that get premium service at the expense of others).
It was this institutionalized "cheating" that led to the development of packet inspection, so ISPs could identify customers' traffic that really did need special service to give the user a good "internet experience" and adjust its handling appropriately (thus giving the ISP an edge on the competition). Of course, like fire, tools to identify traffic type can also be misused to the customers' disadvantage (such as protocols that compete with a product of the ISP's owner). And the requirement for the ISP to pick winning and losing protocols stifles innovation.
Thanks again, Microsoft.
The problem, with bullet wound is... they are not always clean...
Another problem is emergency room doctors who believe the myth of "hydrostatic shock" damage and chop out a core of tissue around the bullet's path (like a linear cancer) rather than treating it properly by cleaning and closing it (like a puncture-with-displacement wound).
Yo, Docs! Even if the bullet WERE somehow traveling faster than the speed of sound in tissue (about mach 4.4), shock waves in tissue aren't any big deal! Think "Lithotripter".
The problem, with bullet wound is...[dirt].
Another problem with bullet wounds is emergency room doctors who believe the myth of "hydrostatic shock" damage and chop out a core of tissue around the bullet's path (as if it were a linear cancer), rather than treating it properly by cleaning and closing the wound (as if it were any other puncture-and-displacement trauma).
Yo, Docs! Even if the bullet somehow WAS traveling faster than the speed of sound in flesh (like about mach 4.4) shock waves aren't any big deal for soft tissue. Think Lithotripter.
The spores germinate only in very alkaline environments...
So what's the limit? The well at my NV place has a pH of eight and the water at my townhouse a pH of nine. Will the city's water system and/or my residential well be plugged with bacterial pseudo-cement, strong as the real stuff? (Note that the well casing has a cement wall - just ideal for them to treat the boundary between it and the dirt as a crack and follow it down.)
Lots of alkaline soil out there (like around my townhouse). Before adding soil amendments it was mostly clay - hard but still workable with a tiller. Will these bugs turn it to concrete requiring a jackhammer?
I could go on...
Notice that the growth only occurs if the muscle was injured. This implies that there's a valid biological mechanism inspiring the body builders' slogan "No pain, no gain!"
I seem to recall that the amygdala mediates a LOT more than just fear. Damage to (both of) it apparently produces: ... overreaction to all objects, hypoemotionality, loss of fear, hypersexuality, and hyperorality, a condition in which inappropriate objects are placed in the mouth. ... an inability to recognize familiar objects, approaching animate and inanimate objects indiscriminately, ... loss of fear towards [possibly threatening organisms] ... reduction in maternal behaviors towards infants, often physically abusing or neglecting them. ... memory loss (amygdala apparently modulates whether memory is referenced), docility, ...
See the Wikipedia articles on the amygdala and Klüver-Bucy syndrome.
A person without fear would rapidly become dead, as there would be nothing preventing them from taking crazy risks.
Actually, a person with extremely attenuated fear tends to become psychopathic.
Apparently an occasional dose of fear/love/hate/excitement is rewarding. (Perhaps not having occasional scary/angry/loving/excited events indicates you're not taking enough risks with your life and as a result are probably missing opportunities.) Thus scary movies and TV shows, risky sports like racing, skydiving, mountain climbing, etc.
One observation about psychopaths is that they do not get excited without extreme stimulus, leading to the speculation that they are experiencing a phenomenon akin to sensory deprivation and part of their behavior (risk taking, harming others, seeking extraordinary rewards with no regard to side-effects, etc.) is an attempt to produce a strong enough stimulus to actually feel some activation of this path.
When the dominant model is to buy the phone with the plan, why should the networks pay extra for the millions of phones they ship with plans when the only benefit is to make it easier for the customer to switch to a competitor? Better for them to ship a cheap phone that can't use all the competitors' services.
On paper, nice concept. But just one minor problem: the power grid's most likely down, or was never built. I served in Iraq ...
But it should work just fine if the military was suppressing a domestic uprising in the US. B-b
(the unshielded cable in the images being for suspension only, and the other two conductors for power).
Sorry, that's wrong.
In the US the typical drop has three wires with split-phase 240V. The two insulated wires are the "hot" ones - 120V each to ground, 240V between them. The bare wire is both for support and for the "neutral" wire, which approximately at ground potential (if it hasn't broken, come lose at one end, had a connection corrode, ...).
When you aren't pulling the same amount of current from the two hot wires the difference current appears primarily in the neutral wire - which may make it a bit "hotter" than ground due to resistive voltage drop. The neutral will typically be tied to the building grounding system at one point near the service entrance of each house and to the pole grounding system at the base of each pole containing a transformer that feeds it (and sometimes other poles as well). But the resistance of the dirt is high enough that the bulk of the neutral current flows through the drop wire.
Besides, if you split the internet into two pipes, one neutral and one non-neutral, you kill net neutrality because you can prioritize the non-neutral bit over the neutral bit. In other words, you can't be a little bit pregnant.
Which is exactly what they need to do anyhow.
The reason the telecom-ISPs need to be able to prioritize certain packets is because they're trying to "converge" the old connection-based infrastructure into an IP-based backbone. This gives great economies by having only one set of boxes, one set of fibers, and so on.
But the connection-based services include quality-of-service guarantees. Once you have established (and are paying for) a connection you have guaranteed packet delivery, data rate, latency, and jitter. Drop too many packets, delay them, or vary the delay too much, and you've failed to perform. This is very bad.
Now basic IP doesn't make any guarantees on delivery and such. That's acceptable if you limit the bandwidth that can be sent into the backbone to less than it can carry. But you'd rather not enforce those limits. Better to do what the net does currently: Throw 'em into the pool and see if they make it to the other side. If there's a traffic jam in the great water-polo game and some of 'em drown, that's fine - more of 'em still get through.
But that's not acceptable when the packets that DIDN'T get through, or got delayed too much, are the ones for which the carriers wrote contracts giving delivery guarantees.
The common internet protocols, including especially TCP itself, work by increasing the number of packets they send until some don't get through, then backing off and ramping up again. If done properly this does a good job of fairly splitting the available data rate among users and using nearly all of it. But it drops OTHER protocols' packets, too. So it guarantees that things like file transfers break the connection-based service guarantees. (It also intermittently fouls up streaming, like VoIP and video.) The two approaches don't play well together in the "fair" environment.
Now the routing equipment COULD split the data rate of the connections between two (or more) services - say a "telecom" service and an "open internet" service - and give each a fixed fraction. Then the packets with guarantees could run in their "private" fraction and meet the requirements, while the "open net" did what the current net does on its fraction. The "open net" would get the bulk of the bandwidth, because the pricier guaranteed stuff is a small part of the total load. And nobody would be arguing with the carriers doing whatever they want with their special packets because they'd be in the special part of the net.
But doing it that way is more expensive than a unified solution. And the bulk of the "telecom" chunk would be idle most of the time. Wouldn't it be better to use that for more internet data?
Problem with doing it that way is that, when the packets with guarantees come along, you have to treat them better than the generic packets. Oops! Now you're not "neutral" by some definitions: You're treating some packets better than others. "Best effort" delivery is no longer your best effort - some packets get treated even better.
One solution is to go ahead and do it that way, marking some packets for special treatment and only allow packets for "connections" that have negotiated their reservations get the magic mark. Then the "best effort" / "fair and neutral internet" gets to use that extra bandwidth when it's not in use for the telephone-like connection service. Good! And it's easy to do it this way. The carrier can also contract with OTHER carriers to let THEM make connections using that reserved bandwidth, and thus emulate the connection-based networks correctly. They can also let internet customers make connections and admit packets up to the agreed data rate, so the service isn't limited to just themselves and other telephone companies. This is called QoS (quality-of-service) marking or DI
Interesting that EDS also processes the insurance claims for Medicare.
By the way: Perot sold EDS a while back. To GM.
I used to think that he shafted GM three times with EDS:
- Once when he convinced them to outsource all their programming and IT to EDS and fire all their non-EDS software consultants. (This in the month before cutting over payroll to a new system at a major plant. LOTS of extra cost - including buying a mainframe for the consultants who had done the work on the payroll, and who refused to walk back into an EDS shop to finish the job for 'em. B-) Lots of other costs with other consultants, too.)
- Once where he sold it to 'em in return for a bunch of stock and a seat on the board.
- Once when he made him pay him more to go away. B-)
But the extra costs to 'em for various consultants who now have to work through middlemen thanks to this mined-harbor business counts as a fourth one. And it wouldn't surprise me if these taken together and/or something else he did contributed to their bankruptcy, which would make it five.
The restrictions which have existed (until Obama overturned them) regard limiting federally funded research to certain pre-existent lines.
If "federally funded" includes "done in a facility which has built part of its infrastructure or is supporting part of its common infrastructure with federal funding", and essentially all organizations capable of doing the work and making the results available publicly (such as university medical centers) meet that definition, how is that different from a federal ban?
Sometimes walk buttons do something. I do know some traffic lights around Austin which will have reds all four ways if the buttons are pressed.
And sometimes they turn on the walk signal and hold the cross-traffic stopped for a long enough time for a pedestrian to cross. Sometimes they also are necessary to even get the light to give the main drag a red with no car traffic on the cross street.
A few elevator "close door" buttons remove or shorten the delay before the door closes. (I've been assuming the rest only did anything useful in the elevator test or fire department modes, or were part of a standard panel design that was used even with controllers that {currently} didn't pay any attention to them.)
It's not just healthcare that keeps people from starting up their own businesses.
There's also the 1980-ish inversion of the "Safe Harbor" provision of the tax code.
In many occupations - from programming to maid service - it is essentially impossible to work directly for a client. You must work through a middleman you don't own.
No matter how much documentation, contractual paperwork, and external stuff you've got in place (office, health insurance, DBA or incorporation, etc.), if you don't keep up your tax payments the IRS will go to your client, say you were really an employee, and demand the withholdings, half the social-security payments, plus penalties and interest. Even if you work through a middleman corporation, if you and/or your family members own half or more of it the IRS will treat it as a scam and non-existent.
So clients mostly won't deal with you unless you're a serf of a company you don't control.
This got flipped about 30 years ago. Before that the "safe harbor" provision of the tax code said essentially that if your client contracted with you in good faith only you were on the hook for your taxes. The flip was due to lobbying by Ross Perot, owner of EDS, a big computer consulting firm.
Those given the current from right to left across the parietal lobe did significantly better when given, compared to those who were given no electrical stimulation. The direction of the current was important -- those given stimulation running in the opposite direction, left to right, did markedly worse at these puzzles than those given no current, with their ability matching that of an average six-year-old.
Right-to-left good, left-to-right bad, eh?
But was this electrons going right-to-left? Or was it classical current (direction of hypothetical positive charge carriers, thanks to Franklin's wrong hypothetical coin flip). It would clearly make a LOT of difference.
(Which also brings up the question of what the charge carriers in the brain are. Cue the "holes in the head" jokes in 3, 2, 1...)
It's also how they were written and interpreted as of 1974, when CONTU was formed to study the existing US copyright law and recommend whether/how it should be applied to/modified to cover software source, object code, databases, other compositions and art "fixed in computer media", photocopying, and other "New Technological Uses". They described the choirmaster precedent in their draft (and, I presume, final) report.
As I've pointed out else where in this thread, it was only in 1997 with the passage of the No-Electronic Theft Act that non-commercial infringement (i.e. trading copies for other copies) was made a criminal offense.
No argument there. But I'm talking about the civil law. My point was that, at least according to CONTU's interpretation of the precedent, the civil statutory damages were intended to be punitive and/or compensatory in the aggregate, so the low probability of whacking any particular mole didn't lead to a situation where, on the average, the moles profit despite some of them occasionally being whacked. (Think of the drug war model.)
(Please note I'm not endorsing CONTU's conclusions, either. Just pointing out that draconian statutory penalties for even relatively innocent, low-volume, non-commercial copyright violation is NOT new.)
I guess that ITU - the organization that defines what constitutes as 4G and what doesn't - does know what 4G means.
The ITU has no more standing to define what 4G means for advertising purposes within the US than you or I do. They can define what it means within their standards regime, but so what? The English language (unlike French) does not have an official standards board.
"nG" is a set of buzzwords used by executives roadmapping their network equipment offerings and network designs. It is not subject to standardization by the ITU - or ANSI, or ISO, or Bellcore, or IEEE, or IETF, or CEN, or ETSI, or (I could go on for pages).
If the ITU standard ever becomes sufficiently adopted in US telecom that ITU-4G is the generally recognized common form of what is sold in the US as "4G service", a competitor that conforms might sue one that does not, or a customer might claim he was defrauded when he was sold something that didn't conform. But with a service that has been advertised as "4G" being the only one sold that way, and being sold that way before the ITU standard was promulgated, it will be hard to claim fraud or false advertising against either that vendor or any other with a comparable product.
The ITU is an INTERnational standards organization promulgating standards for interconnection of telephone equipment. Telephone equipment manufacturers pay attention to their regulations because the carriers want equipment that works together and often puts ITU standards in their requirements when they ask vendors to bid on supplying them with equipment.
That does not give it any standing at all in the US domestic advertising market, or in trademark law. (If ANSI had done it they MIGHT have had more clout...)
Because one US company telecom has used the term in its advertising (to ITS customers) BEFORE ITU promulgated a definition (for the purposes of labeling equipment in situations where the TELECOMS are the customers) it will be easy for their competitors to argue that the term means what the first user meant by it and that any comparable offering is properly labeled with the same term so that consumers can compare them.
IMHO this won't come up anywhere unless/until some competitor of BOTH Clearwire/Sprint and Verizon rolls out a network that DOES meet the ITU definition. Then they'll put out an ad campaign claiming they have REAL 4G and their competitors only have fake, accept no substitutes ...
There was never supposed to be a major nasty punishment in law for noncommercial activity. Copyright laws weren't written to assault someone who photocopies a chapter of a book to study from, or even the whole book. They were written to go after the guy who set up a printing press in a warehouse, or an LP-pressing machine, or a CD-burning machine, started burning bootleg copies and then selling them on the street corner or somewhere else for profit while undercutting the copyright owner.
Actually, they were written to go after anybody that violated them, and whack them with draconian penalties to make up for the low probability of whacking any particular mole.
The classic precedent is a church choir director who bought sheet music for a song that was scored for voices far squeakier than his choir, transposed it down to make it singable by ordinary people, ran off a few copies for his choir, and sent the modification back to the publisher, gratis, to be used in future editions if the publisher chose to do so. He got whacked for the statutory penalty big bucks (which were worth a lot more in those days) for each of the copies of the derived work he made.
As to "the guy who sets up ... in a warehouse ... a CD-burning machine ... started burning bootleg copies and then selling them ... for profit while undercutting the copyright owner", it also applies to the guy who does the same except he gives them away. THAT's what she's charged with. The file-sharing software made them available for some potentially large number of downloads by others.
I don't like it either. But let's not misconstrue the law or the situation. It confuses the issue.
Production lines in other countries don't incur the cost of US worker-safety regulations.
I've seen induction heating used to temper truck axles, among other things.
I saw it used in the manufacture of commutators for starter motors:
- Copper bar bent into circle.
- Induction heat to orange in about 5 seconds, to weld the joint and take the stresses out of it. Result: Stress-free donut.
- Smash the donut into shape (segmented hollow top-hat) with dies.
- Mold plastic into it - to support it and make an insulated press-fit for the shaft).
- Saw the segments apart.
I've been trying to figure out how to make slip rings for a windmill. Seems like bending, welding, and annealing a copper bar would do the trick. And an induction hot plate ought to be just the ticket for the welding/annealing step.