I'd like to propose an enhancement to the SMTP protocol: The MTA which receives the mail on behalf of the user should answer to the question "SPAM_OK" with either "Yes, TTL=x" or "No, TTL=x". Not following the answer should be made illegal (high fines or "downtime" attached). Every sender should be required to explain why he thinks his mail is not SPAM - failure to do so or unability to prove an existing business relationship, see above...
This is neither an opt-in nor an opt-out situation. Instead, people get to choose wether they want opt-in (Answer: No SPAM) or opt-out (Answer: Yes, SPAM is ok).
Redirecting and rerouting can not be used to transparently and permanently move a server or an entire subnet to a different IP-space. While transparency could temporarily be achieved with redirection, redirecting permanently would defeat the purpose. An abstract layer of names is necessary for the internet to work. My proposal is this: Create a top-level domain with a meaningless name, like ".TPMNKT". Then, just like the current domain name system, delegate subdomains, but require that these subdomains are handed must be random strings of consonants. Also require that these strings are at least 20 chars long and contain at least 10 different consonants. This way, all vanity should be removed from the game and domain names become just an abstract addressing layer. With that layer of fallback names in place, additional naming schemes can be proposed and tested by anyone without harming the basic operability of the net.
The WWW needs the domain name system. Virtual hosting relies on the domainname to be transmitted in the HTTP request. Otherwise the server wouldn't know which of the many websites hosted on a single IP is the one you want to see.
Other important aspects of the web would also fall victim to the no-DNS approach, for example multihoming and transparent migration of servers (which would be deadly for the email-system).
A ringtone is just a specially formatted short message. These short messages would use much less than a second of airtime, IF they were using communication slots at all. The same applies to music: How often have you heard that people don't understand CD prices? A compact disc can be made for under 2$, including booklets. How comes they are sold for so much more?
So, that's not the product, you say. The information is the product. But that's not how people see it. That ringtone is just a bunch of keypresses if you do it yourself, and it needs to be programmed just once, even if you transmit it to thousands of cellphones. Can't be the effort of programming or making it available which justifies the prices. Intellectual property implies stuff like "someone thinks of something in a lucky moment and gets paid for it for the rest of his life while I have to go to work over and over again to keep getting paid", so that's easily put aside. I'd say it's simply the feeling of getting ripped off that keeps people away from paying for content.
Almost all code reuse is done in the scope of a single project or a single company. No this isn't ideal, but it is reality.
While that is naturally true for application specific code, there are many examples of successful and widely used interfaces: Microsoft's COM, KDE's KParts, CORBA, XML, etc. Code reuse is not limited to what is compiled into the application. Any form of code reuse on a noteworthy scale requires some sort of interface definition. Understanding and developing interfaces is therefore one of the important goals of programming. The Linux kernel with its modular architecture is actually a good example: Drivers implement a certain interface, the "real" kernel does not need to know about the peculiarities of each piece of hardware (or data structures such as filesystems).
Lego's interface is easily understandable, yet allows for a great variety of models. Lego Technics and Lego Mindstorms build on the same "interface". There are drawbacks to the Lego interface: Models need more room for complex functionality and are not as rugged as Meccano models. Meccano's interface on the other hand has drawbacks as well: It requires you to bend pieces, which "destroys" them - reuse is limited. The pieces are made of metal and you can easily cut yourself if you're not careful. Personally I tend to think that Lego is for the younger child or someone who is more into the software side of a model, while Meccano is better suited for older children and those who want to create more lifelike models. Both are lightyears ahead of a playstation in terms of stimulating creativity.
Interfaces are the more important part of programming. Given a reasonable specification, any "idiot" can write a specialized piece of code. The true challenge is writing universal code. To make your code universal, you have to create universal interfaces and you have to recognize the abstract nature of the problem you're trying to solve, not just the special case you're looking at.
If people where still writing PUTs and POPs instead of programming on higher abstraction levels, nothing would work, neither in time nor at all.
The problem with selling Legos is that they are made of plastic (which makes them last about forever) and the universal parts are of course the most interesting. The whole "build your own action figure from specially designed Legos" is not going to help, since the fascination about these models wears off far too quickly. There are almost no reasons to buy new stuff from Lego, except for truly new universal parts like Lego Mindstorms. Everything else is just handed over from generation to generation. Lego could be selling more if they made the bricks rot away after some time...;)
Suppose you put stuff on your website which, by chinese law, must not be made available to chinese people. Suppose you don't restrict access to the site. That means you are breaking the law in china, and should someone sue you in a chinese court, the verdict then has to be enforced by your country, unless "enforcement would be manifestly incompatible with the public policy" of your country. Whatever that means.
Here's some more information on "The Big Dig". Scroll down to "Also known as" and don't forget to read some user reviews: http://us.imdb.com/Title?0065076
What if there are problems on the network (like inactivity kicks from your provider) or the local system has the "problem" of not being powered-on all the time? What if you want to switch desktops without having to restart applications on the server just for redirecting them to a different X-server?
You wouldn't end up with a coke in your hand if you had a 100$ bill either. There are however several problems with a truly gold-backed currency:
There is a fixed amount of gold. Productivity however increases, resulting in increasing "value" of gold in the long run, in other words deflation. Deflation is counterproductive because it stiffles investments. Keep your gold, get more for it tomorrow.
The amount of accessible gold is not fixed and not coupled to relevant economic processes or goals. "Printing money" is up to mining corporations. Advancements in mining technology cause your gold to drop in value, a flooded mine make it more valuable. Not exactly predictable or desirable.
But at least with gold there is a "hard" reference of value. So we won't tumble into a situation where suddenly all our money is worthless. Bzzzt. Wrong. If I get the impression that I can't buy what I want or need, because no one has the skill or will to produce it, I won't give my products to you, no matter if you pay in gold or not. In return, you refuse to sell your products for gold. So do your former customers. Suddenly gold will buy you nothing. It has become worthless. The "intrinsic value" of gold is what it can be used for in production, no more no less.
Gold is untraceable. This creates all sorts of problems, for example it makes fighting organised crime much harder.
The most important part is of course the fact that while money is seen as a placeholder for gold in a gold-backed currency system, gold itself is just a placeholder for trust. If people stop trusting an economic system, basic needs dictate what is most valuable. Most people do not count possession of gold as one of their basic needs.
VNC lets you keep the windows open even if you close your local X-server. If you have a system that you just want to look after from time to time, you don't need to reopen applications everytime. Just leave them running on the virtual VNC X-server. An added bonus is that you can then download a small and free application for any OS to see the screen of that headless box. If you have the cycles to spare, that is a lot more convenient than restoring the session everytime you want to use that machine.
This ain't for Californians. Plasma displays' power consumption...
Re:How long until this article will not be availab
on
The Future Of The Book
·
· Score: 2
I didn't mean to say that asking the author is the best option when an online document has moved. First you should of course try to find its new location. In the scientific field, which you mentioned, there is a good chance that the document has just moved instead of having been removed. References are not locators. They are sets of information which describe the cited text in enough detail, enabling the reader to find it. Maybe people need to be educated that removing something from the web is the equivalent of recalling every unsold printed copy and almost all library copies of a book. But it is important to see that this problem is not inherent to the medium. The problem is more likely to be noticed because most of the information on the web isn't indexed and archived externally. Part of that problem is the extremely fast growing pile of information. Another part is the lack of an "official" effort to provide such an index/archive.
Reading "Art of War" when waiting in line at the supermarket is not for the aggresive type of person. Kids, do not try this at the supermarket. He's a trained professional.
Re:How long until this article will not be availab
on
The Future Of The Book
·
· Score: 2
What good is a reference if you can't get the referenced text? There is not much difference between an out-of-print paper book and a moved website. If you really want the text, you can always ask the one who wrote it (or the publisher). You do of course need to give proper references, and that means to name much more than a URI.
The near-term future of the book doesn't look to bad, but mid- to long-term, paper books will be an absolute niche product. What are the key arguments against electronic books?
Can't be read whereever you want, in bright daylight for example
Are heavier and less robust
They need a lot of battery power to work
Stupid licensing schemes scare readers away
There are however important advantages to electronic books:
More compact if you need access to many conventional books' contents
Can be updated, which is an excellent feature for technical manuals or other rapidly changing content
Electronic books can contain active content. A dictionary could help reading foreign texts, for example.
I'm not counting environmental aspects on either side, because it isn't obvious which side has the better balance in the long run.
New display technologies like high resolution electronic paper will remove the bad readability, the high weight and energy consumption and probably the not-too-stellar robustness from the cons list. That leaves us with stupid licensing schemes to battle. I get the impression that more and more developers are jumping on the open hardware development train again, after that has been totally "out" for some time now. Microcontroller based this and thats spring up everywhere. And they are not left out in the cold by the software guys: Linux is running on anything from watches to gaming consoles to homebuilt mp3 players based on you-haven't-heard-of-that-one-before processors. I don't think it's too optimistic to say that should corporate marketing insist on pay-per-read, the "underground" will have an alternative ready, either in the form of selfmade machines or in the form of "debugged" firmwares for commercially available "books".
Nobody's installing "special flourescent lamp networks". Only the electronic ballasts in existing lamps need to be replaced by just marginally more expensive versions which can be programmed to emit a short programmable data sequence. This way each lamp tags a certain area. The receiver uses that information to show your position on a map or whatever.
You just program the ballast and replace the old one with it. From then on this light continuously broadcasts its short data sequence, over and over again. No network attached, no power supply, no batteries. Just a slightly more expensive electronic ballast.
The idea is to "tag" areas and have receivers which can use the location information to give advice to the handicapped (like "you're supposed to be in the kitchen and you're not"). The technology could be used to distribute more data, but for this application it's just used as "low-res indoor GPS".
Slashdot Mirror
I'd like to propose an enhancement to the SMTP protocol: The MTA which receives the mail on behalf of the user should answer to the question "SPAM_OK" with either "Yes, TTL=x" or "No, TTL=x". Not following the answer should be made illegal (high fines or "downtime" attached). Every sender should be required to explain why he thinks his mail is not SPAM - failure to do so or unability to prove an existing business relationship, see above...
This is neither an opt-in nor an opt-out situation. Instead, people get to choose wether they want opt-in (Answer: No SPAM) or opt-out (Answer: Yes, SPAM is ok).
all objects closer to another but because of their speed, I doubt many will actually collide
I'll cite that when I get pulled over by the police for speeding again.
Why are they using the Startrek font on that "Quicktime" mpeg? Are they trying to get some funding for an escape plan?
Redirecting and rerouting can not be used to transparently and permanently move a server or an entire subnet to a different IP-space. While transparency could temporarily be achieved with redirection, redirecting permanently would defeat the purpose. An abstract layer of names is necessary for the internet to work. My proposal is this: Create a top-level domain with a meaningless name, like ".TPMNKT". Then, just like the current domain name system, delegate subdomains, but require that these subdomains are handed must be random strings of consonants. Also require that these strings are at least 20 chars long and contain at least 10 different consonants. This way, all vanity should be removed from the game and domain names become just an abstract addressing layer. With that layer of fallback names in place, additional naming schemes can be proposed and tested by anyone without harming the basic operability of the net.
The WWW needs the domain name system. Virtual hosting relies on the domainname to be transmitted in the HTTP request. Otherwise the server wouldn't know which of the many websites hosted on a single IP is the one you want to see. Other important aspects of the web would also fall victim to the no-DNS approach, for example multihoming and transparent migration of servers (which would be deadly for the email-system).
A ringtone is just a specially formatted short message. These short messages would use much less than a second of airtime, IF they were using communication slots at all. The same applies to music: How often have you heard that people don't understand CD prices? A compact disc can be made for under 2$, including booklets. How comes they are sold for so much more?
So, that's not the product, you say. The information is the product. But that's not how people see it. That ringtone is just a bunch of keypresses if you do it yourself, and it needs to be programmed just once, even if you transmit it to thousands of cellphones. Can't be the effort of programming or making it available which justifies the prices. Intellectual property implies stuff like "someone thinks of something in a lucky moment and gets paid for it for the rest of his life while I have to go to work over and over again to keep getting paid", so that's easily put aside. I'd say it's simply the feeling of getting ripped off that keeps people away from paying for content.
Almost all code reuse is done in the scope of a single project or a single company. No this isn't ideal, but it is reality.
While that is naturally true for application specific code, there are many examples of successful and widely used interfaces: Microsoft's COM, KDE's KParts, CORBA, XML, etc. Code reuse is not limited to what is compiled into the application. Any form of code reuse on a noteworthy scale requires some sort of interface definition. Understanding and developing interfaces is therefore one of the important goals of programming. The Linux kernel with its modular architecture is actually a good example: Drivers implement a certain interface, the "real" kernel does not need to know about the peculiarities of each piece of hardware (or data structures such as filesystems).
Lego's interface is easily understandable, yet allows for a great variety of models. Lego Technics and Lego Mindstorms build on the same "interface". There are drawbacks to the Lego interface: Models need more room for complex functionality and are not as rugged as Meccano models. Meccano's interface on the other hand has drawbacks as well: It requires you to bend pieces, which "destroys" them - reuse is limited. The pieces are made of metal and you can easily cut yourself if you're not careful. Personally I tend to think that Lego is for the younger child or someone who is more into the software side of a model, while Meccano is better suited for older children and those who want to create more lifelike models. Both are lightyears ahead of a playstation in terms of stimulating creativity.
Interfaces are the more important part of programming. Given a reasonable specification, any "idiot" can write a specialized piece of code. The true challenge is writing universal code. To make your code universal, you have to create universal interfaces and you have to recognize the abstract nature of the problem you're trying to solve, not just the special case you're looking at.
If people where still writing PUTs and POPs instead of programming on higher abstraction levels, nothing would work, neither in time nor at all.
The problem with selling Legos is that they are made of plastic (which makes them last about forever) and the universal parts are of course the most interesting. The whole "build your own action figure from specially designed Legos" is not going to help, since the fascination about these models wears off far too quickly. There are almost no reasons to buy new stuff from Lego, except for truly new universal parts like Lego Mindstorms. Everything else is just handed over from generation to generation. Lego could be selling more if they made the bricks rot away after some time... ;)
Grandpa, what is a floppy drive?
Suppose you put stuff on your website which, by chinese law, must not be made available to chinese people. Suppose you don't restrict access to the site. That means you are breaking the law in china, and should someone sue you in a chinese court, the verdict then has to be enforced by your country, unless "enforcement would be manifestly incompatible with the public policy" of your country. Whatever that means.
Here's some more information on "The Big Dig". Scroll down to "Also known as" and don't forget to read some user reviews:
http://us.imdb.com/Title?0065076
What if there are problems on the network (like inactivity kicks from your provider) or the local system has the "problem" of not being powered-on all the time? What if you want to switch desktops without having to restart applications on the server just for redirecting them to a different X-server?
- There is a fixed amount of gold. Productivity however increases, resulting in increasing "value" of gold in the long run, in other words deflation. Deflation is counterproductive because it stiffles investments. Keep your gold, get more for it tomorrow.
- The amount of accessible gold is not fixed and not coupled to relevant economic processes or goals. "Printing money" is up to mining corporations. Advancements in mining technology cause your gold to drop in value, a flooded mine make it more valuable. Not exactly predictable or desirable.
- But at least with gold there is a "hard" reference of value. So we won't tumble into a situation where suddenly all our money is worthless. Bzzzt. Wrong. If I get the impression that I can't buy what I want or need, because no one has the skill or will to produce it, I won't give my products to you, no matter if you pay in gold or not. In return, you refuse to sell your products for gold. So do your former customers. Suddenly gold will buy you nothing. It has become worthless. The "intrinsic value" of gold is what it can be used for in production, no more no less.
- Gold is untraceable. This creates all sorts of problems, for example it makes fighting organised crime much harder.
The most important part is of course the fact that while money is seen as a placeholder for gold in a gold-backed currency system, gold itself is just a placeholder for trust. If people stop trusting an economic system, basic needs dictate what is most valuable. Most people do not count possession of gold as one of their basic needs.It boils down to this: Do you trust a company enough to accept a new currency from them?
VNC lets you keep the windows open even if you close your local X-server. If you have a system that you just want to look after from time to time, you don't need to reopen applications everytime. Just leave them running on the virtual VNC X-server. An added bonus is that you can then download a small and free application for any OS to see the screen of that headless box. If you have the cycles to spare, that is a lot more convenient than restoring the session everytime you want to use that machine.
Can you say "Denial of Service"?
Everyone seems to be doing fine.
So it worked out as expected. The text was published in electronic form with exactly that kind of distribution in mind. Do follow that link.
This ain't for Californians. Plasma displays' power consumption...
I didn't mean to say that asking the author is the best option when an online document has moved. First you should of course try to find its new location. In the scientific field, which you mentioned, there is a good chance that the document has just moved instead of having been removed. References are not locators. They are sets of information which describe the cited text in enough detail, enabling the reader to find it. Maybe people need to be educated that removing something from the web is the equivalent of recalling every unsold printed copy and almost all library copies of a book. But it is important to see that this problem is not inherent to the medium. The problem is more likely to be noticed because most of the information on the web isn't indexed and archived externally. Part of that problem is the extremely fast growing pile of information. Another part is the lack of an "official" effort to provide such an index/archive.
Reading "Art of War" when waiting in line at the supermarket is not for the aggresive type of person. Kids, do not try this at the supermarket. He's a trained professional.
What good is a reference if you can't get the referenced text? There is not much difference between an out-of-print paper book and a moved website. If you really want the text, you can always ask the one who wrote it (or the publisher). You do of course need to give proper references, and that means to name much more than a URI.
The near-term future of the book doesn't look to bad, but mid- to long-term, paper books will be an absolute niche product. What are the key arguments against electronic books?
- Can't be read whereever you want, in bright daylight for example
- Are heavier and less robust
- They need a lot of battery power to work
- Stupid licensing schemes scare readers away
There are however important advantages to electronic books:- More compact if you need access to many conventional books' contents
- Can be updated, which is an excellent feature for technical manuals or other rapidly changing content
- Electronic books can contain active content. A dictionary could help reading foreign texts, for example.
I'm not counting environmental aspects on either side, because it isn't obvious which side has the better balance in the long run.New display technologies like high resolution electronic paper will remove the bad readability, the high weight and energy consumption and probably the not-too-stellar robustness from the cons list. That leaves us with stupid licensing schemes to battle. I get the impression that more and more developers are jumping on the open hardware development train again, after that has been totally "out" for some time now. Microcontroller based this and thats spring up everywhere. And they are not left out in the cold by the software guys: Linux is running on anything from watches to gaming consoles to homebuilt mp3 players based on you-haven't-heard-of-that-one-before processors. I don't think it's too optimistic to say that should corporate marketing insist on pay-per-read, the "underground" will have an alternative ready, either in the form of selfmade machines or in the form of "debugged" firmwares for commercially available "books".
Nobody's installing "special flourescent lamp networks". Only the electronic ballasts in existing lamps need to be replaced by just marginally more expensive versions which can be programmed to emit a short programmable data sequence. This way each lamp tags a certain area. The receiver uses that information to show your position on a map or whatever.
You just program the ballast and replace the old one with it. From then on this light continuously broadcasts its short data sequence, over and over again. No network attached, no power supply, no batteries. Just a slightly more expensive electronic ballast.
The idea is to "tag" areas and have receivers which can use the location information to give advice to the handicapped (like "you're supposed to be in the kitchen and you're not"). The technology could be used to distribute more data, but for this application it's just used as "low-res indoor GPS".