Step #2: Congratulations, you have found Slashdot. Now present a theory and have the/. users write down the practical ideas which will be "Leech Computing, Part 2".
There is a difference between being in a public space, being watched and having your every move recorded.
First, if you are in a public space, usually nobody pays attention to you except for those who are interacting with you. That's fine.
Second, being watched: This is already an unusual situation, in which many people feel uncomfortable. Think stars: They regularly complain about lack of privacy. Being watched usually has greater impact on the behaviour of people because they notice that they are watched, which is not the case with CCTV cameras. One of the activities of anti-surveillance groups is therefore to create the same level of awareness by showing people that they are watched (pointing to the camera, creating camera-maps, etc).
Third, recordings: If you think about it, you not only have to relate your actions to the current situation but you also need to think about how your actions may look like from a distance, space and timewise. This can seriously inhibit natural behaviour and is the reason why, for example, many people liked usenet until archives of it were created.
Some people feel that even perfectly normal behaviour can have a negative impact on them when seen in a different, maybe willfully distorting, context.
DSL is capable of several MBits/s. There are different types of DSL, but for ADSL, an FAQ mentions up to 8 MBit downstream and 1 MBit upstream. Since residential customers usually pay for and get less than that, there is extra bandwith which could be used for providing wireless access.
This is the same thing. I was misled by the "Rolling DSL and Wireless Access Out In One Swoop" headline. Nevertheless, I think that using otherwise unused bandwith from customer DSL lines (the difference between what the customer is paying for and what the DSL-modem is capable of) to provide wireless access to an area is quite an intriguing idea.
Ok, looks like the article is a little misleading. "Wireless coverage" is not really the goal of this approach, but merely the means to provide residential broadband internet access. It makes connecting distant customers feasible because up to 40 subscribers can be connected to one one "airhead", which is a special box connected to an uplink, which in turn can be either wired or wireless (but not through other airheads). The cost of installing an uplink can therefore be split among 40 customers, and because the "airhead" is so small, renting an office for hub hardware isn't necessary. This whole concept doesn't appear to be based on 802.11b at all.
Sorry, now I get it. I had thought they wanted to provide a wireless network as a bonus to normal DSL. Stupid me. I guess I'll watch the demonstration from Nokia before I continue commenting.
If they do it in a responsible way, they split the bandwith (x MBit for the customer, rest for the WAP), but without sharing the broadcast domain. DSL modems usually have an ATM mode which could be used to create separate channels, guaranteeing QOS to the customer and keeping the wireless traffic away from his network socket.
Yes, but typically a DSL connection is not using the maximum possible bandwith. They probably use a fixed part of the bandwith for the paying customer and any excess bandwith is allocated to the wireless access point.
There are quite a lot of reasons not to use.doc as the file format:
.doc is not an open format, which makes verification impossible. That is also the reason why there can never be a 100% compatible alternative program.
.doc is not a fixed format. This can lead to trouble with archives. What looks correct may not work in a future version of Word. It's also a major drawback for interchangeability.
.doc contains much more information than is necessary or even visible under normal circumstances. It is an indiscreet file format.
.doc is context dependant. It is not a complete description of a document.
.doc can contain malicious code. Some form of active content may be desirable, but if it is implemented, the effects of the code should be restricted to the document itself.
One file format which does fit most of the requirements is PDF.
Re:video interview with one of the creators
on
Berlin's Robotic Pub
·
· Score: 1
They do have a website (but it's quite minimalistic, no fotos). I also found this
introductory text (use the fish). There you'll find some more info about the installations and links to the otherwise hidden membership applications on the Automatenbar website.
Current codecs are a lot more efficient than you think. 32MB per minute is about 500kB/s, which is enough to transmit almost DVD-quality video. For small displays, like those of handhelds, the rough estimate is more like 1 or 2 MB per minute of video.
Maybe if they had just played PI as sound (base 16384) their ratings wouldn't suck. There's some pretty neat stuff in there. Sure, there's a lot of rubbish in between, but...
In a supposedly free world, the question to be asked is why it shouldn't be allowed to exist. The default is "anything goes", right? It doesn't hurt you if you ignore it, unless of course the majority thinks it's cool and buys the stuff. Which brings us back to "Don't buy it if you don't want it." Ensuring that customers make informed decisions really is the key.
If people buy it, so be it. It's a majority vote and those who complain that they haven't been heard on the issue just don't recognize the facts. Sometimes the road is a little longer than what looks like a straight connection from a to b to you. The best mechanism which is available to us is market dynamics, or isn't it? You are free to try to convince others, which is what I'm doing when I say "Make sure you really want it before you buy it", but the voting is done with the money being spent on what people want and therefore buy.
I'm referring to the devices and the content. If the device doesn't let you do what you want to do at a price which you are willing to pay, don't hope that somehow someone will offer to change that. If a device is crippled with unavoidable DRM, look for alternatives and if there aren't any, don't buy. People need to look at the whole product, the device and the service which is offered (not what is promised to be available in 6 months).
You can't build the equivalent of a high end Cisco router with off-the-shelf PC hardware. PCI (the 33MHz 32Bit variety found in today's PCs) maxes out at 133 MByte/s in theory. Practical values are much lower than that, so there's no way to route a gigabit stream through a standard PC.
BUT: They don't need to. "Because of the local area, unrouted nature of the network, Citylink can use inexpensive Cisco 3524 switches instead of more costly T1 routers". They only use routers at the edges of the network, where speeds are much less than 1 GBit/s. And even if someone needed the full speed there, the router is the customer's hardware, they don't have to go for the 2500$ PC hardware thing if they need more.
The problem with this approach may be scalability: "De Wit says adding quality of service (QoS) features isn't necessary." and "We can fit all the traffic we want onto our Ethernet, so why do we need to worry about prioritizing?" Also, ethernet switching doesn't work beyond a certain network size. Sooner or later, they will have to route. If they're lucky, gigabit router prices have dropped, 10GBit ethernet is ready and everyone accepts the upgrade cost by the time the assumption "The network is all switched and offers all the speed we need, don't worry" is no longer true.
But wouldn't you have to reveal who used a given ip address at a specified time in case law enforcement asks you for it? How would you do that, considering network access is anonymous, and failing that, wouldn't you become liable yourself?
Let's suppose you provided freely accessible wireless network infrastructure to anonymous users, what would be the legal implications of such an endeavour? Would you become liable for damage created with the help of your network? What if you provided no access to the internet? Would you still be liable? Is there a way around these problems? I guess once again lawyers may be the biggest hurdle before public networks become a reality.
Sorry to disappoint you, but automatic parallelization of algorithms is a field of research in which quite a few obstacles still need to be overcome before something like you propose would enter the mainstream. What does work right now is the distribution of already parallel processes/threads to several cpus, but that requires either several applications to run or a multithreaded application. For anything worth running on multi-processor hardware, the parallelism inside the operating system gives only negligible advantages.
I'm supposed to download a program that sits in the tray and records all the keys I press? I mean, to count them, they all have to pass their code, right? And it has network functionality... HELLO? Security?
Wireless networking on the whole campus is nice, of course, but it isn't educational if there is no educational content or projects which make use of the network. Looking at stories about the bandwith demand at universities, I guess the networks are mostly there (although not always wireless), but the on-topic content is missing. I'd say, put the money into virtualizing lecture material and developing new forms of presenting educational material. Some things can be expressed much better in an animation or interactive 3d-model for example, ways of presentation which are usually unvailable today.
There's another thing vinyl DJs might miss: The switching of real records. Grabbing a record from your set and selecting the song by setting the needle to the right part of it (using the visual cues you described) is probably faster than relying on a scroll-point-click-point-click interface on the computer. BTW: The silent time between tracks is deliberately printed with a bigger distance between the turns of the groove to create that visual cue.
Slashdot Mirror
1st of April appears to be troll rehabilitation day. Free karma for everyone who asks for it.
I have not laughed that hard the whole day. Seriously. Yawn.
Step #2: Congratulations, you have found Slashdot. Now present a theory and have the /. users write down the practical ideas which will be "Leech Computing, Part 2".
There is a difference between being in a public space, being watched and having your every move recorded.
First, if you are in a public space, usually nobody pays attention to you except for those who are interacting with you. That's fine.
Second, being watched: This is already an unusual situation, in which many people feel uncomfortable. Think stars: They regularly complain about lack of privacy. Being watched usually has greater impact on the behaviour of people because they notice that they are watched, which is not the case with CCTV cameras. One of the activities of anti-surveillance groups is therefore to create the same level of awareness by showing people that they are watched (pointing to the camera, creating camera-maps, etc).
Third, recordings: If you think about it, you not only have to relate your actions to the current situation but you also need to think about how your actions may look like from a distance, space and timewise. This can seriously inhibit natural behaviour and is the reason why, for example, many people liked usenet until archives of it were created.
Some people feel that even perfectly normal behaviour can have a negative impact on them when seen in a different, maybe willfully distorting, context.
DSL is capable of several MBits/s. There are different types of DSL, but for ADSL, an FAQ mentions up to 8 MBit downstream and 1 MBit upstream. Since residential customers usually pay for and get less than that, there is extra bandwith which could be used for providing wireless access.
This is the same thing. I was misled by the "Rolling DSL and Wireless Access Out In One Swoop" headline. Nevertheless, I think that using otherwise unused bandwith from customer DSL lines (the difference between what the customer is paying for and what the DSL-modem is capable of) to provide wireless access to an area is quite an intriguing idea.
Ok, looks like the article is a little misleading. "Wireless coverage" is not really the goal of this approach, but merely the means to provide residential broadband internet access. It makes connecting distant customers feasible because up to 40 subscribers can be connected to one one "airhead", which is a special box connected to an uplink, which in turn can be either wired or wireless (but not through other airheads). The cost of installing an uplink can therefore be split among 40 customers, and because the "airhead" is so small, renting an office for hub hardware isn't necessary. This whole concept doesn't appear to be based on 802.11b at all.
Sorry, now I get it. I had thought they wanted to provide a wireless network as a bonus to normal DSL. Stupid me. I guess I'll watch the demonstration from Nokia before I continue commenting.
If they do it in a responsible way, they split the bandwith (x MBit for the customer, rest for the WAP), but without sharing the broadcast domain. DSL modems usually have an ATM mode which could be used to create separate channels, guaranteeing QOS to the customer and keeping the wireless traffic away from his network socket.
Yes, but typically a DSL connection is not using the maximum possible bandwith. They probably use a fixed part of the bandwith for the paying customer and any excess bandwith is allocated to the wireless access point.
There are quite a lot of reasons not to use .doc as the file format:
One file format which does fit most of the requirements is PDF.
They do have a website (but it's quite minimalistic, no fotos). I also found this introductory text (use the fish). There you'll find some more info about the installations and links to the otherwise hidden membership applications on the Automatenbar website.
Current codecs are a lot more efficient than you think. 32MB per minute is about 500kB/s, which is enough to transmit almost DVD-quality video. For small displays, like those of handhelds, the rough estimate is more like 1 or 2 MB per minute of video.
Maybe if they had just played PI as sound (base 16384) their ratings wouldn't suck. There's some pretty neat stuff in there. Sure, there's a lot of rubbish in between, but...
In a supposedly free world, the question to be asked is why it shouldn't be allowed to exist. The default is "anything goes", right? It doesn't hurt you if you ignore it, unless of course the majority thinks it's cool and buys the stuff. Which brings us back to "Don't buy it if you don't want it." Ensuring that customers make informed decisions really is the key.
If people buy it, so be it. It's a majority vote and those who complain that they haven't been heard on the issue just don't recognize the facts. Sometimes the road is a little longer than what looks like a straight connection from a to b to you. The best mechanism which is available to us is market dynamics, or isn't it? You are free to try to convince others, which is what I'm doing when I say "Make sure you really want it before you buy it", but the voting is done with the money being spent on what people want and therefore buy.
I'm referring to the devices and the content. If the device doesn't let you do what you want to do at a price which you are willing to pay, don't hope that somehow someone will offer to change that. If a device is crippled with unavoidable DRM, look for alternatives and if there aren't any, don't buy. People need to look at the whole product, the device and the service which is offered (not what is promised to be available in 6 months).
Don't buy what you don't want. It is that simple.
You can't build the equivalent of a high end Cisco router with off-the-shelf PC hardware. PCI (the 33MHz 32Bit variety found in today's PCs) maxes out at 133 MByte/s in theory. Practical values are much lower than that, so there's no way to route a gigabit stream through a standard PC.
BUT: They don't need to. "Because of the local area, unrouted nature of the network, Citylink can use inexpensive Cisco 3524 switches instead of more costly T1 routers". They only use routers at the edges of the network, where speeds are much less than 1 GBit/s. And even if someone needed the full speed there, the router is the customer's hardware, they don't have to go for the 2500$ PC hardware thing if they need more.
The problem with this approach may be scalability: "De Wit says adding quality of service (QoS) features isn't necessary." and "We can fit all the traffic we want onto our Ethernet, so why do we need to worry about prioritizing?" Also, ethernet switching doesn't work beyond a certain network size. Sooner or later, they will have to route. If they're lucky, gigabit router prices have dropped, 10GBit ethernet is ready and everyone accepts the upgrade cost by the time the assumption "The network is all switched and offers all the speed we need, don't worry" is no longer true.
But wouldn't you have to reveal who used a given ip address at a specified time in case law enforcement asks you for it? How would you do that, considering network access is anonymous, and failing that, wouldn't you become liable yourself?
Let's suppose you provided freely accessible wireless network infrastructure to anonymous users, what would be the legal implications of such an endeavour? Would you become liable for damage created with the help of your network? What if you provided no access to the internet? Would you still be liable? Is there a way around these problems? I guess once again lawyers may be the biggest hurdle before public networks become a reality.
Sorry to disappoint you, but automatic parallelization of algorithms is a field of research in which quite a few obstacles still need to be overcome before something like you propose would enter the mainstream. What does work right now is the distribution of already parallel processes/threads to several cpus, but that requires either several applications to run or a multithreaded application. For anything worth running on multi-processor hardware, the parallelism inside the operating system gives only negligible advantages.
I'm supposed to download a program that sits in the tray and records all the keys I press? I mean, to count them, they all have to pass their code, right? And it has network functionality... HELLO? Security?
Wireless networking on the whole campus is nice, of course, but it isn't educational if there is no educational content or projects which make use of the network. Looking at stories about the bandwith demand at universities, I guess the networks are mostly there (although not always wireless), but the on-topic content is missing. I'd say, put the money into virtualizing lecture material and developing new forms of presenting educational material. Some things can be expressed much better in an animation or interactive 3d-model for example, ways of presentation which are usually unvailable today.
There's another thing vinyl DJs might miss: The switching of real records. Grabbing a record from your set and selecting the song by setting the needle to the right part of it (using the visual cues you described) is probably faster than relying on a scroll-point-click-point-click interface on the computer. BTW: The silent time between tracks is deliberately printed with a bigger distance between the turns of the groove to create that visual cue.