I agree these are nice speakers - I have a Monsoon MM-700 system on my desk.
However these are not "the speakers you'll find in Humvees". They're sold under the same Monsoon brand, but the flat-panel computer speakers are designed and manufactured by a company called Sonigistix in Richmond, BC, Canada. Last I checked, none of the Sonigistix flat-panel technology had yet made it into an automotive application.
The Monsoon speakers are based on planar-magnetic technology, partially licensed from a company called Eminent Technology. Sonigistix took the basic design, and adapted it so that the speakers could be mass-produced and sold at a lower price point. Note that there are several other flat-panel computer speakers on the market that do not use planar-magnetic technology, but instead use a "distributed-mode" technology developed by a company called "NXT". In my (biased, as a former Sonigistix employee) opinion, the planar-magnetic ones sound a lot better.
Similar systems are available commercially. Here is one that:
[...]will produce a maximum power of 12 watts. This system contains a unique DC/DC boost voltage converter and charge control adjustable from 5 to 14.8 V.and can be used to charge a battery and provide over 200 watt-hr per day of energy for various uses. It can be installed in 6" and up, firetube type heating systems with a minimum 125,000 btu. rating and is completely hidden with-in existing equipment.
"Eco-Drive Thermo converts the temperature difference between the user's body and the surrounding air into electrical energy to power the movement. [...] The original Eco-Drive Thermo was launched to great acclaim at the 1999 Basel Show."
Don't know if it's shipping to consumers yet, but the technology's been around for a while.
Several years ago I worked in a university lab that had a prototype of a 1 kW sulfur microwave lamp. It was very bright, but the light had a distinctive green tint. After a while your eyes would adapt, but then when you looked out the window (or at anything lit with a "conventional" light source) everything looked pink. Maybe the newer models have solved this problem, but if not I wouldn't want it as indoor lighting.
As for getting your skin colouring, one of the advantages of the sulfur lamp was that it put out much less UV radiation than other light sources. Good for museums where the UV would damage old paintings and documents; bad for getting a tan in your cubicle.
A straight port of CygnusED would be difficult, as it was a commercial product and the source code is probably long-buried by now. Also, it probably used a lot of Amiga-specific features that wouldn't map well to Linux.
However, I've found NEdit to be an acceptable replacement. It has many of the same features as CygnusED, such as the ability to cut-and-paste a column of text. It shouldn't be that hard to patch in the rest of CED's essential features.
(and just so I'm not completely off-topic, let me add that vi[m] is still my editor of choice for non-GUI situations, such as configuring a 'headless' server in a remote location over a 9600-baud serial line).
The largest recorded since we've started measuring these things was an X20 in 1989. Quebec's power system overloaded that year.
According to this table, the 1989 flare is tied for #1 with one from April of this year (at least, in terms of X-ray intensity). However I don't remember hearing about any significant power or communication disruptions from the April flare.
If you go here, you can get source code for a simple multicast client and server program. Run a sender on one PC, a receiver on another, then add as many intermediate routers and boxes as you want.
I am presently working for a company that sells real-time stock market information, and the incoming data feeds from NASDAQ and SIAC (=NYSE) are in multicast format delivered over a private network. So I have a bit of experience with it, although our application doesn't need anything fancy like dynamic routing or supporting a large number of listening clients. I'm using SuSE Linux for some of the application servers, but I've found that OpenBSD is a better solution for routing the data, tunelling it over a secure VPN, etc.
p.s. If you want to know if your ISP supports multicast, just run tcpdump on your gateway and look for multicast or IGMP traffic.
I understand your point. However, one difficulty with modern cryptography is that it's technically very hard to give up "just enough" of your freedom.
You can't just weaken the algorithms (e.g. by only allowing 40-bit keys). Sure it makes it easier for the "good guys" to intercept the communications of the "bad guys", but the converse is also true. Even the terrorists will be able to listen in on your business communications, find out what flight your company's president is taking to the trade show, etc.
The only option that makes sense is to require government access to keys. In this scenario, the algorithms operate at full strength and users are secure against other users. However, if the government agencies want to investigate you, they will have a copy of your key on file.
However, there are problems. Under this scenario, the government has to have your key on file in advance of when they want to use it. It's no good for them to ask you for your key after they begin to suspect you; you'd know they were on to you.
Now this means that you have to send to the government a copy of every cryptographic key you ever generate. As it happens, this week I've been working on setting up some software-based VPNs to tunnel corporate data across the Internet. I've got a half-dozen OpenBSD boxes running OpenSSH, IPSec, and so on. There are randomly-generated (temporary) keys all over the place. These are just development boxes and will probably get wiped clean in a couple of weeks when I install the real servers. However, right now, each of these boxes is capable of creating a secure communications channel across the Internet. Therefore, under certain legal conditions, I would be required to register all of the keys with a central government agency.
When you're dealing with government access to keys, it's not good enough to have 99% compliance with the regulations. Your targets, the terrorists, are going to be in that last 1%. So you need to ensure 100% compliance with the regulations.
One way to do this is to allow only pre-approved, secure crypto implementations that have an automatic and tamper-proof key-forwarding mechanism. Guess what, you've just made it illegal for users to use any open-source crypto software. You've also made it illegal to run any open-source operating systems or hardware emulators, because these give you enough control that you could modify the operation of any software-based cryptographic utility (think of running a program in a VMWare sandbox, or single-stepping through it in an emulator, and modifying the key data that it sends to the government agency). Suddenly, the only legal "PC" would be a sealed box with a factory-installed Microsoft operating system, on a "CPRM"-style hard drive that couldn't be re-formatted. General-purpose compilers and scripting languages would have to be outlawed, so that ultimately the "PC" could only be used for online shopping, games, or listening to "pay-per-view" downloaded musical content (from RIAA-approved central servers, of course).
If you allow people to keep on using general-purpose computers, the only way to ensure that the government has access to your keys is to randomly audit the process, to try decrypting the messages of a user, and ensuring that you get a legible plain-text. If you don't, you have to haul that user in for a severe punishment (severe enough that it scares everyone else into making sure the government always has an up-to-date version of their keys).
So now you have a scenario where otherwise-innocent citizens have to be audited, and severely punished for the "crime" of not proactively ensuring that the central government agency has a copy of every random key they might ever happen to generate in the course of their daily lives. And you have to audit almost every citizen if you want to have a good chance of catching the terrorists (who make up an extremely small fraction of the population).
And if that's not bad enough, remember that most governments are not 100% free of corruption. Maybe some bored clerk in the crypto bureau will decide he wants to read the full plaintext of the day's decryption-compliance audits, rather than just feeding them to the AI system (maybe Carnivore, maybe just a simple algorithm to see if the decryption result looks like a standard document format). Or maybe the mafia pays off a clerk to deliver them a copy of your financial records, so they can blackmail you. Or maybe a large international aerospace corporation convinces the governenment agencies to let them peek at your bid on a large contract, so they can undercut your company and get the business.
It's a very slippery slope, when you start trying to introduce a "controlled leak" into a system that is inherently strong and secure. That's why I advocate a strong right to privacy, even though I feel the government may be morally entitled to read my "private" communications as long as due process is followed (present their case before a judge, get a 'search warrant' if they can demonstrate probable cause, etc).
I'd be interested to hear if anyone knows of technical solutions that would allow a "reasonable compromise" between freedom and security. Otherwise, I think we just have to take it as a modern "law of nature" that people can communicate privately without the possibility of interception. We as a society have to grow up. We're not children any more, and our "parents" (governemnts) can't protect us from all the evils of the world any more. We've all gotten too good at killing and hurting each other for the "parental" safeguards to work any more. All we can do is recognize our mutually vulnerable nature, and try to work together to build peace in the world. Some other Net poster suggested that the US take some fraction of the $BIGNUM dollars recently approved to "fight terrorism", and use it to sponsor humanitarian and educational initiatives in Afghanistan, thus undercutting some of the "Great Satan" imagery that fuels the passion of the terrorist sympathizers. IMHO, this is a very interesting idea and is probably far more effective on a per-dollar basis than cruise missiles.
p.s. Just for the record, I am a Canadian citizen who luckily didn't lose any friends in the WTC attack (as far as I know). Apply appropriate weighting factors to my opinions.
your PD would probably have to be something like: seabrookpd.city.state.us which no one can remember
Well, I would hope that anybody in that city would be able to remember at least the ".city.state.us" part of the name. And, in that relatively small namespace, there should be no problem assigning the name "police" to the local police department.
I think the DNS system needs to be re-worked to use deeper sub-domain paths. It just doesn't work to have everybody in the world fighting over a small number of flat namespaces (especially when the holder of a trademark gets to claim that substring in every TLD, past present or future). Adding more TLDs is just making the problem worse.
Here's one possible alternative:
- Each TLD would be a category, like ".com" for businesses, ".org" for non-profit organizations, ".ind" for individual people's homepages, etc. Categories would overlap as little as possible, so that any registrant would clearly "belong" to one of them.
- Most names could not be registered directly within the TLD. Instead, they would be registered at a geographic sub-level corresponding to the scope at which the registering entity existed.
So, a local business "ABC Carpet Cleaning" would be able to register the name "abc-carpet-cleaning.vancouver.bc.ca.com" while a fedarally-incorporated business could register "aircanada.ca.com". A different local business in Toronto could register "abc-carpet-cleaning.toronto.on.ca.com" without creating any conflicts. Businesses that had operations in several countries would be allowed to register in the TLD, like "coca-cola.com".
To save some typing, a user's browser could support an abbreviated notation like "circuitcity_com". The DNS system would first try to match "circuitcity.vancouver.bc.ca.com" (or whatever the user's local context was), and would then look for matches all the way up the tree: "circuitcity.bc.ca.com", "circuitcity.ca.com", "circuitcity.com". If there were multiple matches, a page could pop up asking the user which one he wanted. Frequently visited sites would be in the user's bookmark file anyway, so the length of the name wouldn't be an issue.
Anyway, it's an thought. Maybe somebody already wrote up an RFC on this, or wrote a long essay explaining why it's a stupid idea. I haven't looked.
Envelopes are not the real-world equivalent of strong cryptography. Envelopes deter casual snooping, but it's very easy for a government agency to open one and examine the contents (probably without leaving any indication that it had been opened).
When dealing with strong cryptography, the equivalent question would have to be "would it make sense to ban welded-shut, lead-lined steel boxes in the postal system"? And to an average person on the street, the answer would probably be 'yes'.
How many other things started out as an April Fool's day joke and then actually got implemented?
Well, the classic example has to be RFC1149, A Standard for the Transmission of IP Datagrams on Avian Carriers
Re:Yet another failed .COM.
on
Rebel.com Autopsy
·
· Score: 1, Insightful
I feel no sympathy for these people.
I feel sorry for the Netwinder engineers. They had put together quite a nice little product, and there probably was a real market opportunity for it at the time (if they could have brought the unit cost down through increased production volume). Like the Amiga, however, there was just too much cluelessness at the top for the company to have any chance of succeeding.
And yes, at home I have both an Amiga (boxed up) and a Netwinder (running as a DNS/FTP/NTP server).
I agree, every Canadian should take a few hours to read the documents on that site and send in some sort of comment. Even if your comments don't affect their ultimate decision, at least you will be on record as standing up for your beliefs.
Let the government know that we assert its primary responsibility is to represent the interests of Canadian citizens, not those of international corporations.
Let them know that networked personal computers are more than mere content-delivery devices. Protest the fact that (for example) the music industry is charging you $0.21 for every CD-R you buy, even if you're using it to back up personal data or record your own music. Tell the government that you will not tolerate any further power-grabs by the music and video industries. Assert your right to create and manipulate your own data on your own PC, without interference.
Let them know that due process must be applied whenever a corporation wants to have a user's website shut down for alleged copyright infringement. Demand penalties and the user's right to seek compensation whenever a corporation makes a false accusation.
Refute the notion that content publishers are entitled to make the same profits they did with their pre-Internet business models. Times change, and economic factors change (just as they did when technology such as the phonograph and printing press were invented, essentially creating the publishing industry).
Demand the freedom to analyze, reverse-engineer, and discuss the technological systems used to control content. Prohibit only the disclosure of the encryption keys used by a particular protection method (following the general cryptographic principle that all the security is in the key, not the algorithm). Don't let corporations use your tax dollars to subsidize poor engineering (by arresting anyone who dares to point out the weakness in their ultra-secure "XOR-with-constant" protection technology).
Warn the government that extending the term of copyright (lifetime of author + 70 years, up from + 50 years), coupled with technology that replaces "sale" with "pay-per-view", will reduce the incentive to create new content (a goal of the copyright act) because it will be cheaper to just re-distribute the same old content over and over.
Say something!
p.s. Tell your friends, too. Many of my co-workers (at a dot.com) had never heard of Dmitry or the DMCA. Just because we see these stories every day, it doesn't mean the rest of the world does.
Only if you know your exact distance from the earth's center (as well as the earth's mass). How do you determine that? To measure G you need to use a torsion balance, and that requires a definition of kg (since G is defined in terms of N*m^2/kg^2).
From the description of the device, it relates the mass of the standard kilogram to an electromagnetic force, by balancing that electromagnetic force against the gravitational force on the mass (== the weight of the mass).
Therefore, the only relevant parameter is the one relating the mass of an object to the gravitational force on that object. This is "little g", the local gravitational acceleration, and it can be measured directly with only a length and a time standard. You don't need to know "big G" or anything about the earth.
The gravitational force on an object is given by:
Force(N) = mass(kg) * g(N/kg)
According to NIST [nist.gov], they've got a variance on the order of 3% per century in the observed mass (probably measured by weight) of the
standard kilo brick.
3 percent??? Do you have any idea how HUGE a variation that would be in a primary standard? Maybe if they polished it with a belt sander before every measurement...
The link you provided says:
[...]are causing the mass of the kilogram to vary by about 3 parts in 108 per century relative to sister prototypes.
Now I'm not certain about this, but I'd wager that the "108" is actually a "10 to the 8th power" that got mangled somewhere in the conversion to HTML. If so, it would represent a more plausible 0.000003% per century variation.
Call me a computer scientist, but isn't there something recursive about defining the prototype kilogram with gravity and then measuring gravity at the same time?
There are ways of measuring gravity without a known mass standard. For example, you could measure the acceleration of an object falling freely in a vacuum chamber (only need length and time measurements).
However, it does seem strange to me that they'd base the kilogram standard on something as indirect as the local gravitational field.
I've never used them, but Intel does provide high-performance math libraries. So, their compilers probably have real technical optimizations as well (not just marketing fluff).
It's the speed of light in a vacuum that's constant. In a material like water or glass, the speed of light is slower by a factor known as the refractive index. And yes, the refractive index of a material can be a function of the light's frequency (color).
About ten years ago there was a postscript virus that Did Things to printers
There's some info about it here. Was apparantly quite nasty on some hardware, as it changed a password that required an EPROM replacement to correct. This might have been more a "trojan" than a "virus", as I didn't find any references to it spreading itself (just that it could be a payload in clipart or other EPS files).
I thought that there was also something a few years ago where viewing a postscript file could alter files on your local machine (buffer overflow in a particular viewer program, unsafe default security settings, or something). However I couldn't find any information, so I might be mis-remembering.
I have been needing the help and advice on some things, but files send I to people, no response!
The thing is, if you want free help and advice from people, you need to show them some courtesy in return. Try putting a nice friendly "ILOVEYOU" in the subject line next time, and I'm sure you'll get a much better response. If that doesn't work, send them a picture of Anna Kournikova too.
What is a DC-DC converter? Why would one need to convert DC power to DC power? Stepping it to a different voltage, or something?
Yes, DC-DC converters are used to convert DC power to a different voltage (or polarity). One place you see them is on computer motherboards, where you have (for example) 3.3V available but you need 1.8V to power your CPU. A DC-DC converter can step the voltage down without wasting much of the power (as you would if you used a linear voltage regulator, or a resistor). DC-DC converters can also step voltages up, for example when you need to generate 300V for a camera flash from a 1.5V battery.
What I'd really would like is to have their special commbox for my otherwise ordinary PC. Something along the lines of one box by the computer where all the video, keyboard, mouse, sound and the like is plugged in, a cable, and another box at my desk where all connectors are available locally.
Check out the Cybex "Longview" - it carries KVM, audio, and a serial port up to 500 feet on standard Cat5/RJ45 cable (but it's analog, not ethernet). Combine it with some of their rack-mounted KVM switches, and you can run a whole server room remotely.
Slashdot Mirror
I agree these are nice speakers - I have a Monsoon MM-700 system on my desk.
However these are not "the speakers you'll find in Humvees". They're sold under the same Monsoon brand, but the flat-panel computer speakers are designed and manufactured by a company called Sonigistix in Richmond, BC, Canada. Last I checked, none of the Sonigistix flat-panel technology had yet made it into an automotive application.
The Monsoon speakers are based on planar-magnetic technology, partially licensed from a company called Eminent Technology. Sonigistix took the basic design, and adapted it so that the speakers could be mass-produced and sold at a lower price point. Note that there are several other flat-panel computer speakers on the market that do not use planar-magnetic technology, but instead use a "distributed-mode" technology developed by a company called "NXT". In my (biased, as a former Sonigistix employee) opinion, the planar-magnetic ones sound a lot better.
Similar systems are available commercially. Here is one that:
[...]will produce a maximum power of 12 watts. This system contains a unique DC/DC boost voltage converter and charge control adjustable from 5 to 14.8 V.and can be used to charge a battery and provide over 200 watt-hr per day of energy for various uses. It can be installed in 6" and up, firetube type heating systems with a minimum 125,000 btu. rating and is completely hidden with-in existing equipment.
thermoelectric batteries are totally new...
Link to the Citizen Eco-Drive Thermo watch...
"Eco-Drive Thermo converts the temperature difference between the user's body and the surrounding air into electrical energy to power the movement. [...] The original Eco-Drive Thermo was launched to great acclaim at the 1999 Basel Show."
Don't know if it's shipping to consumers yet, but the technology's been around for a while.
Several years ago I worked in a university lab that had a prototype of a 1 kW sulfur microwave lamp. It was very bright, but the light had a distinctive green tint. After a while your eyes would adapt, but then when you looked out the window (or at anything lit with a "conventional" light source) everything looked pink. Maybe the newer models have solved this problem, but if not I wouldn't want it as indoor lighting.
As for getting your skin colouring, one of the advantages of the sulfur lamp was that it put out much less UV radiation than other light sources. Good for museums where the UV would damage old paintings and documents; bad for getting a tan in your cubicle.
A straight port of CygnusED would be difficult, as it was a commercial product and the source code is probably long-buried by now. Also, it probably used a lot of Amiga-specific features that wouldn't map well to Linux.
However, I've found NEdit to be an acceptable replacement. It has many of the same features as CygnusED, such as the ability to cut-and-paste a column of text. It shouldn't be that hard to patch in the rest of CED's essential features.
(and just so I'm not completely off-topic, let me add that vi[m] is still my editor of choice for non-GUI situations, such as configuring a 'headless' server in a remote location over a 9600-baud serial line).
So did Vigor, the vi paperclip, make it into the 6.0 release?
The largest recorded since we've started measuring these things was an X20 in 1989. Quebec's power system overloaded that year.
According to this table, the 1989 flare is tied for #1 with one from April of this year (at least, in terms of X-ray intensity). However I don't remember hearing about any significant power or communication disruptions from the April flare.
If you go here, you can get source code for a simple multicast client and server program. Run a sender on one PC, a receiver on another, then add as many intermediate routers and boxes as you want.
I am presently working for a company that sells real-time stock market information, and the incoming data feeds from NASDAQ and SIAC (=NYSE) are in multicast format delivered over a private network. So I have a bit of experience with it, although our application doesn't need anything fancy like dynamic routing or supporting a large number of listening clients. I'm using SuSE Linux for some of the application servers, but I've found that OpenBSD is a better solution for routing the data, tunelling it over a secure VPN, etc.
p.s. If you want to know if your ISP supports multicast, just run tcpdump on your gateway and look for multicast or IGMP traffic.
I understand your point. However, one difficulty with modern cryptography is that it's technically very hard to give up "just enough" of your freedom.
You can't just weaken the algorithms (e.g. by only allowing 40-bit keys). Sure it makes it easier for the "good guys" to intercept the communications of the "bad guys", but the converse is also true. Even the terrorists will be able to listen in on your business communications, find out what flight your company's president is taking to the trade show, etc.
The only option that makes sense is to require government access to keys. In this scenario, the algorithms operate at full strength and users are secure against other users. However, if the government agencies want to investigate you, they will have a copy of your key on file.
However, there are problems. Under this scenario, the government has to have your key on file in advance of when they want to use it. It's no good for them to ask you for your key after they begin to suspect you; you'd know they were on to you.
Now this means that you have to send to the government a copy of every cryptographic key you ever generate. As it happens, this week I've been working on setting up some software-based VPNs to tunnel corporate data across the Internet. I've got a half-dozen OpenBSD boxes running OpenSSH, IPSec, and so on. There are randomly-generated (temporary) keys all over the place. These are just development boxes and will probably get wiped clean in a couple of weeks when I install the real servers. However, right now, each of these boxes is capable of creating a secure communications channel across the Internet. Therefore, under certain legal conditions, I would be required to register all of the keys with a central government agency.
When you're dealing with government access to keys, it's not good enough to have 99% compliance with the regulations. Your targets, the terrorists, are going to be in that last 1%. So you need to ensure 100% compliance with the regulations.
One way to do this is to allow only pre-approved, secure crypto implementations that have an automatic and tamper-proof key-forwarding mechanism. Guess what, you've just made it illegal for users to use any open-source crypto software. You've also made it illegal to run any open-source operating systems or hardware emulators, because these give you enough control that you could modify the operation of any software-based cryptographic utility (think of running a program in a VMWare sandbox, or single-stepping through it in an emulator, and modifying the key data that it sends to the government agency). Suddenly, the only legal "PC" would be a sealed box with a factory-installed Microsoft operating system, on a "CPRM"-style hard drive that couldn't be re-formatted. General-purpose compilers and scripting languages would have to be outlawed, so that ultimately the "PC" could only be used for online shopping, games, or listening to "pay-per-view" downloaded musical content (from RIAA-approved central servers, of course).
If you allow people to keep on using general-purpose computers, the only way to ensure that the government has access to your keys is to randomly audit the process, to try decrypting the messages of a user, and ensuring that you get a legible plain-text. If you don't, you have to haul that user in for a severe punishment (severe enough that it scares everyone else into making sure the government always has an up-to-date version of their keys).
So now you have a scenario where otherwise-innocent citizens have to be audited, and severely punished for the "crime" of not proactively ensuring that the central government agency has a copy of every random key they might ever happen to generate in the course of their daily lives. And you have to audit almost every citizen if you want to have a good chance of catching the terrorists (who make up an extremely small fraction of the population).
And if that's not bad enough, remember that most governments are not 100% free of corruption. Maybe some bored clerk in the crypto bureau will decide he wants to read the full plaintext of the day's decryption-compliance audits, rather than just feeding them to the AI system (maybe Carnivore, maybe just a simple algorithm to see if the decryption result looks like a standard document format). Or maybe the mafia pays off a clerk to deliver them a copy of your financial records, so they can blackmail you. Or maybe a large international aerospace corporation convinces the governenment agencies to let them peek at your bid on a large contract, so they can undercut your company and get the business.
It's a very slippery slope, when you start trying to introduce a "controlled leak" into a system that is inherently strong and secure. That's why I advocate a strong right to privacy, even though I feel the government may be morally entitled to read my "private" communications as long as due process is followed (present their case before a judge, get a 'search warrant' if they can demonstrate probable cause, etc).
I'd be interested to hear if anyone knows of technical solutions that would allow a "reasonable compromise" between freedom and security. Otherwise, I think we just have to take it as a modern "law of nature" that people can communicate privately without the possibility of interception. We as a society have to grow up. We're not children any more, and our "parents" (governemnts) can't protect us from all the evils of the world any more. We've all gotten too good at killing and hurting each other for the "parental" safeguards to work any more. All we can do is recognize our mutually vulnerable nature, and try to work together to build peace in the world. Some other Net poster suggested that the US take some fraction of the $BIGNUM dollars recently approved to "fight terrorism", and use it to sponsor humanitarian and educational initiatives in Afghanistan, thus undercutting some of the "Great Satan" imagery that fuels the passion of the terrorist sympathizers. IMHO, this is a very interesting idea and is probably far more effective on a per-dollar basis than cruise missiles.
p.s. Just for the record, I am a Canadian citizen who luckily didn't lose any friends in the WTC attack (as far as I know). Apply appropriate weighting factors to my opinions.
your PD would probably have to be something like: seabrookpd.city.state.us which no one can remember
Well, I would hope that anybody in that city would be able to remember at least the ".city.state.us" part of the name. And, in that relatively small namespace, there should be no problem assigning the name "police" to the local police department.
I think the DNS system needs to be re-worked to use deeper sub-domain paths. It just doesn't work to have everybody in the world fighting over a small number of flat namespaces (especially when the holder of a trademark gets to claim that substring in every TLD, past present or future). Adding more TLDs is just making the problem worse.
Here's one possible alternative:
- Each TLD would be a category, like ".com" for businesses, ".org" for non-profit organizations, ".ind" for individual people's homepages, etc. Categories would overlap as little as possible, so that any registrant would clearly "belong" to one of them.
- Most names could not be registered directly within the TLD. Instead, they would be registered at a geographic sub-level corresponding to the scope at which the registering entity existed.
So, a local business "ABC Carpet Cleaning" would be able to register the name "abc-carpet-cleaning.vancouver.bc.ca.com" while a fedarally-incorporated business could register "aircanada.ca.com". A different local business in Toronto could register "abc-carpet-cleaning.toronto.on.ca.com" without creating any conflicts. Businesses that had operations in several countries would be allowed to register in the TLD, like "coca-cola.com".
To save some typing, a user's browser could support an abbreviated notation like "circuitcity_com". The DNS system would first try to match "circuitcity.vancouver.bc.ca.com" (or whatever the user's local context was), and would then look for matches all the way up the tree: "circuitcity.bc.ca.com", "circuitcity.ca.com", "circuitcity.com". If there were multiple matches, a page could pop up asking the user which one he wanted. Frequently visited sites would be in the user's bookmark file anyway, so the length of the name wouldn't be an issue.
Anyway, it's an thought. Maybe somebody already wrote up an RFC on this, or wrote a long essay explaining why it's a stupid idea. I haven't looked.
Envelopes are not the real-world equivalent of strong cryptography. Envelopes deter casual snooping, but it's very easy for a government agency to open one and examine the contents (probably without leaving any indication that it had been opened).
When dealing with strong cryptography, the equivalent question would have to be "would it make sense to ban welded-shut, lead-lined steel boxes in the postal system"? And to an average person on the street, the answer would probably be 'yes'.
How many other things started out as an April Fool's day joke and then actually got implemented?
Well, the classic example has to be RFC1149, A Standard for the Transmission of IP Datagrams on Avian Carriers
I feel no sympathy for these people.
I feel sorry for the Netwinder engineers. They had put together quite a nice little product, and there probably was a real market opportunity for it at the time (if they could have brought the unit cost down through increased production volume). Like the Amiga, however, there was just too much cluelessness at the top for the company to have any chance of succeeding.
And yes, at home I have both an Amiga (boxed up) and a Netwinder (running as a DNS/FTP/NTP server).
I agree, every Canadian should take a few hours to read the documents on that site and send in some sort of comment. Even if your comments don't affect their ultimate decision, at least you will be on record as standing up for your beliefs.
Let the government know that we assert its primary responsibility is to represent the interests of Canadian citizens, not those of international corporations.
Let them know that networked personal computers are more than mere content-delivery devices. Protest the fact that (for example) the music industry is charging you $0.21 for every CD-R you buy, even if you're using it to back up personal data or record your own music. Tell the government that you will not tolerate any further power-grabs by the music and video industries. Assert your right to create and manipulate your own data on your own PC, without interference.
Let them know that due process must be applied whenever a corporation wants to have a user's website shut down for alleged copyright infringement. Demand penalties and the user's right to seek compensation whenever a corporation makes a false accusation.
Refute the notion that content publishers are entitled to make the same profits they did with their pre-Internet business models. Times change, and economic factors change (just as they did when technology such as the phonograph and printing press were invented, essentially creating the publishing industry).
Demand the freedom to analyze, reverse-engineer, and discuss the technological systems used to control content. Prohibit only the disclosure of the encryption keys used by a particular protection method (following the general cryptographic principle that all the security is in the key, not the algorithm). Don't let corporations use your tax dollars to subsidize poor engineering (by arresting anyone who dares to point out the weakness in their ultra-secure "XOR-with-constant" protection technology).
Warn the government that extending the term of copyright (lifetime of author + 70 years, up from + 50 years), coupled with technology that replaces "sale" with "pay-per-view", will reduce the incentive to create new content (a goal of the copyright act) because it will be cheaper to just re-distribute the same old content over and over.
Say something!
p.s. Tell your friends, too. Many of my co-workers (at a dot.com) had never heard of Dmitry or the DMCA. Just because we see these stories every day, it doesn't mean the rest of the world does.
Only if you know your exact distance from the earth's center (as well as the earth's mass). How do you determine that? To measure G you need to use a torsion balance, and that requires a definition of kg (since G is defined in terms of N*m^2/kg^2).
From the description of the device, it relates the mass of the standard kilogram to an electromagnetic force, by balancing that electromagnetic force against the gravitational force on the mass (== the weight of the mass).
Therefore, the only relevant parameter is the one relating the mass of an object to the gravitational force on that object. This is "little g", the local gravitational acceleration, and it can be measured directly with only a length and a time standard. You don't need to know "big G" or anything about the earth.
The gravitational force on an object is given by:
Force(N) = mass(kg) * g(N/kg)
(note that 1 N/kg == 1 m/(s^2))
According to NIST [nist.gov], they've got a variance on the order of 3% per century in the observed mass (probably measured by weight) of the
standard kilo brick.
3 percent??? Do you have any idea how HUGE a variation that would be in a primary standard? Maybe if they polished it with a belt sander before every measurement...
The link you provided says:
[...]are causing the mass of the kilogram to vary by about 3 parts in 108 per century relative to sister prototypes.
Now I'm not certain about this, but I'd wager that the "108" is actually a "10 to the 8th power" that got mangled somewhere in the conversion to HTML. If so, it would represent a more plausible 0.000003% per century variation.
Call me a computer scientist, but isn't there something recursive about defining the prototype kilogram with gravity and then measuring gravity at the same time?
There are ways of measuring gravity without a known mass standard. For example, you could measure the acceleration of an object falling freely in a vacuum chamber (only need length and time measurements).
However, it does seem strange to me that they'd base the kilogram standard on something as indirect as the local gravitational field.
I've never used them, but Intel does provide high-performance math libraries. So, their compilers probably have real technical optimizations as well (not just marketing fluff).
It's the speed of light in a vacuum that's constant. In a material like water or glass, the speed of light is slower by a factor known as the refractive index. And yes, the refractive index of a material can be a function of the light's frequency (color).
I thought there were some weather satellites
The site http://www.ssec.wisc.edu/data/ has some nice pictures, including full-hemisphere views from GOES-8 and GOES-10.
Of course these are in a fixed position with respect to the earth's surface, while GoreSat would have been fixed with respect to the sun's position.
Also of interest is the SOHO spacecraft currently orbiting L1 and observing the sun.
There's already at least one spacecraft stationed at L1 - SOHO (Solar and Heliospheric Observatory).
There's a page here showing its orbit. It doesn't sit right at the L1 point, but rather moves in a "halo orbit" around L1.
About ten years ago there was a postscript virus that Did Things to printers
o stv.txt
There's some info about it here. Was apparantly quite nasty on some hardware, as it changed a password that required an EPROM replacement to correct. This might have been more a "trojan" than a "virus", as I didn't find any references to it spreading itself (just that it could be a payload in clipart or other EPS files).
http://catless.ncl.ac.uk/Risks/10.32.html#subj1
ftp://ftp.minolta-qms.com/pub/cts/out_going/dos/p
http://www.sevenlocks.com/password/pspass.txt
I thought that there was also something a few years ago where viewing a postscript file could alter files on your local machine (buffer overflow in a particular viewer program, unsafe default security settings, or something). However I couldn't find any information, so I might be mis-remembering.
I have been needing the help and advice on some things, but files send I to people, no response!
The thing is, if you want free help and advice from people, you need to show them some courtesy in return. Try putting a nice friendly "ILOVEYOU" in the subject line next time, and I'm sure you'll get a much better response. If that doesn't work, send them a picture of Anna Kournikova too.
Stupid question...
What is a DC-DC converter? Why would one need to convert DC power to DC power? Stepping it to a different voltage, or something?
Yes, DC-DC converters are used to convert DC power to a different voltage (or polarity). One place you see them is on computer motherboards, where you have (for example) 3.3V available but you need 1.8V to power your CPU. A DC-DC converter can step the voltage down without wasting much of the power (as you would if you used a linear voltage regulator, or a resistor). DC-DC converters can also step voltages up, for example when you need to generate 300V for a camera flash from a 1.5V battery.
What I'd really would like is to have their special commbox for my otherwise ordinary PC. Something along the lines of one box by the computer where all the video, keyboard, mouse, sound and the like is plugged in, a cable, and another box at my desk where all connectors are available locally.
Check out the Cybex "Longview" - it carries KVM, audio, and a serial port up to 500 feet on standard Cat5/RJ45 cable (but it's analog, not ethernet). Combine it with some of their rack-mounted KVM switches, and you can run a whole server room remotely.