The interfaces that 2.4.x exposed to the world have been quite stable during the branch's lifetime. However, the internal interfaces have and always willbe subject to change, and this break VMWare's kernel modules.
It is not the role of the kernel maintainers to cater to commerical software vendors. The kernel hackers want to push out the best standalone product possible, and this involves making modifications to the internal architecture when nessicary. These sorts of changes will break compatibility with kernel modules external to the kernel tree. They are quite aware of this, and are knowingly making the comprimise
Anyways, if you want the kernel evolution to standstill just so your stupid commerical apps will always run the way they always have, why don't you just refrain from upgrading your kernel?
Although you're ultimately correct, your example is a little deceptive. In fact, for all pratical purposes, an anonymous P2P network can be built.
In your examples (IP address, postal address), address were used to uniquely and permantely identify a person or other entity. This is fast and efficent for networks that are not required to function anonymouly. The router 'knows' which router it needs to send a packet to to get it closer to it's final destination, and a postal office knows which other piece of postal infrastructure to send a letter off to to get it closer to the person. However, this is not the only way to do addressing.
It is possible to do addressing on the basis of 'conversations'. Your computer sends a packet off to the P2P network with a unique(ish) idenitfying number. The other nodes of the P2P network broadcast your packet to each other, but each node remembers which node immediately before them sent them the packet. This builds sort of a 'temporary routing entry', which allows each router to appropriately route responses closer to where they have to be. However, each router only knows the identity of the router that came before it, so you would need to have control of all the routers in the route from the originating host to to your host to fully trace a packet.
Although it is still possible to find the identity of an originating host, it's entirely impractical in all but an extremely comprimised network, where the attacker controls a large number of nodes. Also, it's not only impratical to associate a packet with an originating internet host, it's also impratical to tell if two packets are originating from the same node, because the addressing scheme is conversation based instead of node based.
I have a strong feeling that this is exactly how Freenet operates, but I don't know, as I haven't done much research in to it. This is just how I would architect a P2P network.
I was just reading this article, and thinking about how much I would not implement portable internet gaming that way. At all. Maybe I'm a Unix weenie, but I think the philosophy of "do one thing and do it well" applies here. This is how I'd do it:
Do not put games on a cell phone. The cell phone is a communications device, and should be nothing more. Needlessly bloating it with features will make the user find it klunky and complicated to use. Also, it'd be much harder to properly design and test a device that performs so many different functions. Nobody wants a cell phone with a cucumber slicer if it bluescreens every 5 minutes
When the phone is on, provide a local Bluetooth IPv6 network (look, now I'm buzz-word compliant). It could IPv6 autoregister an IP from the cell phone provider servers, and use that for the duration of the phone being on. Other devices could hook in to this wireless network and use the cell phone as a gateway to the internet (keep in mind IPv6->IPv4 and back is nearly trivial once the infrastructure is up).
Make the gaming device use this internet field around the cell phone. It should be about cell phone size, maybe a bit bigger for easier handling. The screen could be much larger, and less of the device would be dedicated to buttons. It would be like a cross between a Gameboy pocket (a bit too large to be uberportable) and a Sega VMU (a bit too small to be fun to use). You would probably want to throw in a color screen. Firstly, for how impressive they look, and secondly because when a screen is relatively small, you want to get as much information packed in there as possible. Clever use of color could actually improve gameplay.
For the love of god, don't use Java. Java has it's place, but not in portable gaming devices, were every cycle counts. Make a standard gaming system specification everyone could live with. I'd look in to something resembling a fast m68k compatible processor, 256KB of RAM with impressive memory bandwidth, and a custom 2D chip with respectable on-board 2D rasterization abilities. Any game worth playing could be implemented on a system like that (it'd end up being slightly more powerful than a SNES), and the hardware could probably be mostly implemented on a single chip
Then again, I'd probably be one of the only people in the known world that would enjoy playing Chu Chu Rocket Embedded infinitely more because I knew the technology was cleanly implemented.
I'm no troll, but the human eye can't even tell the difference between that many colors.
Well, you might not be a a troll, but you're painfully uninformed.
This article was not about a display with ridiculous color depth, but one with a high resolution
The human eye can distinguish between more than 9.2 million colors
Modern displays cannot produce images anywhere near what the human eye is capable of perceiving. We are very visual animals, and our eyes are quite precise organs. Although printers and cameras have acheived very high resolutions, monitors still more or less suck. This display can do about 3000x3000 (if it was square), which is quite an improvement, but I wouldn't call it revolutionary.
The claim that it requires 16 Pentium 4's to drive seems a bit dubious, either producing images on the screen requires a lot of math, or the media is just confused again. The latter seems more likely, but anyone have any ideas what sort of display technology would require that sort of CPU horsepower?
Slashdot Mirror
The interfaces that 2.4.x exposed to the world have been quite stable during the branch's lifetime. However, the internal interfaces have and always willbe subject to change, and this break VMWare's kernel modules.
It is not the role of the kernel maintainers to cater to commerical software vendors. The kernel hackers want to push out the best standalone product possible, and this involves making modifications to the internal architecture when nessicary. These sorts of changes will break compatibility with kernel modules external to the kernel tree. They are quite aware of this, and are knowingly making the comprimise
Anyways, if you want the kernel evolution to standstill just so your stupid commerical apps will always run the way they always have, why don't you just refrain from upgrading your kernel?
-Bodnar42
Although you're ultimately correct, your example is a little deceptive. In fact, for all pratical purposes, an anonymous P2P network can be built.
In your examples (IP address, postal address), address were used to uniquely and permantely identify a person or other entity. This is fast and efficent for networks that are not required to function anonymouly. The router 'knows' which router it needs to send a packet to to get it closer to it's final destination, and a postal office knows which other piece of postal infrastructure to send a letter off to to get it closer to the person. However, this is not the only way to do addressing.
It is possible to do addressing on the basis of 'conversations'. Your computer sends a packet off to the P2P network with a unique(ish) idenitfying number. The other nodes of the P2P network broadcast your packet to each other, but each node remembers which node immediately before them sent them the packet. This builds sort of a 'temporary routing entry', which allows each router to appropriately route responses closer to where they have to be. However, each router only knows the identity of the router that came before it, so you would need to have control of all the routers in the route from the originating host to to your host to fully trace a packet.
Although it is still possible to find the identity of an originating host, it's entirely impractical in all but an extremely comprimised network, where the attacker controls a large number of nodes. Also, it's not only impratical to associate a packet with an originating internet host, it's also impratical to tell if two packets are originating from the same node, because the addressing scheme is conversation based instead of node based.
I have a strong feeling that this is exactly how Freenet operates, but I don't know, as I haven't done much research in to it. This is just how I would architect a P2P network.
-Bodnar42
I was just reading this article, and thinking about how much I would not implement portable internet gaming that way. At all. Maybe I'm a Unix weenie, but I think the philosophy of "do one thing and do it well" applies here. This is how I'd do it:
Then again, I'd probably be one of the only people in the known world that would enjoy playing Chu Chu Rocket Embedded infinitely more because I knew the technology was cleanly implemented.
-RyanWell, you might not be a a troll, but you're painfully uninformed.
Modern displays cannot produce images anywhere near what the human eye is capable of perceiving. We are very visual animals, and our eyes are quite precise organs. Although printers and cameras have acheived very high resolutions, monitors still more or less suck. This display can do about 3000x3000 (if it was square), which is quite an improvement, but I wouldn't call it revolutionary.
The claim that it requires 16 Pentium 4's to drive seems a bit dubious, either producing images on the screen requires a lot of math, or the media is just confused again. The latter seems more likely, but anyone have any ideas what sort of display technology would require that sort of CPU horsepower?
-Bodnar42
No. I saw a photograph of Sealand once, thus some of the data went directly to the USA without going through the EU Internet infrastructure.
Wouldn't that be implying that some of Sealand's data went directly to the USA because you saw a photograph of Sealand?
Sorry, I couldn't help but leap on a grammar error by the infamous grammar nazi...
-Bodnar42