Being a native speaker of German, I have some trouble understanding your.sig - it looks like an online translation of something like "I'd like to have sex with my own army of Llama whatever". Could you share this precious piece of information with the rest of us?
You may accuse me of talking about a very limited field, but I need to do multi-language word processing, and in that market there is practically nothing on the PC market that comes even close to the comfort and ease of Nisus Writer 5.1.3 on the Macintosh with Apple's Arabic, Cyrillic, Hebrew and Japanese language kits installed under MacOS 9. The closest thing in terms of output quality is TeX; the closest thing in terms of input ease is MS Office 2000 under Windows 2000, but the hardware requirements are terrible. Nisus runs fine on an old $100 PowerBook 170 (albeit under OS 7.6, for which language kits exist as well, though), while for Office 2000 you need something in the $1000-and-up range.
I'm running a small server in a bit of a jeopardized environment (it runs, among other things, an anonymous remailer). It runs a software called Port Sentry by Psionic, among other stuff. I think the software is qute qood, and I'm posting it here so that you might give it a try.
Portsentry works by attaching itself to a configurable number of ports. When it detects a port scan, it simply uses firewall rules to block IP traffic to the address the scan came from. You can easily configure your firewall to drop the respective rule after some time. It detects TCP as well as UDP scans at a configurable sensitivity. If you're evil-hearted, you can configure it to put in some retaliatory action, but I wouldn't recommend that because you don't know whom you're hurting. On systems that don't run a firewall, the IP address can be dropped by assigning it a dead route.
While the methods employed are quite basic, it's a very useful, more or less minimalistic tool that definitely adds to the security of the site. I'm using it under FreeBSD 4.1-STABLE, where it's in the ports collection, but it says it works for other BSDs as well as Linux and ipchains-based firewalls just as well.
After parting with the Atari Mega ST keyboard with RTS keycaps that I was using for ages (because it had one microswitch for each key - unbeatable reliability and quite good feedback, and with the better keycaps the keys were even discernible without looking;-)), I sticked for some time (literally) with an Escom $8 keyboard that I eventually replaced after I had spilled some apple juice on it. I tried to remove it all, but half a year later suddenly some keys started working rather erratically. When I took the keyboard apart, I found a think layer of green fungus that was happily evolving in my keyboard. While it was a bit sad to throw this #1 biological experiment away, it needed replacement sooner or later. Now I'm on a Logitech cordless keyboard which, even though blamed a lot, feels quite nice. The click might be more audible, but the tactile feedback is there (and I'm a heavy typer).
Well, that was my share in the universal disgust contest that seems to be developing here.:-))
I'm using OS 9 at the moment (no, not OS-9, rather Mac OS 9), and its footprint is somewhere around 27 megabytes. Still quite a bit if one has only 32 MB of RAM, but when I disable stuff like OpenGL that I usually don't need, I can cut it doen to 19,8 MB which is quite reasonable. LinuxPPC uses more on my PB G3 even without X running.
I have used various portables for a while, and I must say that even though 1024x768 on a 12" screen is a bit small, one can work with it quite comfortably with things like font antialiasing turned on.
On the Mac, the problem is, however, that some software does not support enlarging the screen. On X11, I can manually adjust the screen resolution, thus gaining some control over the desktop's appearance. Even Windows with its font enlargement mechanism caters to this, even though the two levels it offers are both a bit odd and everything in between looks a bit chunky.
For example, when I use software like Nisus, the best multilingual word processor I've ever had the chance to work with, I find that it displays font sizes as pixel sizes - a 12 pt font is displayed as 12 pixel screen font. You can't adjust this manually, so even on the 14,1" screen on the PowerBook G3 that I'm using at the moment the text looks a bit small. Antialiasing doesn't help - while it usually yields better legibility, at these small sizes it makes no difference, and in special cases when you need stuff like "h with dot under" or Arabic vowel signs, it is plain impossible to discern things with antialiasing turned on, so I usually leave it off. That way, however, fonts look a bit clobbered. That effect will be a bit worse on the small screen.
I realize, though, that this problem isn't as bad as it sounds now, because Apple is introducing OS X in parallel anyway, where most of the display routines are better virtualized, being more or less directly PostScript-based. It will be much more easily possible to customize screen display than on most other platforms I know, so in the end you get the added resolution on the smaller LCD for finer display of details. I think it will be possible to live with it.
In Bonn university, where I study, it is quite common that people copy entire books if they need them for study. This happens for good reasons, because when you need books that were published in 1767 for your studies rather extensively, the university can't expect you to buy the thing antiquarically.
I think that at least over here in Germany, it cannot be guaranteed that everybody can buy all of the books that they need for studying, even when it's not the extreme case of books out of print for 200 years. Personally, I study computer science, neural information processing, Islamic studies and Arabic linguistics. In three of these subjects, it is extremely difficult to get hold of the needed books at affordable prices. If I want to read something by an Arabic author, nobody can expect me to hop on a plane to Cairo to buy it there. And if it's a library where more than one person may need the book in question, the logical solution is: copy it and carry it home.
As far as I know, copying excerpts from books for educational reasons is legal in Germany at any rate, and as far as I can remember it, copying books for personal educational use is free even if you take the whole book.
It seems that Mr. Metallica has never been studying or even bothered to ask anyone familiar with university environments before making such statements. Honestly, even though the average Harvard student is richer already than I can hope ever to be, I don't believe that you're officially forbidden to copy a book if you happen to need it.
The problem I'm having with this piece of technology is that I utterly fail to see any purpose apart from telling the time of the day.
You may rightfully accuse me of being a bit backwards with my choice of gadgetry, but I'm afraid that my present watch, a mechanical Swatch Irony that runs Coil 1.0 and Mechanical Windup 0.99b3, does not do. I realize that the Swatch is not skinnable, though, and not released under the GPL either.
Is there anyone who has a formal copyright on any of the classic rogue-style games, like Rogue, Amulet of Yendor (aka Hack), NetHack, Omega, Moria, Angband and so on? Come on, let's sue Blizzard for creating Diablo. It's a bit like that, isn't it?
As I said, we're talking about one-use-only numbers anyway. I don't see very much of a difference between transmitting the number to the merchant encrypted or unencrypted, since it's going invalid right after that anyway.
Well, since we're talking about one-use-only numbers anyway, I don't see very much of a difference between transmitting the number to the merchant encrypted or unencrypted, since it's going invalid right after that anyway.
Someone else encrypts your card number as well as the amount of money using the merchant' card number. You're basically right where you are now without encryption.
This may sound like a good idea, but it has its drawbacks.
The first drawback is granularity.
If the value of every single card is large (few hundred dollars), it would be a mugger's paradise because people don't usually carry around much in cash, but a potentually valuable payment card would be a good target.
If the value of every single card is small, no one will use it for larger transactions. You can buy your roll of bread quite comfortable using real money, and if you have to enter a dozen numbers when you buy your new $99 sound card online, the system is not going to be very popular. We've had this in Germany because it was considered to use prepaid phonecards for transactions. The idea was dropped, however.
The second drawback is non-rechargability. If recharging devices were available, people would start stealing those and recharging their cards at will. To make this impossible, one has to provide each card with a sort of "shadow bank account" and have the recharger communicate with some central authority. Then, you could desable known stolen rechargers.
The third and worst drawbacks is that if it's an electronic device, you can fake it. I spent some time in 1996 assembling a microcontroller-based board that could pretend it was a German phonecard. No one would introduce a payment card that could be faked this way. In order to stop this, one has to introduce either advanced secret card signing algorithms, which are sure to either leak out or be faked sooner or later, or use shadow accounting like with the German GeldKarte ("money card"). Again, anonymity and non-traceability can no longer be guaranteed, and the advantage will be gone.
A very good introduction how the German GeldKarte payment card system works can be found here. I'm sorry that it's all in German, but the system is specific to Germany, so most people wouldn't bother to translate it. You can try the fish, though. An English introduction can be found at Manni's page.
This may sound like a good idea, but it has its drawbacks.
The first drawback is granularity.
If the value of every single card is large (few hundred dollars), it would be a mugger's paradise because people don't usually carry around much in cash, but a potentually valuable payment card would be a good target.
If the value of every single card is small, no one will use it for larger transactions. You can buy your roll of bread quite comfortable using real money, and if you have to enter a dozen numbers when you buy your new $99 sound card online, the system is not going to be very popular. We've had this in Germany because it was considered to use prepaid phonecards for transactions. The idea was dropped, however.
The second drawback is non-rechargability. If recharging devices were available, people would start stealing those and recharging their cards at will. To make this impossible, one has to provide each card with a sort of "shadow bank account" and have the recharger communicate with some central authority. Then, you could desable known stolen rechargers.
The third and worst drawbacks is that if it's an electronic device, you can fake it. I spent some time in 1996 assembling a microcontroller-based board that could pretend it was a German phonecard. No one would introduce a payment card that could be faked this way. In order to stop this, one has to introduce either advanced secret card signing algorithms, which are sure to either leak out or be faked sooner or later, or use shadow accounting like with the German GeldKarte ("money card"). Again, anonymity and non-traceability can no longer be guaranteed, and the advantage will be gone.
A very good introduction how the German GeldKarte payment card system works can be found here. I'm sorry that it's all in German, but the system is specific to Germany, so most people wouldn't bother to translate it. You can try the fish, though. An English introduction can be found at Manni's page
This is a really good idea, a bit like the unique transaction number system used in electronic banking - once the number is used, you can't use it anymore, nor can anyone else.
The problem I am seeing with it, however, is that it gives a whole new field of application to the credit card number generator programs that the usual cracking groups have been cranking out for the last year or so (like DisCard online, available from places like New Order). The algorithm by which AmEx would be generating their credit card numbers is going to be a company secret, of course, but so are the algorithms by which they are generated now. If anyone was able to provide a similar algorithm that generates disposable credit card numbers from someone's semi-public fixed account numbers or whatever static personal data AmEx would be using, it would probably lead to a new generation of credit card frauds ("What do you mean, you haven't used this number yet? So what about the transaction on XX-XX-XXXX?").
Well, to pick up your example, neural networks often adapt more quickly to changes in the input set and tend to yield better mappings in the input/output space than procedural solutions.
Applying this to IT employees, non-trained IT folk are often more flexible and willing to try something new than trained people are. Trained people, however, have often a better potential for abstracting causes out of problems. (Or am I underfitting here?;-))
Result is: both have their places.
(Wouldn't have needed to descend into AI theory to come to this point, though)
I was in the first wave of membership registrations from Germany after the excellent publicity campaigns run by c't and Spiegel online, and hence I received my PIN letter quite early. I was a bit suprised when recently I found another letter with another PIN and along with it received an email with another number and password.
I managed to activate the second membership without further problems. I was not able to endorse another candidate, though, because the server said that my second membership had been "placed at -3" for security reasons and refused to let me place another vote. I sent mail to ICANN asking for deletion of my second account, to which they haven't answered, though.
It might be interesting to know if there are others out there enlisted at ICANN with more than one membership account, and if any of these were able to cast more than one vote, which I would find rather disturbing and democratically unsatisfying. As far as I am concerned, I mainly see technical problems behind this, but for the conspiracy theorists among you, this is probably another bit of meat to chew on, as is this article on Spiegel online (in German) about the "revolt of the voting cattle" and fears of technoanarchists dominating ICANN.
I've got two questions that were left unanswered so far, so if anyone could help me out with these, I'd really appreciate it.
Since a satellite usually does not project directional signals for different customers, it would seem that the signal coming from the satellite contains all the downstream data for all customers within a certain geographic region. This leads to two subproblems:
Bandwidth might become a problem in more densely settled areas. You might argue that satellite is not for densely settled areas anyway, but the area covered by a satellite in geosynchronous orbit is rather large.
Security can not be guaranteed since anyone with slightly modified software can eavesdrop on every other connection through the same satellite. You might argue that it'd probably be encrypted, but still, I have a problem with the theoretical possibility of this happening. I'll be waiting for the first public decryption challenge: "Participate in our Satellite Decryption challenge! Clients for all major free operating systems included!!";-)
How does the satellite distinguish between different uplinks from different customers? They can't probably do that simply by discerning the direction vectors of the incoming signals because the resolution necessary to do that in geosynchronous orbit isn't even possible in military satellites, at least in the comparatively long wavelength regions used for data communications. The solution is probably that they assign time frames to the clients and each client gets to occupy time frames in order. That means that when a lot of clients are online, upstreams are likely to be rather chunky and unreliable. Is this the case or am I missing something?
If there's anyone out there who knows a bit about how these technical details are tackled, I'd appreciate to hear about it...;-)
I don't know what's going on in the powerline market over in the States, but here in Germany this has been in the media for quite some time (if you can read German, then check out this list of articles that were on the Heise news service over the last few months.
Development in the field is rather active over here. You can buy home spin-off solutions that are based on powerline communication already. One application that is already being sold is a (phone line) modem whose serial port is replaced by a powerline adapter, and by hooking up your computer to a second adapter you're able to access your modem from anywhere within your local house circuit. I am not quite sure what implications this has on security, but I am fairly sure that some measures are being taken. This is not really high-speed, though, even though it seems fairly reliable.
Powerline Internet access is a different matter, of course. In Cologne, they will be starting to sell powerline-based Internet and telephony by the beginning of 2001. A couple of field tests, also on a larger scale in a somewhat more public environment, have already been conducted.
The speed that is being claimed varies. Preussen Elektra (recently merger'ed into e.on Energie) claim that their particular system is capable of reaching 10 Mbps in-house and 2 Mbps for incoming/outgoing Internet access. Siemens claim they reach 1.3 Mbps over public lines and plan to extend this to 10 Mbps. So as far as cable or DSL are concerned, this is quite a competition.:-)
The central problem with powerline communications is that your average powerline is just a pair of wires arranged in an unpredictable network topology, and that the behaviour of the electrical properties of the system tends to be a bit difficult to handle because most electric devices emit quite a bit of noise. Take a look at the noise emissions from a 100 to 300 W dimmable ceiling lamp, for example, and then you'll immediately see why powerline network access took this long to develop. It appears that they got this quite under control, though.
The final problems introducing this over here appear to be of a legal sort, because there are quite strict regulations in Germany as to which emissions are allowed in which part of the spectrum. With powerline communications, one has the problem of the non-shielded wires acting like a very large antenna, so they have to take care as to which frequencies they're using and how they're reducing emissions. The carriers needed for 1 to 10 Mbit are well in the amateur radio spectrum, for example. Nevertheless, powerline internet has good prospects for the future over here because it is by far the least expensive way by which to hook up people to the network (and since all major electricity corporations here also sell network services, they are quite interested in extending their customer base) - practically every house is connected to the powerline network already and has quite a bit of wiring installed as well.
Slashdot Mirror
Being a native speaker of German, I have some trouble understanding your .sig - it looks like an online translation of something like "I'd like to have sex with my own army of Llama whatever". Could you share this precious piece of information with the rest of us?
I wonder of what relevance this is, seeing that it's possibly only generating a lot of slashdotting on RedHat's servers?
Wouldn't it have been a bit more cooperative to let the news simmer down a bit?
You may accuse me of talking about a very limited field, but I need to do multi-language word processing, and in that market there is practically nothing on the PC market that comes even close to the comfort and ease of Nisus Writer 5.1.3 on the Macintosh with Apple's Arabic, Cyrillic, Hebrew and Japanese language kits installed under MacOS 9. The closest thing in terms of output quality is TeX; the closest thing in terms of input ease is MS Office 2000 under Windows 2000, but the hardware requirements are terrible. Nisus runs fine on an old $100 PowerBook 170 (albeit under OS 7.6, for which language kits exist as well, though), while for Office 2000 you need something in the $1000-and-up range.
I'm running a small server in a bit of a jeopardized environment (it runs, among other things, an anonymous remailer). It runs a software called Port Sentry by Psionic, among other stuff. I think the software is qute qood, and I'm posting it here so that you might give it a try.
Portsentry works by attaching itself to a configurable number of ports. When it detects a port scan, it simply uses firewall rules to block IP traffic to the address the scan came from. You can easily configure your firewall to drop the respective rule after some time. It detects TCP as well as UDP scans at a configurable sensitivity. If you're evil-hearted, you can configure it to put in some retaliatory action, but I wouldn't recommend that because you don't know whom you're hurting. On systems that don't run a firewall, the IP address can be dropped by assigning it a dead route.
While the methods employed are quite basic, it's a very useful, more or less minimalistic tool that definitely adds to the security of the site. I'm using it under FreeBSD 4.1-STABLE, where it's in the ports collection, but it says it works for other BSDs as well as Linux and ipchains-based firewalls just as well.
After parting with the Atari Mega ST keyboard with RTS keycaps that I was using for ages (because it had one microswitch for each key - unbeatable reliability and quite good feedback, and with the better keycaps the keys were even discernible without looking ;-)), I sticked for some time (literally) with an Escom $8 keyboard that I eventually replaced after I had spilled some apple juice on it. I tried to remove it all, but half a year later suddenly some keys started working rather erratically. When I took the keyboard apart, I found a think layer of green fungus that was happily evolving in my keyboard. While it was a bit sad to throw this #1 biological experiment away, it needed replacement sooner or later. Now I'm on a Logitech cordless keyboard which, even though blamed a lot, feels quite nice. The click might be more audible, but the tactile feedback is there (and I'm a heavy typer).
Well, that was my share in the universal disgust contest that seems to be developing here. :-))
I'm using OS 9 at the moment (no, not OS-9, rather Mac OS 9), and its footprint is somewhere around 27 megabytes. Still quite a bit if one has only 32 MB of RAM, but when I disable stuff like OpenGL that I usually don't need, I can cut it doen to 19,8 MB which is quite reasonable. LinuxPPC uses more on my PB G3 even without X running.
I have used various portables for a while, and I must say that even though 1024x768 on a 12" screen is a bit small, one can work with it quite comfortably with things like font antialiasing turned on.
On the Mac, the problem is, however, that some software does not support enlarging the screen. On X11, I can manually adjust the screen resolution, thus gaining some control over the desktop's appearance. Even Windows with its font enlargement mechanism caters to this, even though the two levels it offers are both a bit odd and everything in between looks a bit chunky.
For example, when I use software like Nisus, the best multilingual word processor I've ever had the chance to work with, I find that it displays font sizes as pixel sizes - a 12 pt font is displayed as 12 pixel screen font. You can't adjust this manually, so even on the 14,1" screen on the PowerBook G3 that I'm using at the moment the text looks a bit small. Antialiasing doesn't help - while it usually yields better legibility, at these small sizes it makes no difference, and in special cases when you need stuff like "h with dot under" or Arabic vowel signs, it is plain impossible to discern things with antialiasing turned on, so I usually leave it off. That way, however, fonts look a bit clobbered. That effect will be a bit worse on the small screen.
I realize, though, that this problem isn't as bad as it sounds now, because Apple is introducing OS X in parallel anyway, where most of the display routines are better virtualized, being more or less directly PostScript-based. It will be much more easily possible to customize screen display than on most other platforms I know, so in the end you get the added resolution on the smaller LCD for finer display of details. I think it will be possible to live with it.
Does FPS stand for first-person shooter or for frames per second? Sounds like they're directly related to each other.
In Bonn university, where I study, it is quite common that people copy entire books if they need them for study. This happens for good reasons, because when you need books that were published in 1767 for your studies rather extensively, the university can't expect you to buy the thing antiquarically.
I think that at least over here in Germany, it cannot be guaranteed that everybody can buy all of the books that they need for studying, even when it's not the extreme case of books out of print for 200 years. Personally, I study computer science, neural information processing, Islamic studies and Arabic linguistics. In three of these subjects, it is extremely difficult to get hold of the needed books at affordable prices. If I want to read something by an Arabic author, nobody can expect me to hop on a plane to Cairo to buy it there. And if it's a library where more than one person may need the book in question, the logical solution is: copy it and carry it home.
As far as I know, copying excerpts from books for educational reasons is legal in Germany at any rate, and as far as I can remember it, copying books for personal educational use is free even if you take the whole book.
It seems that Mr. Metallica has never been studying or even bothered to ask anyone familiar with university environments before making such statements. Honestly, even though the average Harvard student is richer already than I can hope ever to be, I don't believe that you're officially forbidden to copy a book if you happen to need it.
You probably figured out quite well what I meant. :-) Kinda happens when you don't use Preview.
The problem I'm having with this piece of technology is that I utterly fail to see any purpose apart from telling the time of the day.
You may rightfully accuse me of being a bit backwards with my choice of gadgetry, but I'm afraid that my present watch, a mechanical Swatch Irony that runs Coil 1.0 and Mechanical Windup 0.99b3, does not do. I realize that the Swatch is not skinnable, though, and not released under the GPL either.
Now that's REALLY good.
Is there anyone who has a formal copyright on any of the classic rogue-style games, like Rogue, Amulet of Yendor (aka Hack), NetHack, Omega, Moria, Angband and so on? Come on, let's sue Blizzard for creating Diablo. It's a bit like that, isn't it?
As I said, we're talking about one-use-only numbers anyway. I don't see very much of a difference between transmitting the number to the merchant encrypted or unencrypted, since it's going invalid right after that anyway.
Well, since we're talking about one-use-only numbers anyway, I don't see very much of a difference between transmitting the number to the merchant encrypted or unencrypted, since it's going invalid right after that anyway.
Someone else encrypts your card number as well as the amount of money using the merchant' card number. You're basically right where you are now without encryption.
This may sound like a good idea, but it has its drawbacks.
The first drawback is granularity.
The second drawback is non-rechargability. If recharging devices were available, people would start stealing those and recharging their cards at will. To make this impossible, one has to provide each card with a sort of "shadow bank account" and have the recharger communicate with some central authority. Then, you could desable known stolen rechargers.
The third and worst drawbacks is that if it's an electronic device, you can fake it. I spent some time in 1996 assembling a microcontroller-based board that could pretend it was a German phonecard. No one would introduce a payment card that could be faked this way. In order to stop this, one has to introduce either advanced secret card signing algorithms, which are sure to either leak out or be faked sooner or later, or use shadow accounting like with the German GeldKarte ("money card"). Again, anonymity and non-traceability can no longer be guaranteed, and the advantage will be gone.
A very good introduction how the German GeldKarte payment card system works can be found here. I'm sorry that it's all in German, but the system is specific to Germany, so most people wouldn't bother to translate it. You can try the fish, though. An English introduction can be found at Manni's page.
(Sorry for posting this multiple times :-))
This may sound like a good idea, but it has its drawbacks.
The first drawback is granularity.
The second drawback is non-rechargability. If recharging devices were available, people would start stealing those and recharging their cards at will. To make this impossible, one has to provide each card with a sort of "shadow bank account" and have the recharger communicate with some central authority. Then, you could desable known stolen rechargers.
The third and worst drawbacks is that if it's an electronic device, you can fake it. I spent some time in 1996 assembling a microcontroller-based board that could pretend it was a German phonecard. No one would introduce a payment card that could be faked this way. In order to stop this, one has to introduce either advanced secret card signing algorithms, which are sure to either leak out or be faked sooner or later, or use shadow accounting like with the German GeldKarte ("money card"). Again, anonymity and non-traceability can no longer be guaranteed, and the advantage will be gone.
A very good introduction how the German GeldKarte payment card system works can be found here. I'm sorry that it's all in German, but the system is specific to Germany, so most people wouldn't bother to translate it. You can try the fish, though. An English introduction can be found at Manni's page
.This is a really good idea, a bit like the unique transaction number system used in electronic banking - once the number is used, you can't use it anymore, nor can anyone else.
The problem I am seeing with it, however, is that it gives a whole new field of application to the credit card number generator programs that the usual cracking groups have been cranking out for the last year or so (like DisCard online, available from places like New Order). The algorithm by which AmEx would be generating their credit card numbers is going to be a company secret, of course, but so are the algorithms by which they are generated now. If anyone was able to provide a similar algorithm that generates disposable credit card numbers from someone's semi-public fixed account numbers or whatever static personal data AmEx would be using, it would probably lead to a new generation of credit card frauds ("What do you mean, you haven't used this number yet? So what about the transaction on XX-XX-XXXX?").
Well, to pick up your example, neural networks often adapt more quickly to changes in the input set and tend to yield better mappings in the input/output space than procedural solutions.
Applying this to IT employees, non-trained IT folk are often more flexible and willing to try something new than trained people are. Trained people, however, have often a better potential for abstracting causes out of problems. (Or am I underfitting here? ;-))
Result is: both have their places.
(Wouldn't have needed to descend into AI theory to come to this point, though)
I was in the first wave of membership registrations from Germany after the excellent publicity campaigns run by c't and Spiegel online, and hence I received my PIN letter quite early. I was a bit suprised when recently I found another letter with another PIN and along with it received an email with another number and password.
I managed to activate the second membership without further problems. I was not able to endorse another candidate, though, because the server said that my second membership had been "placed at -3" for security reasons and refused to let me place another vote. I sent mail to ICANN asking for deletion of my second account, to which they haven't answered, though.
It might be interesting to know if there are others out there enlisted at ICANN with more than one membership account, and if any of these were able to cast more than one vote, which I would find rather disturbing and democratically unsatisfying. As far as I am concerned, I mainly see technical problems behind this, but for the conspiracy theorists among you, this is probably another bit of meat to chew on, as is this article on Spiegel online (in German) about the "revolt of the voting cattle" and fears of technoanarchists dominating ICANN.
That was really informative :-)
I've got two questions that were left unanswered so far, so if anyone could help me out with these, I'd really appreciate it.
If there's anyone out there who knows a bit about how these technical details are tackled, I'd appreciate to hear about it... ;-)
I don't know what's going on in the powerline market over in the States, but here in Germany this has been in the media for quite some time (if you can read German, then check out this list of articles that were on the Heise news service over the last few months.
Development in the field is rather active over here. You can buy home spin-off solutions that are based on powerline communication already. One application that is already being sold is a (phone line) modem whose serial port is replaced by a powerline adapter, and by hooking up your computer to a second adapter you're able to access your modem from anywhere within your local house circuit. I am not quite sure what implications this has on security, but I am fairly sure that some measures are being taken. This is not really high-speed, though, even though it seems fairly reliable.
Powerline Internet access is a different matter, of course. In Cologne, they will be starting to sell powerline-based Internet and telephony by the beginning of 2001. A couple of field tests, also on a larger scale in a somewhat more public environment, have already been conducted.
The speed that is being claimed varies. Preussen Elektra (recently merger'ed into e.on Energie) claim that their particular system is capable of reaching 10 Mbps in-house and 2 Mbps for incoming/outgoing Internet access. Siemens claim they reach 1.3 Mbps over public lines and plan to extend this to 10 Mbps. So as far as cable or DSL are concerned, this is quite a competition. :-)
The central problem with powerline communications is that your average powerline is just a pair of wires arranged in an unpredictable network topology, and that the behaviour of the electrical properties of the system tends to be a bit difficult to handle because most electric devices emit quite a bit of noise. Take a look at the noise emissions from a 100 to 300 W dimmable ceiling lamp, for example, and then you'll immediately see why powerline network access took this long to develop. It appears that they got this quite under control, though.
The final problems introducing this over here appear to be of a legal sort, because there are quite strict regulations in Germany as to which emissions are allowed in which part of the spectrum. With powerline communications, one has the problem of the non-shielded wires acting like a very large antenna, so they have to take care as to which frequencies they're using and how they're reducing emissions. The carriers needed for 1 to 10 Mbit are well in the amateur radio spectrum, for example. Nevertheless, powerline internet has good prospects for the future over here because it is by far the least expensive way by which to hook up people to the network (and since all major electricity corporations here also sell network services, they are quite interested in extending their customer base) - practically every house is connected to the powerline network already and has quite a bit of wiring installed as well.
Sorry. Typos happen every now and then. Apology goes to the OpenBSD enthusiasts out there. :-))