Well actually every good consumer router already does that. For example with mine, I can define 4 types of packages and impose certain bandwidth levels on them, as well as set priorities. It's a basic feature of many DSL-Routers.
40 years seriously is _fast_ in the computing business.
Just look at the GRAIL-system from the end of the 1960s. (It's somewhere in the 3rd 3rd I think, it has flowchart like graphics on the screen) http://www.archive.org/details/AlanKeyD1987 That old system is still more advanced than anything you can buy today.
40 years is essentially nothing in UI development. There are lots of usefull concepts to be explored. Like typed natural language interfaces.
Well how could one exploit such a system. It doesn't cost anything more to route than to just be a part of the network.
Well tracking is an integral issue with those cell-"phone" networks. Unfortunately the operators even want it to provide location-based services.
I guess that 5G networks will be meshed and user-driven. It's just an ingenious business model. The users provide an open network, and phone-operators can just use that network, simmilar to VoIP-Providers using the internet. It's just so much cheaper to give away a few thousand $20 WLAN-Routers as well as free low bandwidth (1MBit max) internet access as to build up a few base-stations.
A business-model could look like this: ISPs agree on technical standards for meshed wireless networking. Then they let companies build boxes for that kind of networking and sell them, or give them away for free. After a network has been buildt, those ISPs offer VPN-tunnels into the internet, as well as other services for a fee, or finance that with advertisements. It's essentially the web business-model brought to networking. You have a big network where everybody shares the cost, and you can use it any way you want.
User-owned wireless networks can provide 2 features, cell-phone networks cannot.
1. They are affordable.
2. They are free of corporate censorship.
Cell phone networks are still designed with low-bandwidth voice in mind. They do not really scale beyond a few megabits per cell, even with 3G networks which we will be stuck with for the next decades.
There also is a more dangerous consequence. Traditional cell phone networks need to have a preety good clue to where your mobile station is. That means people will be able to easily track you and follow every move, even without the need to get out of their office. In times of governemnts all over the world getting more and more fashist, this is definitely not good.
Actually in many cases all that's needed is a wireless network card and a good antenna. Every computer can easily act as a meshed networking router.
For example we currently set up such a system at our dormatory. We will set up one server for a wiki and some ftp-space, but otherwise the only infrastructure will be the computers of the users.
We do not have any internet access in that network, except for a proxy that relays into the tor network.
If there is any interrest you can read my article in the Winter 2007-2008 issue of 2600, page 9. If there is even more interrest, I can post my raw version here.
Just install an anonymous proxy so all the http-trafic goes via tor or something.
Besides, there hasn't been a single case yet in which something like that has happened.
Furthermore, who says you need to provide internet access via such a network? Just a general local network would be sufficient. One lawyers cannot look into easily without actually going to the neighbourhood.
I guess the main issue will be backward compatibility. This preety much seems to be the main complaint about Windows Vista. A "new" version will even be less compatible. Even if they have some VM-based compatibility system it's not likely to be based on XP or 2000.
The main selling point of Windows is not the OS itself, but the fact that you can run software on it written over the last decade. Even in the year 2000, many businesses were running DOS applications. In fact it is very rare that you find Windows applications actually using new features.
Most Windows software is legacy software. No developer would write for Windows 7 if that meant the software wouldn't run under Windows 2000 or Windows XP anymore. They might as well write it under GTK and just port it to whatever platform they want.
Some people talk about a recession, of course that is not quite true. What will come is a stagflation which is a recession combined with an inflation.
Of course engeneers might be laid off first because of the idiocy of the managers, but that will not make the situation any 'better'. With no developers or service-persons, and no money to afford foreign workforce they will not be able to sustain their business.
But unlike them you are able to go on living. You can use an old car and convert it into a power plant for generating electricity. You can use LEDs from old devices to give off a bit of light. You can turn junk into something usable. Those will be sought after then. You will be able to earn your food while ecconomists will have to starve, or try to continue to live their highly parasitic lifestyle.
Of course COBOL is still around. There currently is virtually no other language beeing able to replace it because they lack BCD arithmetics for easy to predict precision.
Also there is another one missing. "Developer for organically grown commercial computing enviroments". If you ever had to programm something serious under Windows, and then later tried another OS you will know what I mean. You have Interfaces which are borderline bizarre, partly undocumented, and only plausible if you take decades of company policy into account. It is even more frustrating when you find an error in the interfaces you are using. It is virtually impossible to reach anybody who even remotely cares for your problem, even if the code which shows the problem only has 5 lines and is an exact copy of one of the examples giving out by the people writing the interface.
There may be a need for a "new" net, but the goals would be completely different.
Such a network would need to provide things like distributed caching by default and censorship resistance, as well as anonymity.
For example the network would cache all cachable protocolls by default, as often as it can be done. Then no site could be slashdotted again as many of the routers in between would just cache the content. A great side effect is that the identity of the originator of the request would be obscured by the routers.
Another important point is that it must not have any "single points of control" like the DNS-system or IP allocations. Furthermore we would need to focus on every participant beeing able to route. The network must not be tree-like anymore. If you have wireless LAN and your neighbour has, too, there must be automatic peering.
Another idea would be to make it work on scaresly connected networks. Imagine you have a mobile device. It could try to fetch your encrypted (!) e-mail and fetch it whenever you have a connection. Every router in the connection would try to accept the request and cache the response until you have a connection again.
While in most commercial software packages, the developers and users are strictly seperated, in FOSS they are typically closely coupled.
So more often than not, commercial companies bring out software that nobody wants and wont work propperly. A good example is all the malware around. Software that spies on you, displays ads on your screen, sends out spam or makes your computer participate in a DDOS.
While with FOSS you actually get software which is typically as good as resources allow, and not just as good as the business model allows.
I guess most POS equipment runs on dedicated hardware.
There is no reason not to use Linux. Barcode scanners emulate keypresses. There are no usable software interfaces for POS devices like customer displays or bon printers. Those are written by directly accessing the serial and parallel ports. A task that is extremely frustrating under Windows.
Depending on what you want. You could work with simple shell scripts and perhaps some awk or perl. If you don't want to store information about the customer it can be extremely simple. Essentially a table in an SQL database which stores the article number, the price and the date/time. You can easily get many pieces of information from that.
This seriously works for commentaries. Al Jazeera International does it since it's start. You can just send your video commentary about any topic to them, and if it's good they will publish it.
And seriously, at the current state of CNNs news coverage, even getting reports about broken pickle jars in supermarkets would be an improvement. I mean they will surely censor out the good stuff anyhow. Stuff that would deserve to be in the media.
Seriously, do you really think the US television would even report on it? Wars aren't that popular anymore, so it's probably not a good idea to brag about it in the media.
The war will start soon, but probably the only tv station reporting about it will be Al Jazeera International.
Ohh yes, that guy who found out that the wiimote exist.
Well my personal favourite idea I want to do some research in is to build a huge "Rand-Tablet". Such a tablet is essentially a simplified Wacom tablet. You have a grid of thin wires behind your surface through which you run pulses or bursts of current. Your pen consists of a coil (perhaps with amplifier) which picks up the magnetic fields generated by the coils. Now as you know which wire is acticated when, you can simply measure the relative distances of your pen tip from the nearest wires and therefore calculate the precise position of your pen.
As those wires can be really thin, you can simply place them a few inches behind a translucent projection screen, making them essentially invisible.
Of course you can use multiple pens, and if you are good, you can even try to use the wires as antennas to read back the signal of the pen wirelessly. But for public terminals wired pens might be a better idea.
1. Computers won't get much smaller. Just think of it, you can now buy a cheap C64 clone which consists of a single chip, but yet they build a bulky case around it so it will be usefull. Just look at Laptops, they get bigger every year. Today you can buy laptops with 17 inch screens.
2. Computers won't have hundreds of Chips. The number of Chips will most likely decrease with nanotechnology. Even today we see a trent towards as little chips as possible. What the author might have meant is that we might have multi processor systems. Yes, that will probably happen.
Anyhow my guesses are as follows: There will be a new medium sized breakthrought like the Internet, and Microsoft will again, completely miss it.
Maybe there will be a network access revolution: The Internet has abolished the traditional publisher/consumer relationship as everyone could publish for about the same cost as reading. There are already the first data-networks which try to push that idea down to the network level. In those networks you don't have an ISP anymore, but you just peer with your neighbours (for example over WLAN). Just think about it, if you lay a cable to your neighbour, you would easily get a gigabit of bandwidth. And even with moderately cheap switches you can easily scale that up. There may still be ISPs, but their importance will fade. This is the vital key to having network access everywhere. Without that you just will be carrying around homing devices which you don't dare using because every use costs you a significant amount of money. User generated networks are just far more efficient.
Voice access will come, but mostly to cheap devices. Over the next years the cost of implementing voice recognition might be significantly lower than the one of implementing buttons and displays.
Software might _slowly_ approach the state of the art of the 1970s again. We might get systems the average person can actually use after a day of training or so. Maybe we need to get rid of "I want that feature"-languages like C, C++, C# or Java and get back to the more "scientific" languages like Smalltalk or Lisp. One yet has to show me a system that actually does anything which would be hard to implement in Assembler in one of those "new" languages.
But realistically, I don't see that much change in the future. The computer industry just got to commercial. Back till the mit 90s you were at least able to catch a glimpse of the next 10 years by looking at 'workstations', but today the workstation market has completely died.
ARCNet is still alive and well doing. It's used in areas where precise timing is important and you'd like to have LAN-like networks. For example quite a bit of TV studio equipment has ARCNet to communicate.
Well as long as the average user still doesn't do anything, or constantly clicks 'accept' everywhere there might not be much 'extremely malicious' software around. (Althought I have to say some ways of controlling a web-browser are simpler than opening TCP/IP connections.)
Temporarily disconnecting a computer won't do much good, as the software could just wait, or refuse to start without an internet connection.
Then there are the holes created by 'personal firewalls'. Not only have there been dozens of buffer overruns, but also logic errors. For example some 'personal firewalls' allow you to block transmitting personal information. They run a little proxy server which filters out queries containing certain names and numbers. Let's imagine you entered your 4-digit PIN number there. An attacker could just get you to a website (or several) with 10000 tiny little images numbered 0000.jpg to 9999.jpg. Your browser would eventually load all of them, except for the one containing your PIN-Number. With a bit more thought that can be changed to to give you hints what the personal informations are.
You do realice that a piece of software on your own computer cannot stop any of that. Simply put, how should that programm find out if you are surfing the web or if a programm is controlling your browser to phone home?
Personal firewalls suck, it has been proven that they usually can be penetrated from both sides. They also confuse the user and teach him to "click accept or nothing will work", which is barely something you want your user to do.
It's a much more sensible advise to tell your user to turn off unnessesary services, especially since there are now simple applications which do that work for you.
Essentially just because vendor A has devices which consume "up to 18 watts" it doesn't mean that nobody can build devices which take less than 13 Watts.
That's just 5 watts difference. You could probably achieve this by switching to higher efficiency components. Or you could store some energy for the short bursts of transmission, getting a steady power of 13 watts.
Keep in mind that most vendors probably still have the very first itteration of hardware. It will significantly improve over the next years anyhow.
Like many other technologies there are also positive aspects.
Think about it, RFIDs need to be made cheaply. Sooner or later we will have printable microcontrollers costing a cent each, including a radio interface and a solar panel. You could do extremely subversive things with those, like building a wireless meshed network which is virtually untraceable by governments and large cooperations. You would just need to glue your little computers to lamp-posts and they would relay.
The sound quality is not the point. Playing a phonographic recording is just a whole different experience, you more or less celebrate playing it. MP3s on the other hand just are there at a touch of a button.
I guess one reasons for the rise of the vinyl record also was, that it was, till recently, one of the few media to get music legally from. MP3s were not commecially available till recently and legal CDs are hard to find. (of course there are a lot of pirated CDs on the market)
There are types of IR recievers which not only recieve IR as they were designed for, but also demodulate RF. If you cover them with tape they will still work if the remote is close enought.
On the other hand, imagine someone modulating a 10 Watt transmitter the right way, you might be able to turn off all TV sets in your room, without anything looking to suspicious. And you have a good excuse, "There was no way I could have used a TV-B-Gone as I was behind the set.", etc.
I still wonder why TV companies still bother building remote controll recievers into their sets.
Slashdot Mirror
Well actually every good consumer router already does that. For example with mine, I can define 4 types of packages and impose certain bandwidth levels on them, as well as set priorities. It's a basic feature of many DSL-Routers.
40 years seriously is _fast_ in the computing business.
Just look at the GRAIL-system from the end of the 1960s. (It's somewhere in the 3rd 3rd I think, it has flowchart like graphics on the screen)
http://www.archive.org/details/AlanKeyD1987
That old system is still more advanced than anything you can buy today.
40 years is essentially nothing in UI development. There are lots of usefull concepts to be explored. Like typed natural language interfaces.
Actually when the lightpen was invented they already found out why it was a bad idea.
It's very unformfortable to use it for prolonged periods.
Well how could one exploit such a system. It doesn't cost anything more to route than to just be a part of the network.
Well tracking is an integral issue with those cell-"phone" networks. Unfortunately the operators even want it to provide location-based services.
I guess that 5G networks will be meshed and user-driven. It's just an ingenious business model. The users provide an open network, and phone-operators can just use that network, simmilar to VoIP-Providers using the internet. It's just so much cheaper to give away a few thousand $20 WLAN-Routers as well as free low bandwidth (1MBit max) internet access as to build up a few base-stations.
A business-model could look like this:
ISPs agree on technical standards for meshed wireless networking. Then they let companies build boxes for that kind of networking and sell them, or give them away for free.
After a network has been buildt, those ISPs offer VPN-tunnels into the internet, as well as other services for a fee, or finance that with advertisements. It's essentially the web business-model brought to networking. You have a big network where everybody shares the cost, and you can use it any way you want.
I think you are missing the point here.
User-owned wireless networks can provide 2 features, cell-phone networks cannot.
1. They are affordable.
2. They are free of corporate censorship.
Cell phone networks are still designed with low-bandwidth voice in mind. They do not really scale beyond a few megabits per cell, even with 3G networks which we will be stuck with for the next decades.
There also is a more dangerous consequence. Traditional cell phone networks need to have a preety good clue to where your mobile station is. That means people will be able to easily track you and follow every move, even without the need to get out of their office. In times of governemnts all over the world getting more and more fashist, this is definitely not good.
We need user-operated networks.
Actually in many cases all that's needed is a wireless network card and a good antenna. Every computer can easily act as a meshed networking router.
For example we currently set up such a system at our dormatory. We will set up one server for a wiki and some ftp-space, but otherwise the only infrastructure will be the computers of the users.
We do not have any internet access in that network, except for a proxy that relays into the tor network.
If there is any interrest you can read my article in the Winter 2007-2008 issue of 2600, page 9. If there is even more interrest, I can post my raw version here.
There are simple ways to solve that problem.
Just install an anonymous proxy so all the http-trafic goes via tor or something.
Besides, there hasn't been a single case yet in which something like that has happened.
Furthermore, who says you need to provide internet access via such a network? Just a general local network would be sufficient. One lawyers cannot look into easily without actually going to the neighbourhood.
I guess the main issue will be backward compatibility. This preety much seems to be the main complaint about Windows Vista. A "new" version will even be less compatible. Even if they have some VM-based compatibility system it's not likely to be based on XP or 2000.
The main selling point of Windows is not the OS itself, but the fact that you can run software on it written over the last decade. Even in the year 2000, many businesses were running DOS applications.
In fact it is very rare that you find Windows applications actually using new features.
Most Windows software is legacy software. No developer would write for Windows 7 if that meant the software wouldn't run under Windows 2000 or Windows XP anymore. They might as well write it under GTK and just port it to whatever platform they want.
Some people talk about a recession, of course that is not quite true. What will come is a stagflation which is a recession combined with an inflation.
Of course engeneers might be laid off first because of the idiocy of the managers, but that will not make the situation any 'better'. With no developers or service-persons, and no money to afford foreign workforce they will not be able to sustain their business.
But unlike them you are able to go on living. You can use an old car and convert it into a power plant for generating electricity. You can use LEDs from old devices to give off a bit of light. You can turn junk into something usable. Those will be sought after then. You will be able to earn your food while ecconomists will have to starve, or try to continue to live their highly parasitic lifestyle.
It sounds like written by a 15 year old.
Of course COBOL is still around. There currently is virtually no other language beeing able to replace it because they lack BCD arithmetics for easy to predict precision.
Also there is another one missing. "Developer for organically grown commercial computing enviroments". If you ever had to programm something serious under Windows, and then later tried another OS you will know what I mean. You have Interfaces which are borderline bizarre, partly undocumented, and only plausible if you take decades of company policy into account. It is even more frustrating when you find an error in the interfaces you are using. It is virtually impossible to reach anybody who even remotely cares for your problem, even if the code which shows the problem only has 5 lines and is an exact copy of one of the examples giving out by the people writing the interface.
There may be a need for a "new" net, but the goals would be completely different.
Such a network would need to provide things like distributed caching by default and censorship resistance, as well as anonymity.
For example the network would cache all cachable protocolls by default, as often as it can be done. Then no site could be slashdotted again as many of the routers in between would just cache the content. A great side effect is that the identity of the originator of the request would be obscured by the routers.
Another important point is that it must not have any "single points of control" like the DNS-system or IP allocations.
Furthermore we would need to focus on every participant beeing able to route. The network must not be tree-like anymore. If you have wireless LAN and your neighbour has, too, there must be automatic peering.
Another idea would be to make it work on scaresly connected networks. Imagine you have a mobile device. It could try to fetch your encrypted (!) e-mail and fetch it whenever you have a connection. Every router in the connection would try to accept the request and cache the response until you have a connection again.
Ohh, you are confusing something.
While in most commercial software packages, the developers and users are strictly seperated, in FOSS they are typically closely coupled.
So more often than not, commercial companies bring out software that nobody wants and wont work propperly. A good example is all the malware around. Software that spies on you, displays ads on your screen, sends out spam or makes your computer participate in a DDOS.
While with FOSS you actually get software which is typically as good as resources allow, and not just as good as the business model allows.
I guess most POS equipment runs on dedicated hardware.
There is no reason not to use Linux. Barcode scanners emulate keypresses. There are no usable software interfaces for POS devices like customer displays or bon printers. Those are written by directly accessing the serial and parallel ports. A task that is extremely frustrating under Windows.
Depending on what you want. You could work with simple shell scripts and perhaps some awk or perl. If you don't want to store information about the customer it can be extremely simple. Essentially a table in an SQL database which stores the article number, the price and the date/time. You can easily get many pieces of information from that.
This seriously works for commentaries. Al Jazeera International does it since it's start. You can just send your video commentary about any topic to them, and if it's good they will publish it.
And seriously, at the current state of CNNs news coverage, even getting reports about broken pickle jars in supermarkets would be an improvement. I mean they will surely censor out the good stuff anyhow. Stuff that would deserve to be in the media.
Seriously, do you really think the US television would even report on it?
Wars aren't that popular anymore, so it's probably not a good idea to brag about it in the media.
The war will start soon, but probably the only tv station reporting about it will be Al Jazeera International.
Ohh yes, that guy who found out that the wiimote exist.
Well my personal favourite idea I want to do some research in is to build a huge "Rand-Tablet". Such a tablet is essentially a simplified Wacom tablet. You have a grid of thin wires behind your surface through which you run pulses or bursts of current. Your pen consists of a coil (perhaps with amplifier) which picks up the magnetic fields generated by the coils. Now as you know which wire is acticated when, you can simply measure the relative distances of your pen tip from the nearest wires and therefore calculate the precise position of your pen.
As those wires can be really thin, you can simply place them a few inches behind a translucent projection screen, making them essentially invisible.
Of course you can use multiple pens, and if you are good, you can even try to use the wires as antennas to read back the signal of the pen wirelessly. But for public terminals wired pens might be a better idea.
1. Computers won't get much smaller. Just think of it, you can now buy a cheap C64 clone which consists of a single chip, but yet they build a bulky case around it so it will be usefull. Just look at Laptops, they get bigger every year. Today you can buy laptops with 17 inch screens.
2. Computers won't have hundreds of Chips. The number of Chips will most likely decrease with nanotechnology. Even today we see a trent towards as little chips as possible. What the author might have meant is that we might have multi processor systems. Yes, that will probably happen.
Anyhow my guesses are as follows:
There will be a new medium sized breakthrought like the Internet, and Microsoft will again, completely miss it.
Maybe there will be a network access revolution: The Internet has abolished the traditional publisher/consumer relationship as everyone could publish for about the same cost as reading. There are already the first data-networks which try to push that idea down to the network level. In those networks you don't have an ISP anymore, but you just peer with your neighbours (for example over WLAN). Just think about it, if you lay a cable to your neighbour, you would easily get a gigabit of bandwidth. And even with moderately cheap switches you can easily scale that up. There may still be ISPs, but their importance will fade. This is the vital key to having network access everywhere. Without that you just will be carrying around homing devices which you don't dare using because every use costs you a significant amount of money. User generated networks are just far more efficient.
Voice access will come, but mostly to cheap devices. Over the next years the cost of implementing voice recognition might be significantly lower than the one of implementing buttons and displays.
Software might _slowly_ approach the state of the art of the 1970s again. We might get systems the average person can actually use after a day of training or so. Maybe we need to get rid of "I want that feature"-languages like C, C++, C# or Java and get back to the more "scientific" languages like Smalltalk or Lisp. One yet has to show me a system that actually does anything which would be hard to implement in Assembler in one of those "new" languages.
But realistically, I don't see that much change in the future. The computer industry just got to commercial. Back till the mit 90s you were at least able to catch a glimpse of the next 10 years by looking at 'workstations', but today the workstation market has completely died.
ARCNet is still alive and well doing. It's used in areas where precise timing is important and you'd like to have LAN-like networks. For example quite a bit of TV studio equipment has ARCNet to communicate.
For example this controll panel for video cameras http://www.ikegami.com/br/products/hdtv/mcp200.html has ARCNet connectors.
Well as long as the average user still doesn't do anything, or constantly clicks 'accept' everywhere there might not be much 'extremely malicious' software around. (Althought I have to say some ways of controlling a web-browser are simpler than opening TCP/IP connections.)
Temporarily disconnecting a computer won't do much good, as the software could just wait, or refuse to start without an internet connection.
Then there are the holes created by 'personal firewalls'. Not only have there been dozens of buffer overruns, but also logic errors. For example some 'personal firewalls' allow you to block transmitting personal information. They run a little proxy server which filters out queries containing certain names and numbers. Let's imagine you entered your 4-digit PIN number there. An attacker could just get you to a website (or several) with 10000 tiny little images numbered 0000.jpg to 9999.jpg. Your browser would eventually load all of them, except for the one containing your PIN-Number. With a bit more thought that can be changed to to give you hints what the personal informations are.
You do realice that a piece of software on your own computer cannot stop any of that. Simply put, how should that programm find out if you are surfing the web or if a programm is controlling your browser to phone home?
Personal firewalls suck, it has been proven that they usually can be penetrated from both sides.
They also confuse the user and teach him to "click accept or nothing will work", which is barely something you want your user to do.
It's a much more sensible advise to tell your user to turn off unnessesary services, especially since there are now simple applications which do that work for you.
http://www.dingens.org/index.html.en
Essentially just because vendor A has devices which consume "up to 18 watts" it doesn't mean that nobody can build devices which take less than 13 Watts.
That's just 5 watts difference. You could probably achieve this by switching to higher efficiency components. Or you could store some energy for the short bursts of transmission, getting a steady power of 13 watts.
Keep in mind that most vendors probably still have the very first itteration of hardware. It will significantly improve over the next years anyhow.
Like many other technologies there are also positive aspects.
Think about it, RFIDs need to be made cheaply. Sooner or later we will have printable microcontrollers costing a cent each, including a radio interface and a solar panel.
You could do extremely subversive things with those, like building a wireless meshed network which is virtually untraceable by governments and large cooperations. You would just need to glue your little computers to lamp-posts and they would relay.
The sound quality is not the point.
Playing a phonographic recording is just a whole different experience, you more or less celebrate playing it.
MP3s on the other hand just are there at a touch of a button.
I guess one reasons for the rise of the vinyl record also was, that it was, till recently, one of the few media to get music legally from. MP3s were not commecially available till recently and legal CDs are hard to find. (of course there are a lot of pirated CDs on the market)
There are types of IR recievers which not only recieve IR as they were designed for, but also demodulate RF. If you cover them with tape they will still work if the remote is close enought.
On the other hand, imagine someone modulating a 10 Watt transmitter the right way, you might be able to turn off all TV sets in your room, without anything looking to suspicious. And you have a good excuse, "There was no way I could have used a TV-B-Gone as I was behind the set.", etc.
I still wonder why TV companies still bother building remote controll recievers into their sets.